2018 healthatglance rep en

Transcript

1 Health at a Glance: Europe 2018 STATE OF HEALTH IN THE EU CYCLE Health at a Glance: Europe 2018 STATE OF HEALTH IN THE EU CYCLE

2

3 Health at a Glance: Europe 2018 STATE OF HEALTH IN THE EU CYCLE

4 This work is published under the responsibility of the Secretary-General of the OECD. This publication has been produced with the financial and substantive assistance of the European Union. The opinions expressed and arguments employed herein do not necessarily reflect the official views of OECD member countries or the European Union. This document, as well as any data and any map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Please cite this publication as: , OECD Publishing, Paris. Health at a Glance: Europe 2018: State of Health in the EU Cycle OECD/EU (2018), https://doi.org/10.1787/health_glance_eur-2018-en ISBN 978-92-64-30334-8 (print) ISBN 978-92-64-30335-5 (PDF) ISBN 978-92-64-30686-8 (HTML) ISBN 978-92-64-30685-1 (epub) Series: Health at a Glance: Europe ISSN 2305-607X (print) ISSN 2305-6088 (online) European Union ISBN 978-92-79-88852-6 (print) ISBN 978-92-79-88853-3 (PDF) Catalogue number: EW-01-18-697-EN-C (print) Catalogue number: EW-01-18-697-EN-N (PDF) Revised version, February 2019. Details of revisions available at: www.oecd.org/about/publishing/Corrigendum_Health_at_a_Glance_Europe_2018.pdf The information in this document with reference to “Cyprus” relates to the southern part of the Note by Turkey: Island. There is no single authority representing both Turkish and Greek Cypriot people on the Island. Turkey recognises the Turkish Republic of Northern Cyprus (TRNC). Until a lasting and equitable solution is found within the context of the United Nations, Turkey shall preserve its position concerning the “Cyprus issue”. Note by all the European Union Member States of the OECD and the European Union: The Republic of Cyprus is recognised by all members of the United Nations with the exception of Turkey. The information in this document relates to the area under the effective control of the Government of the Republic of Cyprus. Photo credits: Cover © baselinearts.co.uk. Corrigenda to OECD publications may be found on line at: www.oecd.org/about/publishing/corrigenda.htm. © OECD/European Union 2018 You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications, databases and multimedia products in your own documents, presentations, blogs, websites and teaching materials, provided that suitable acknowledgment of the source and copyright owner is given. All requests for public or commercial use and translation rights should be [email protected] . Requests for permission to photocopy portions of this material for public or commercial use shall be submitted to or the Centre français d’exploitation du droit de copie addressed directly to the Copyright Clearance Center (CCC) at [email protected] . [email protected] (CFC) at

5 FOREWORD Foreword Health at a Glance: Europe marks the start of a new State of Health in This 2018 edition of the EU cycle by the European Commission designed to assist EU Member States in improving the health of their citizens and the performance of their health systems. Two overarching trends warrant special mention. First, the steady increase in life expectancy has slowed considerably in many EU countries due to a slower rate of reduction of cardiovascular deaths and an increase in the number of deaths among the elderly during winter months in recent years. Second, large inequality in life expectancy persists. Across the EU, people with a low level of education can expect to live six years less than those with a high level of education. More than 1.2 million people die prematurely We need more protection and prevention. every year in EU countries – this could be avoided through better disease prevention policies and more effective health care interventions. On the one hand, we must tackle the misinformation about vaccines and address population hesitancy about childhood vaccination, as outlined in the recommendation proposed to the Council of the EU earlier this year. At the same time, many lives could be saved by redoubling efforts to prevent unhealthy lifestyles. Some 790 000 EU citizens die prematurely each year from tobacco smoking, alcohol consumption, unhealthy diets and lack of physical activity. Policies to control tobacco and harmful consumption of alcohol or to halt obesity therefore need to be actively pursued. also makes a strong case for promoting mental Health at a Glance: Europe This edition of health and preventing mental illness. The total costs of mental health problems – which include the costs to health systems and social security programmes, but also lower employment and worker productivity – are estimated to amount to more than 4% of GDP across EU countries, equivalent to over EUR 600 billion per year. Promoting mental health and improving access to treatment for people with poor mental health should be a priority. Health systems have We need more effective and people-centred health systems. achieved remarkable progress in treating life-threatening diseases such as heart attacks, strokes and various cancers, yet wide disparities in survival rates persist not only between countries but also among hospitals and health care providers within each country. It is not enough to only collect data on mortality. Health care needs to place people at the centre, which requires asking patients more systematically whether they are better, or worse, following different health care interventions. We must also measure how well the primary care sector is managing the growing number of people living with one or more chronic conditions. The OECD and the European Commission are working together with countries to fill these critical data gaps on patient-reported experience and outcome measures. We need to improve access to health care. Universal health coverage – a key Sustainable Development Goal – and timely access to affordable, preventive and curative health care – a key European Pillar of Social Rights principle of the – should remain central to policy action. Recent data on the unmet health care needs are encouraging; fewer EU citizens report foregoing care due to HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 3

6 FOREWORD financial reasons, distance from services or waiting times. The gap between the poor and the wealthy, however, remains too large. Poor Europeans are on average five times more likely to have problems accessing health care than richer ones, and policies must prioritise financial protection for disadvantaged groups. Finally, we need more resilient health systems. As health systems evolve, they must become more resilient and adapted to rapidly changing environments and needs. In this edition of Health at a Glance: Europe, we highlight the importance of reducing wasteful spending, and the potential gains for efficiency and sustainability of health systems. Evidence from various countries suggests that up to one-fifth of health spending is wasteful and could be reallocated to better use. For example, too many hospital admissions reflect failures in the management of health problems in the community and consume over 37 million bed days each year across the EU. The digital , offers transformation of health and care, a key component of the EU’s Digital Single Market tremendous potential for improving the prevention, detection and management of chronic diseases, as well as improving health system management and research. The OECD and the European Commission will work closely together with policymakers and cycle, to help promote policies that other key stakeholders throughout the State of Health in the EU will deliver both longer and healthier lives for all EU citizens. Angel Gurría Vytenis Andriukaitis Secretary-General European Commissioner for Health and Food Safety Organisation for Economic Co-operation and Development HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 4

7 ACKNOWLEDGEMENTS Acknowledgements H ealth at a Glance: Europe 2018 , the first step in the State of Health in the EU cycle, is the result of close co-operation between the OECD and the European Commission.The preparation of this publication was led by the OECD, and the Commission provided guidance and technical support throughout the process. This publication would not have been possible without the effort of national data correspondents from the 36 countries who have provided most of the data and the metadata presented in this report, and financial support provided by the European Union. This publication also benefitted from many useful comments from members of the Commission’s Expert Group on Health Information (EGHI), as well as from several officials in the European Centre for Disease Prevention and Control (ECDC), the Joint Research Centre (JRC) and the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). This report was prepared by a team from the OECD Health Division under the co-ordination of Gaétan Lafortune. Chapter 1 on mental health promotion was prepared by Emily Hewlett, Gaétan Lafortune, Eileen Rocard and David Morgan. The preparation of Chapter 2 on strategies to reduce wasteful spending in health systems was led by Agnès Couffinhal, Ruth Lopert and Gaétan Lafortune, with inputs by Elina Suzuki, Michael Padget, Gaëlle Balestat, Eileen Rocard, Marie-Clémence Canaud and Roi Meshulam. Chapter 3 on the health status of populations was prepared by Eileen Rocard and Gaétan Lafortune. Chapter 4 on risk factors to health was prepared by Marion Devaux and IvanTzintzun, with inputs by Joao Matias and Katerina Skarupova from the EMCDDA for indicators related to alcohol consumption among children and illegal drug consumption among both children and adults, and with inputs from Marta Buoncristiano and Joao Breda from WHO Europe for the indicator on childhood obesity. Michael Mueller, James Cooper, David Morgan and Jens Wilkens prepared Chapter 5 on health expenditure and financing, with input by Sebastiano Lustig. Chapter 6 on effectiveness (including quality of care and patient experience) was prepared by Rie Fujisawa, with inputs by Eileen Rocard and Gaétan Lafortune for the indicator on avoidable mortality, by Yuka Nishina for the indicators on cancer care and by Michael Padget from the OECD Health Division and Carl Suetens from the ECDC for the indicator related to health care-associated infections. Chapter 7 on accessibility was prepared by Gaétan Lafortune, Gaëlle Balestat, Michael Mueller and Marie-Clémence Canaud, with input by Jon Cylus, Sarah Thomson and Tamas Evetovits from the European Observatory on Health Systems and Policies for the indicator related to the financial goods. burden of out-of-pocket spending for health services and Chapter 8 on the r esilience of health systems was prepared by Luke Slawomirski, Gaétan Lafortune, David Morgan and Gaëlle Balestat, with inputs by Marc Struelens and Katrin Leitmeyer from the ECDC for the indicator related to public health laboratory capacity. Editorial assistance was provided by Ruth Lopert, Marie-Clémence Canaud, Lucy Hulett and Kate Cornford (on Chapter 1). This HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 5

8 ACKNOWLEDGEMENTS publication also benefited from useful comments from Francesca Colombo, Chris James and Valérie Paris from the OECD Health Division. Many useful comments were also received from the European Commission. Special thanks go to DG SANTE’s State of Health in the EU team for their advice during the project and the co-ordination of inputs from different officials across the Commission. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 6

9 TABLE OF CONTENTS Table of contents ... 11 Executive summary Readers’ guide ... 15 Part I tic chapters on public health and health care issues Thema Promoting mental health in Europe: Why and how? ... 19 Chapter 1. Intr oduction ... 2 0 illness affects tens of millions of Europeans every year ... 21 Mental The costs of mental health problems exceed 4% of GDP ... 26 Actions to promote mental health and prevent mental illness in Europe ... 32 Conclusions ... 39 Note ... 40 References ... 40 Chapter 2. Strategies to reduce wasteful spending: Turning the lens to hospitals and pharmaceuticals ... 45 ... 46 oduction Intr Addr ... 47 essing wasteful spending in hospitals Addressing wasteful spending on pharmaceuticals ... 60 Conclusions ... 72 Notes ... 73 References ... 74 Part II Overview of health indicators Chapter 3. Health status ... 81 T rends in life expectancy ... 82 Inequalities in life expectancy ... 84 Healthy life expectancy at birth and at age 65 ... 86 Main causes of mortality ... 88 Mortality from circulatory diseases ... 90 Mortality from cancer ... 92 Mortality from respiratory diseases ... 94 Infant health ... 96 Self-reported health and disability ... 98 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 7

10 TABLE OF CONTENTS Notified cases of vaccine-preventable diseases ... 100 ... 102 New reported cases of HIV and tuberculosis ... 104 Cancer incidence Diabetes prevalence ... 106 Dementia prevalence ... 108 Risk factors ... 111 Chapter 4. en... 112 Smoking among c hildr ... 114 Smoking among adults Alcohol consumption among children ... 116 Alcohol consumption among adults ... 118 Illicit drug consumption among children ... 120 Illicit drug consumption among adults ... 122 Obesity among children ... 124 Obesity among adults ... 126 Mortality due to air pollution and extreme weather conditions ... 128 Chapter 5. Health expenditure and financing ... 131 ... 132 per capita e Health expenditur Health expenditure in relation to GDP ... 134 Health expenditure by type of good and service ... 136 Health expenditure in hospitals ... 138 Pharmaceutical expenditure ... 140 Financing of health expenditure ... 142 ... 145 Chapter 6. Effectiveness: Quality of care and patient experience le mortality (preventable and amenable) ... 146 Avoida b Childhood vaccinations ... 148 Patient experience with ambulatory care ... 150 Mortality following acute myocardial infarction (AMI) ... 152 Mortality following stroke ... 154 Waiting times for hip fracture surgery ... 156 Screening, survival and mortality for cervical cancer ... 158 Screening, survival and mortality for breast cancer ... 160 Survival and mortality for colorectal cancer ... 162 Late-diagnosed HIV and tuberculosis treatment outcomes ... 164 Healthcare-associated infections ... 166 Chapter 7. Accessibility: Affordability, availability and use of services ... 169 needs ... 170 e Unmet health car Financial burden of out-of-pocket expenditure ... 172 Population coverage for health care ... 174 Extent of health care coverage ... 176 Availability of doctors ... 178 Availability of nurses ... 180 Consultations with doctors ... 182 Availability and use of diagnostic technologies ... 184 Hospital beds and discharges ... 186 Waiting times for elective surgery ... 188 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 8

11 TABLE OF CONTENTS Chapter 8. ... 191 Resilience: Innovation, efficiency and fiscal sustainability Electronic Medical Records and ePrescribing ... 192 Adoption and use of Individuals using the internet to access health services and health information . . 194 Public health laboratory capacity to control infectious diseases threats ... 196 Average length of stay in hospital ... 198 Daysurgery... 200 Capital expenditure in the health sector ... 202 Projections of public expenditure on health and long-term care ... 204 Statistical annex ... 207 Follow OECD Publications on: http://twitter.com/OECD_ Pubs http://www.facebook.com/OECDPublications http://www.linkedin.com/groups/OECD-Publications-4645871 http://www.youtube.com/oecdilibrary OECD lerts A http://www.oecd.org/oecddirect/ StatLinks 2 This book has... ® A service that delivers Excel files from the printed page! Look for the StatLinks 2 at the bottom of the tables or graphs in this book. To download the matching Excel® spreadsheet, just type the link into your prefix, or click on the link from Internet browser, starting with the http://dx.doi.org the e-book edition. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 9

12

13 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 Executive summary H ealth at a Glance: Europe 2018 presents comparative analyses of the health status of EU citizens and the performance of the health systems of the 28 EU Member States, 5 candidate countries and 3 EFTA countries. It is the first step in the State of Health in the EU cycle of knowledge brokering. This publication has two parts. Part I comprises two thematic chapters, the first focusing on the need for concerted efforts to promote better mental health, the second outlining possible strategies for reducing wasteful spending in health. In Part II, the most recent trends in key indicators of health status, risk factors and health spending are presented, together with a discussion of progress in improving the effectiveness, accessibility and resilience of European health systems. Making the case for greater priority to improving mental health ● Mental health is critical to individual well-being, as well as for social and economic participation. Yet, according to recent estimates, more than one in six people across EU countries had a mental health issue in 2016, equivalent to about 84 million people. Moreover, in 2015 the deaths of more than 84 000 people in EU countries were attributed to mental illness or suicide. “The total costs of mental ill-health are estimated at more than 4% of GDP – or over EUR 600 billion – across the 28 EU countries” ● The economic and social costs of mental illness are substantial.The total costs of mental ill-health are estimated at more than 4% of GDP – or over EUR 600 billion – across the 28 EU countries. EUR 190 billion (or 1.3% of GDP) reflects direct spending on health care, another EUR 170 billion (1.2% of GDP) is spent on social security programmes, while a further EUR 240 billion (1.6% of GDP) represents indirect costs to the labour market due to lower employment and productivity. ● The heavy individual, economic and social burdens of mental illness are not inevitable. Many European countries have in place policies and programmes to address mental illness at different ages. However, much more can be done to manage and promote mental health. Reducing wasteful spending to make health systems more effective and resilient “Evidence from various countries suggests that up to one-fifth of health spending is wasteful and could be reallocated to better use” ● Wasteful spending occurs when patients receive unnecessary tests or treatments or when care could have been provided with fewer and less costly resources. Evidence from various countries suggests that as much as one-fifth of health spending is wasteful and could be reduced or eliminated without undermining quality of care. Reducing wasteful 11

14 EXECUTIVE SUMMARY spending not only contributes to health system resilience, but helps achieve and maintain universal access to effective care. ● When it comes to hospitals, many admissions could be avoided with better management of chronic conditions in the community. Potentially avoidable admissions for conditions such as asthma and diabetes consume over 37 million bed days each year across the EU. Unnecessarily delayed discharges are also costly for hospitals, and many discharge- ready patients occupy beds that could be used for patients with greater needs. ● When it comes to pharmaceuticals, minimising waste and optimising the value derived from medicine spending are also critical to achieving efficient and sustainable health systems. A mix of policy levers can support this goal, including: 1) ensuring value for money in the selection and coverage, procurement and pricing of pharmaceuticals through Health Technology Assessment; 2) exploiting the potential savings from generics and biosimilars; 3) encouraging rational prescribing; and 4) improving patient adherence. Gains in life expectancy have slowed in many EU countries, and large inequalities persist ● While life expectancy increased by at least 2 to 3 years over the decade from 2001 to 2011 in all EU countries, the gains have slowed down markedly since 2011 in many countries particularly in Western Europe, increasing by less than half a year between 2011 and 2016. This slowdown appears to have been driven by a slowdown in the rate of reduction of deaths from circulatory diseases and periodical increases in mortality rates among elderly people due partly to bad flu seasons in some years. “People with a low level of education can expect to live six years less than those with a high level of education” ● Large disparities in life expectancy persist not only by gender, but also by socioeconomic status. On average across the EU, 30-year-old men with a low level of education can expect to live about 8 years less than those with a university degree (or the equivalent), while the “education gap” among women is narrower, at about 4 years. These gaps largely reflect differences in exposure to risk factors, but also indicate disparities in access to care. Putting a greater focus on preventing risk factors ● While smoking rates in both children and adults have declined in most EU countries, about one-fifth of adults still smoke every day, and as many as one in four in countries with less advanced tobacco control policies. ● Alcohol control policies have reduced overall alcohol consumption in several countries, but heavy alcohol consumption among adolescents and adults remains an important public health issue. In EU countries, nearly 40% of adolescents report at least one “binge drinking” event in the preceding month, and more than 40% of young men aged 20-29 also report heavy episodic drinking. “At least one in six adults are obese across EU countries, with wide disparities by socioeconomic status” The prevalence of obesity continues to increase among adults in most EU countries, with at ● least one in six defined as obese. Inequality in obesity remains marked: 20% of adults with a lower education level are obese compared with 12% of those with a higher education. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 12

15 EXECUTIVE SUMMARY Strengthening the effectiveness of health systems can reduce premature mortality “More than 1.2 million deaths could have been avoided in EU countries in 2015 through better public health policies or more effective and timely health care” ● More than 1.2 million people in EU countries died in 2015 from diseases and injuries that could have been avoided either through stronger public health policies or more effective and timely health care. ● Vaccine-preventable diseases have resurged in some parts of Europe in recent years, pointing to the importance of promoting effective vaccination coverage for all children across all EU countries. It is estimated that 790 000 people in EU countries died prematurely in 2016 due to ● tobacco smoking, harmful consumption of alcohol, unhealthy diets and lack of physical activity. ● The quality of acute care for life-threatening conditions has improved in most countries over the past decade. Fewer people die following a hospital admission for acute myocardial infarction (a 30% reduction on average between 2005 and 2015) or stroke (a reduction of over 20% during this same period). However, wide disparities in the quality of acute care persist not only between countries but also between hospitals within each country. Remarkable progress has also been achieved in cancer management through the ● implementation of population-based screening programmes and the provision of more effective and timely care. Survival rates for various cancers have never been higher, yet there is still considerable room for further improvement in cancer management in many countries. Ensuring universal access to care is critical to reducing health inequalities “Unmet health care needs are generally low in EU countries, but low-income households are five times more likely to report unmet needs than high-income households” Unmet health care needs are an important measure of accessibility. Recent survey data ● show that in most EU countries the share of the population reporting unmet care needs is generally low and has declined over the past ten years. Yet, low-income households are still five times more likely to report unmet care needs than high-income households, mainly for financial reasons. ● In addition to being affordable, health services must also be accessible when and where people need them. While the numbers of doctors and nurses in nearly all EU countries have increased over the past decade, shortages of general practitioners are common, particularly in rural and remote areas. ● Long waiting times for elective surgery is an important policy issue in many EU countries as it impedes timely access to care. In many of these countries, waiting times have worsened in recent years as the demand for surgery has increased more rapidly than the supply. Strengthening the resilience of health systems Health systems need to respond more efficiently to changing health care needs driven by ● demographic changes and exploit more fully the potential of new digital technologies to strengthen prevention and care. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 13

16 EXECUTIVE SUMMARY ● In 2017, health spending accounted for 9.6% of GDP in the EU as a whole, up from 8.8% in 2008. Population ageing means not only that health care needs will increase in the future, but also that there will be increasing demand for long-term care. Indeed, spending on long-term care is expected to grow faster than spending on health care. “New digital technologies have the potential to promote more healthy ageing and more people- centred care” ● New digital technologies offer great opportunities to promote healthy ageing and achieve more efficient and people-centred care. The use of Electronic Medical Records and ePrescribing is growing across EU countries, and growing numbers of EU residents use the internet to obtain health information and access health services, although there are disparities by age and socioeconomic groups. Population ageing requires profound transformations in health systems, from a focus on ● acute care in hospitals to more integrated and people-centred care in the community. Many EU countries began this transformation over a decade ago – for example by reducing hospital capacity and average length of stay, and strengthening community care – but the process still requires ongoing, long-term effort. Monitoring and improving the State of Health in the EU Health at a Glance: Europe 2018 is the result of ongoing and close collaboration between the OECD and the European Commission to improve country-specific and EU-wide cycle. knowledge on health issues as part of the Commission’s State of Health in the EU State of Health in the EU cycle to assist EU In 2016, the European Commission launched the Member States in improving the health of their citizens and the performance of their health systems. Health at a Glance: Europe is the first product of the two-year cycle, presenting every even-numbered year extensive data and comparative analyses that can be used to identify both the strengths and the opportunities for improvement in health and health systems. The second step in the cycle is the for all EU countries. The next Country Health Profiles edition of these profiles will be published in 2019 jointly with the European Observatory on Health Systems and Policies, and will highlight the particular characteristics and challenges for each country. After a Companion Report that the European Commission presents along with the profiles, the final step in the cycle is a series of Voluntary Exchanges with Member States. These are opportunities to discuss in more detail some of the challenges and potential policy responses. ec.europa.eu/health/state . Info: HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 14

17 READERS’ GUIDE Readers’ guide ealth at a Glance: Europe 2018 presents key data and analysis of health and health H systems in the 28 EU member states, 5 candidate countries and 3 European Free Trade Association countries. The publication is divided in two parts. Part I contains two thematic chapters focussing on important, but often neglected, public health and health care issues. The first chapter assesses the health and economic burden of mental health problems across EU countries, making the case for greater efforts to promote better mental health at all ages. The second chapter looks at wasteful spending in health systems, focussing in particular on hospitals and pharmaceuticals, and reviewing possible strategies to reduce waste to promote a better allocation of resources. Part II includes six chapters providing an overview of key indicators of health and health systems, based to a large extent on the European Core Health Indicators (ECHI) shortlist ( https://ec.europa.eu/health/indicators/echi/list_en ). The structure of the last three chapters is based on the 2014 Commission Communication on effective, accessible and resilient health systems ( https://ec.europa.eu/health/sites/health/files/healthcare/docs/ ).Newindicatorshavebeenincludedinthiseditiontoreflect com2014_215_final_en.pdf different aspects of the effectiveness, accessibility and resilience of health systems. The data presented in this publication come mostly from official national statistics, and have been collected in many cases through the administration of joint questionnaires by the OECD, Eurostat and WHO. The data have been validated by the three organisations to ensure that they meet high standards of data quality and comparability. Some data also come from European surveys co-ordinated by Eurostat, notably the European Union Statistics on Income and Living Conditions Survey (EU-SILC) and the second wave of the European Health Interview Survey (EHIS), as well as from the European Centre for Disease Prevention and Control (ECDC), the European Commission’s Joint Research Centre (JRC), and other sources. Presentation of indicators and calculation of EU averages With the exception of the first two thematic chapters, all indicators in the rest of the publication are presented over two pages. The first page provides a brief commentary highlighting the key findings conveyed by the data, defines the indicator and signals any significant data comparability limitation. On the facing page is a set of figures. These typically show current levels of the indicator and, where possible, trends over time. For those countries that have a relatively small population (less than 1 million), three-year averages are often calculated to minimise random errors due to small numbers. The average in the figures includes only EU member states and is generally calculated population-weighted average as a of all the EU member states presented (up to 28 if there is full data coverage). In some cases, the average is calculated based on the unweighted HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 15

18 READERS’ GUIDE average of EU countries, notably when there is missing data for several countries or when the data owners have already calculated and reported unweighted EU averages. Population figures The population figures used to calculate rates per capita and the population-weighted EU averages come from the Eurostat demographics database. The data were extracted in early June 2018 and relate to mid-year estimates (calculated as the average between the beginning and end of the year). Population estimates are subject to revision, so they may differ from the latest population figures released by Eurostat or national statistical offices. Data limitations Limitations in data comparability are indicated both in the text (in the box related to “Definition and comparability”) as well as in footnotes underneath the charts. Data sources Readers interested in using the data presented in this publication for further analysis and research are encouraged to consult the full documentation of definitions, sources and for all OECD member countries, including 23 EU OECD Health Statistics methods contained in member states and four additional countries (Iceland, Norway, Switzerland andTurkey).This information is available in OECD.Stat ( http://stats.oecd.org/index.aspx?DataSetCode=HEALTH ). For the nine other countries (Albania, Bulgaria, Croatia, Cyprus, the Former Yugoslav Republic of Macedonia, Malta, Montenegro, Romania and Serbia), readers are invited to http://ec.europa. consult the Eurostat database for more information on sources and methods: eu/eurostat/data/database . Readers interested in an interactive presentation of the European Core Health Indicators (ECHI) can consult DG SANTE’s ECHI data tool at http://ec.europa.eu/health/indicators/indicators/ index_en.htm . Readers interested in indicators that quantify the burden of cancer in Europe can also https://ecis.jrc.ec.europa.eu/ . visit the JRC’s European Cancer Information System (ECIS): HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 16

19 P ART I Thematic chapters on public health and health care issues HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018

20

21 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART I Chapter 1 Promoting mental health in Europe: Why and how? Good mental health is a critical part of individual well-being, and the foundation for happy, fulfilled, productive lives. However, this chapter finds that more than one in six people across EU countries had a mental health problem in 2016. Living with mental ill-health means that individuals are less able to succeed at school and work, are more likely to be unemployed, and may suffer worse physical health. For some, mental illnesses lead to premature mortality: over 84 000 people died of mental health problems and suicides across EU countries in 2015. The economic costs of mental illness are also significant. This chapter estimates total costs related to mental ill-health at more than 4% of GDP – or over EUR 600 billion – across the 28 EU countries in 2015. EUR 190 billion (or 1.3% of GDP) is direct spending on health care, another EUR 170 billion (1.2% of GDP) is spending on social security programmes, while a further EUR 240 billion (1.6% of GDP) is caused by indirect costs in the labour market, driven by lower employment rates and reduced productivity due to mental illness. The heavy economic, social and individual burden of mental illness is not inevitable, and more must be done to prevent and treat mental disorders, and to foster good mental health. The latter part of this chapter explores some effective ways by which European countries are promoting mental well-being and preventing mental illness, and identifies critical gaps where more action is needed. 19

22 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Introduction Good mental health is a critical part of individual well-being, and the foundation for happy, fulfilled, productive lives. Mental ill-health, meanwhile, will affect everyone at some point in their lives – whether experiencing mental illness themselves, or as a family member, friend or colleague of someone living with a mental disorder. Mental ill-health can affect women and men of all ages and backgrounds. Without effective prevention and treatment, mental illnesses can have profound effects on people’s ability to carry out their daily lives and often result in poorer physical health. The impact of poor mental health can affect people throughout their lifetime. Children and adolescents with poor mental health have worse educational outcomes and job opportunities. Adults with mental health problems are less productive at work and more likely to be unemployed. Elderly people with mental problems are more likely to be isolated and be less active in their community. Mental health problems cover a wide range of illnesses, including disorders such as mild or moderate anxiety and depression, drug and alcohol use disorders, and severe disorders such as severe depression, bipolar disorders and schizophrenia. Comorbidity of mental disorders and physical illnesses, and multiple mental health problems, is common. Some mental disorders may affect individuals for only a short time, while others affect individuals their entire life. Mental health problems often result from a complex interplay of many factors, including genetic, social and economic factors, and can be provoked or worsened by behavioural and environmental factors such as alcohol and drug abuse, poverty and debt, trauma, or physical ill-health. The burden of mental health problems in Europe is very high, both in terms of morbidity and mortality. Tens of millions of people across the EU experience at least one mental health problem at any point in time, and tens of thousands die each year either directly from mental health disorders or from suicide (which in many cases are linked to mental health problems, although other factors can also play a role).The economic burden, too, is significant. This chapter estimates total costs related to mental ill-health at more than 4% of GDP – or over EUR 600 billion – across the 28 EU countries in 2015. EUR 190 billion (or 1.3% of GDP) is direct spending on health care, another EUR 170 billion (1.2% of GDP) is spending on social security programmes, while a further EUR 240 billion (1.6% of GDP) is caused by indirect costs in the labour market, driven by lower employment rates and reduced productivity due to mental illness. In response to the health and economic impact of mental illness, European countries are taking actions to both prevent and treat mental illness when it occurs. The economic, societal and individual burden of mental illness is not a foregone conclusion – many interventions exist which can lessen the impact of mental ill-health. While the latter part of this chapter focuses mainly on effective interventions to prevent mental illness and promote mental well-being, improving access to early diagnosis, care and treatment for mental health conditions when they arise remains critical. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 20

23 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Carefully chosen and well-implemented actions to promote better mental health and prevent mental ill-health can lead to significant benefits over time, for individuals and their families, for society, and for economies. Cost-effective and sometimes even cost- saving interventions can help strengthen the mental well-being and resilience of mothers and infants, school-age children, workers, and older populations. Box 1.1. Defining mental health and mental illness The widely used definition established by the WHO emphasises the positive dimension that “mental health is a state of well-being in which the individual realises his or her own abilities, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribution to his or her community” (WHO, 2001). The terms mental health and mental well-being in this chapter draw on this WHO definition of positive mental health. Mental illness is the loss of mental health due to a mental disorder. Mental disorders are defined as those reaching the clinical threshold of a diagnosis according to psychiatric classification systems including disorders such as depression, anxiety, bipolar disorder and schizophrenia. In this chapter, mental illnesses will generally comprise all those included in Chapter 5 of the International Classification of Diseases (ICD-10) on mental and behavioural disorders with the exception of dementia (which is considered, along with Alzheimer’s disease, the main form of dementia, as a neurological disorder). The broad terms “mental ill-health”, “mental illness” and “mental health problems” are used interchangeably and refer to mental disorders but also include psychological distress, i.e. symptoms or conditions that do not reach the clinical threshold of a diagnosis within the classification systems but which can account for significant suffering and hardship, and can be enduring and disabling. Mental illness affects tens of millions of Europeans every year Mental health problems affect about 84 million people across EU countries Although there are significant gaps in information about the prevalence of mental health problems across EU countries, all available evidence suggests that mental health problems affect tens of millions of Europeans every year. The data currently available from population-based surveys are often limited to a few specific mental health disorders, or specific age groups. However, the Institute for Health Metrics and Evaluation (IHME) provides estimates of the prevalence of a wide range of mental health disorders across all age groups based on a wide variety of data sources and a set of assumptions According to the latest IHME estimates, more than one in six people across EU countries 1 igure 1.1) – that is, nearly 84 million people. (17.3%) had a mental health problem in 2016 (F The most common mental disorder across EU countries is anxiety disorder, with an estimated 25 million people (or 5.4% of the population) living with anxiety disorders, followed by depressive disorders, which affect over 21 million people (or 4.5% of the population). An estimated 11 million people across EU countries (2.4%) have drug and alcohol use disorders. Severe mental illnesses such as bipolar disorders affect almost 5 million people (1.0% of the population), while schizophrenic disorders affect another estimated 1.5 million people (0.3%). By country, the estimated prevalence of mental health disorders is highest in Finland, the Netherlands, France and Ireland (with rates of 18.5% or more of the population with at HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 21

24 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? More than one in six people in EU countries have a mental health problem Figure 1.1. Alcohol and drug use disorders Depressive disorders Anxiety disorders Bipolar disorders and schizophrenia Others 25% 20% 18.8% 18.6% 18.5% 18.5% 18.5% 18.4% 18.3% 18.3% 18.3% 18.0% 17.9% 17.9% 17.7% 17.7% 17.7% 17.6% 17.5% 17.3% 17.3% 17.0% 17.0% 16.9% 16.9% 16.7% 15.7% 15.5% 15.4% 15.2% 15.1% 14.9% 14.8% 14.3% 15% 10% 5% 0% IHME, 2018 (these estimates refer to 2016). Source: http://dx.doi.org/10.1787/888933833920 1 2 least one disorder), and lowest in Romania, Bulgaria and Poland (with rates of less than 15% of the population). Some of these cross-country differences may be due to the fact that people living in countries with greater awareness and less stigma associated with mental illness, as well as easier access to mental health services, may be diagnosed more easily or may be more likely to self-report mental ill-health. In many countries, there is still strong stigma associated with various mental health problems, and in some countries this stigma sits alongside a still-widespread belief that it is better to simply avoid talking about mental illness (Munizza et al., 2013). Several mental illnesses are more common amongst women, including anxiety disorders, depressive disorders and bipolar disorders. Some of these gender gaps may be due to a greater propensity of women to report these problems. However, one exception is drug and alcohol use disorders, which are more than two times more likely to occur in men than women on average across EU countries (IHME, 2018). Data from the 2014 European Health Interview Survey confirm a substantial gender gap in self-reported chronic depression, with more than one in twelve women (8.8%) indicating they experience chronic depression, compared with one in nineteen men (5.3%). The prevalence of chronic depression increases steadily with age among both women and men, and is particularly high in middle age (Figure 1.2). At age 55-64, more than 11.4% of women and 7.1% of men reported being chronically depressed across the EU as a whole in 2014. These rates decrease between the age 65 and 74, and then increase again in older ages. This increase in older ages may be partly explained by the fact that depression is often associated with poor physical health, frailty, perceived financial strain and lower social support (Grundy, van den Broek and Keenan, 2017). HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 22

25 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Chronic depression is more often reported by women Figure 1.2. and increases with age in EU countries Men Women % with chronic depression % with chronic depression 14 14 12.4 11.4 12 12 9.9 9.8 10 10 8 8 7.1 7.1 6.7 6.5 6.1 5.4 5.3 6 6 4.7 4.5 4 4 2.7 2 2 0 0 75+ 75+ 65-74 15 - 24 35 - 44 65-74 55 - 64 45 - 54 55 - 64 25 - 34 15 - 24 25 - 34 35 - 44 45 - 54 Source: Eurostat Database (based on EHIS 2014). http://dx.doi.org/10.1787/888933833939 1 2 By level of education, people with at most lower secondary educational attainment are almost two-times more likely to report chronic depression compared to those with higher educational level. This is also the case for people in low-income groups. On average across EU countries, women and men living in the lowest income group are more than two times oup (Figure 1.3). more likely to report chronic depression than those in the highest income gr People who are employed generally report lower levels of depression than those who are not, and people with a mental disorder are more likely to be unemployed (OECD, 2015). People with depression or other mental health problems often see improvement in their condition after finding work, as their labour-force status increases their self-esteem and sense of worth in society, and losing a job generally contributes to worsened mental health (OECD, 2018). A considerable number of children experience mental health problems which, unless they receive appropriate care and support, may have a lasting effect throughout their lives. Evidence suggests that many mental disorders begin at adolescence or even younger; most studies find that roughly half of all lifetime mental disorders start by the mid-teens (Kessler et al., 2007). A 2010 study found that in five of the six EU countries covered (Bulgaria, Germany, Lithuania, the Netherlands and Romania), 10% to 15% of children aged 6-11 years old experience at least one mental health or behavioural disorder (i.e. conduct disorder, emotional disorder, hyperactivity or inattention disorder). Italy is the only country where prevalence was less than 10%, but about 8% of children still had a mental or behaviour disorder (Kovess-Masfety et al., 2016). HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 23

26 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Women and men in the lowest income group are more than two times more likely Figure 1.3. to report chronic depression than those in the highest income group across the EU Men Women Hi High income Low income h income g Low income a R oman i gar i u B l a reece G C yprus o P d an l zec R h C ep. ep. ova k R Sl a uan h i Li t y l ta I i ston E a a i roat C F rance M a l ta i e l g um B EU28 i ustr A a H ungary D enmar k we S d en b L uxem ourg n i S pa a i L atv d an l I re ermany G a Sl oven i om d ng Ki d te i n U d an et h s er l N l P ortuga an Fi n d l N orway d I ce l an % % 15 20 25 10 0 5 20 25 5 0 15 10 Note: High income refers to people in the top income quintile (20% of the population with the highest income), whereas low income refers to people in the bottom income quintile (20% of the population with the lowest income). Countries are listed in order of rate of reported chronic depression by women (from lowest to highest). Data for Switzerland is not available. Source: Eurostat Database (based on EHIS 2014). http://dx.doi.org/10.1787/888933833958 1 2 Mortality related to mental health problems and suicides is substantial Over 84 000 people died of mental health problems and suicides across EU countries in 2015, and this is an under-estimation as many people with mental health problems also die prematurely because of higher rates of physical health problems and chronic diseases that are not properly treated. “Excess mortality” for mental disorders – the gap between the mortality rate of the general population and the mortality rate for people with a mental disorder – is huge. For example, excess mortality amongst women who have been diagnosed with schizophrenia is above 6 in Finland, Norway and Sweden (OECD, 2018). Persons with severe mental illness die 10-20 years earlier than the general population (Liu et al., 2017; OECD, 2014; Coldefy and Gandré, 2018). Of the 84 000 deaths directly related to mental health problems and suicides, most of these deaths were among men, mainly because of higher suicide rates among men (Figure 1.4). Some 43 000 men in EU countries died from suicide in 2015, compared with 13 000 women. However, the gender gap in suicide attempts is much smaller or even reversed in some countries, because women often use less fatal methods. For example, in France, while the completed suicide rate is more than three times greater among men than HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 24

27 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? The number of deaths from mental health problems Figure 1.4. and suicides generally increases with age Men Women Suicide Mental and behavioural disorders Number of deaths Number of deaths 10 000 10 000 9 000 9 000 8 000 8 000 7 000 7 000 6 000 6 000 5 000 5 000 4 000 4 000 3 000 3 000 2 000 2 000 1 000 1 000 0 0 35-44 75+ 65-74 55-64 45-54 0-14 25-34 15-24 55-64 25-34 15-24 0-14 45-54 65-74 75+ 35-44 Note: Mental and behavioural disorders cover all the diseases in the related ICD-10 chapter with the exception of dementia. Eurostat Database (the data refer to 2015). Source: http://dx.doi.org/10.1787/888933833977 1 2 women, hospital discharge rate for suicide attempts was 52% greater among women in 2015 (Observatoire national du suicide, 2018). Many different factors may explain why some people are led to attempt or complete suicide, including major life events (such as the death of a loved one, a divorce or employment loss), social isolation, or socioeconomic or cultural context. However, a high proportion of people who have survived a suicide attempt or died from suicide have experienced a mental health disorder (Hoven, Mandell and Bertolote, 2010; Cavanagh et al., 2003; WHO, 2014). A cross-national analysis based on the WHO World Mental Health Surveys found that a wide range of mental disorders increased the odds of experiencing suicidal thoughts, and a smaller number of disorders increased the odds of acting on such thoughts (Nock et al., 2009). The number of suicides increases steadily with age among both men and women, reaching a peak among 45-64 years-olds (Figure 1.4). Between ages 65 and 74 the number of suicides decreases at least slightly. By country, the suicide rate among the population of all ages is highest, by far, in Lithuania, with (age-standardised) rates of 30 deaths per 100 000 population in 2015. Slovenia, Latvia and Hungary also have high rates at around 20 deaths per 100 000 population, which is almost two times greater than the EU average (11 per 100 000 population).The lowest rates are reported in Southern European countries (Greece, Cyprus, Italy, Malta and Spain) (F igure 1.5). Some caution is required in interpreting suicide rates as these may reflect, at least in part, differences in recording practices. On average across all countries, the suicide rate among men was 3.7 times greater than among women. This gender gap was largest in the four countries with the highest rate, but also in Estonia, Poland and Romania. Despite the relatively low absolute number of suicides among younger age groups, suicide is nonetheless one of the leading causes of death among adolescents and young adults. Some 3 400 young people age 15-24 died from suicide in EU countries in 2015, making this the main cause of death in this age group after road traffic injuries. Young people are more likely to attempt suicide if they have a family history of alcohol and drug HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 25

28 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Men are more likely to die from suicide in all EU countries Figure 1.5. Women Total Men Age-standardised rate per 100 000 population 60 50 40 30 20 10 0 1. Three-year average (2013-15). Source: Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933833996 abuse disorders, have access to firearms, and experience difficult life events at school or at home (McLoughlin, Gould and Malone, 2015). However, it is heartening to note that suicide rates among teenagers have decreased by 20% on average across EU countries between 2000 and 2015. There has been a notable decrease in Finland, reflecting the success of suicide prevention campaigns targeting this age group (see Box 1.5). The costs of mental health problems exceed 4% of GDP The total costs of mental health problems on EU economies are huge, highlighting the need for greater efforts to prevent mental ill-health and to provide timely and effective treatments when it occurs. Besides the costs on health care systems, mental health problems also result in substantial costs in terms of social security benefits as well as negative labour market impacts in terms of reduced employment and productivity. This section provides estimates of the direct and indirect costs related to mental illnesses across EU countries, using different data sources and based on a set of explicit assumptions where x 1.2). necessary (see Bo In 2015, the overall costs related to mental ill-health are estimated to have exceeded 4% of GDP across the 28 EU countries. This equates to more than EUR 600 billion. This total breaks down approximately into the equivalent of 1.3% of GDP (or EUR 190 billion) in direct spending on health systems, 1.2% of GDP (or EUR 170 billion) on social security programmes, and a further 1.6% of GDP (or EUR 240 billion) in indirect costs related to labour market impacts (lower employment and lower productivity). Despite these costs being considerable, they are still a significant under-estimate, as several additional costs have not been taken into account. These include, in particular, social spending related to mental health problems, such as higher social assistance benefits and higher work-injury benefits, and the higher cost of treating a physical illness if the patient also has a mental illness. In addition, some of the indirect impacts of mental health problems on labour market participation such as reduced employment rates or working hours for informal caregivers taking care of people with mental health problems or the impact on co-workers, have not been taken into account. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 26

29 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? By country, the estimated costs related to mental health problems range from 2% to 2.5% of GDP in Romania, Bulgaria and the Czech Republic, to over 5% of GDP in Denmark, Finland, the Netherlands and Belgium (Figure 1.6). These variations are mainly driven by the share of people reporting mental health problems (which may be under-estimated in countries where there is a strong stigma associated with mental health problems) as well as differences in the social security benefits provided to people with mental health problems (in terms of paid sick leave benefits, disability benefits and unemployment insurance benefits), and different levels of spending on mental health care services. Figure 1.6. Estimated direct and indirect costs related to mental health problems across EU countries Asa%ofGDP,2015 4.9 5.3 5.0 4.8 2.8 3.2 5.4 2.6 3.2 5.1 4.1 3.0 4.8 5.1 3.1 2.5 Legend 2.6 2.10% - 3.39% 4.3 3.1 3.5 3.7 3.40% - 4.69% 4.1 2.1 4.0 4.70% - 5.40% 3.3 2.4 4.2 3.7 3.0 (Malta) 3.3 3.2 OECD estimates (see Box 1.2 and Table 1.1 for further information). Source: http://dx.doi.org/10.1787/888933834015 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 27

30 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Table 1.1. Estimates of total costs (direct and indirect) of mental health problems in EU countries, in million EUR and as a share of GDP, 2015 Total costs Direct costs Indirect costs On the labour market On health systems On social benefits %ofGDP in million EUR %ofGDP in million EUR %ofGDP in million EUR in million EUR %ofGDP 4.10 194 139 1.31 EU28 1.15 242 995 1.64 607 074 169 939 14 930 4.33 4 686 1.36 3 902 1.13 6 342 1.84 Austria 20 740 Belgium 5 447 1.33 5 845 1.42 9 448 2.30 5.05 Bulgaria 2.36 448 0.99 299 0.66 320 0.71 1 067 1 785 537 525 1.18 Croatia 1.21 724 1.63 4.01 Cyprus 3.21 203 1.14 144 0.81 223 1.25 569 Czech Republic 4 132 2.45 1 727 1.02 1 046 0.62 1 360 0.81 Denmark 14 627 3 431 1.26 5 563 2.05 5 633 2.07 5.38 572 210 1.03 167 0.82 196 0.96 Estonia 2.81 11 140 5.32 2 576 1.23 Finland 1.85 4 681 2.23 3 884 France 81 345 3.71 29 337 1.34 26 437 1.20 25 570 1.17 Germany 146 536 4.81 43 421 1.43 40 939 1.35 62 177 2.04 Greece 3.01 2 241 1.27 1 440 0.82 1 630 0.92 5 311 Hungary 3 454 3.12 1 417 1.28 703 0.64 1 333 1.20 Ireland 8 299 3.17 2 232 0.85 1 891 0.72 4 176 1.59 0.96 1.12 Italy 54 487 3.30 20 221 1.22 15 787 18 478 Latvia 789 270 1.11 169 0.70 350 1.44 3.24 990 2.64 0.99 266 0.71 352 0.94 Lithuania 372 3.14 1.35 0.79 701 1 634 520 1.00 Luxembourg 413 314 132 1.38 40 0.42 142 1.50 Malta 3.29 34 969 5.12 8 534 1.25 Netherlands 1.62 15 367 2.25 11 069 Poland 3.01 5 113 1.19 3 235 0.75 4 604 1.07 12 952 6 580 1 652 2 048 1.14 Portugal 0.92 2 880 1.60 3.66 Romania 2.12 1 510 0.94 737 0.46 1 153 0.72 3 400 Slovak Republic 2 061 2.61 655 0.83 599 0.76 807 1.02 Slovenia 1 602 507 1.31 308 0.79 786 2.02 4.13 45 058 14 415 1.33 12 318 1.14 18 325 1.70 Spain 4.17 21 677 4.83 5 696 1.27 Sweden 1.68 8 423 1.88 7 558 United Kingdom 106 024 4.07 36 353 1.40 22 704 0.87 46 967 1.80 Iceland 753 4.93 201 1.31 265 1.73 288 1.88 Norway 4.97 4 965 1.43 6 384 1.83 5 950 1.71 17 299 Switzerland 21 679 3.54 5 769 0.94 7 023 1.15 8 888 1.45 Source: OECD estimates based on Eurostat Database and other data sources (see Box 1.2 on sources and methodology on direct and indirect costs). 1 2 http://dx.doi.org/10.1787/888933834034 Box 1.2. Methodology and data sources used to estimate the costs of mental health problems Table 1.2 below summarises the different categories of direct and indirect costs that have been considered in the analysis in this chapter, along with the main data sources used. The direct costs include both those borne by health care systems to provide treatments to mental health problems and additional social security spending, including paid sick leave benefits, disability benefits and unemployment insurance benefits. The indirect costs relate to the labour market impact of mental health problems, and include both lower employment rates for people with mental health problems and lower productivity due to higher absenteeism and lower productivity when at work (“presenteeism”). HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 28

31 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Box 1.2. Methodology and data sources used to estimate the costs of mental health problems (cont.) Summary of direct and indirect costs related to mental health Table 1.2. problems and main data sources Sources Specific cost categories Broad categories Impact on health spending Higher direct health care costs (physician visits, pharmaceutical Cost of disorders of the brain in Europe 2010 costs and hospitalisations, etc.) Eurostat Health Expenditures by Diseases and Conditions 2016 Impact on social spending Higher paid sick leave benefits Eurostat Database and national administrative data (for some countries) Eurostat Database and national administrative Higher disability benefits data (for some countries) Eurostat Database and national survey data for Higher unemployment insurance benefits some countries Impact on labour market Lost income due to mortality from mental illnesses among Eurostat Database (Causes of mortality ) (employment and productivity) working-age population Lost income due to lower employment rate among working-age Eurostat Database (European Health Interview Survey 2014) population with mental health problems European Working Conditions Survey (2015) Lost income due to greater absenteeism (fewer hours worked and Eurostat Database and more sick leaves) among people with mental health problems Lost income due to lower productivity for people with mental European Working Conditions Survey (2015) and Eurostat Database health problems at work (presenteeism) Estimates of direct health care costs are based on a selection of mental health conditions contained in a previous study on the cost of disorders of the brain in Europe (Gustavsson et al., 2011). The original cost estimates have been extrapolated to 2015 using recent health spending data and updated macroeconomic data. Overall estimates have also been corroborated with country-specific health expenditure by disease studies such as the Eurostat Health Expenditures by Diseases and Conditions study in 2016. The assumption has been made that the share of mental health spending remained constant between 2010 and 2015. The main data sources for the estimates on social security benefits are the Eurostat Database, the European Working Conditions Survey, and national data sources. The following assumptions have been made to fill data gaps on the share of social security spending related to mental health problems for countries that did not have the required data readily available: 1) 20% of paid sick leave benefits are related to mental health problems, based on the available evidence from Sweden (OECD, 2012); 2) 37% of disability benefits are related to mental health problems, based on the available evidence from six countries (Austria, Belgium, Denmark, Netherlands, Sweden, United Kingdom) (OECD, 2015) and 3) 15% of unemployment insurance benefits are related to mental health problems, based on the evidence from the same group of six countries that about 30% of people on average who are receiving unemployment insurance benefits also report some mental health problems, but assuming that mental health problems are the leading cause for unemployment for half of these people only. The labour market impact of mental health problems draws also on the Eurostat Database and the European Working Conditions Survey. The approach used to measure the negative employment effect of mental health problems is to assume that people with mental health problems would have had the same employment rate as the rest of the population, and earn the same salary, using the median wage in the economy. The productivity effect is measured by looking at both absenteeism and “presenteeism”. The latter is based on a study that has found that both blue-collar and white-collar workers experiencing mental ill-health are about 6% less productive than those without such problems (Hilton et al., 2008). The assumption is made that this lower productivity at work is reflected in lower wages. The costs throughout the analysis are expressed in euros without any adjustment for variations in the cost of living (no adjustment for purchasing power parity). HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 29

32 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Direct costs of mental health problems on health systems and social security benefits A sizeable share of health spending goes towards mental health problems Spending on the provision of mental health services is estimated to have accounted for about 13% of health spending across EU countries in 2015. This is less than spending on circulatory diseases – the number one cause of mortality in the EU – but similar to spending on cancer care in many countries. This equals 1.3% of GDP or around EUR 194 billion of direct health care spending on a broad range of mental health conditions across the EU. This covers spending on the health services and goods related to the prevention, diagnosis and treatment of mental health disorders (including physician visits, hospitalisations and pharmaceuticals). This spending reaches an estimated 1.4% of GDP in Germany and the United Kingdom. At the lower end, in addition to Luxembourg at 0.8% and Ireland at 0.9%, Lithuania, Bulgaria, Romania and the Slovak Republic are all estimated to have spent less than 1% of GDP on direct health care services for mental health. Mental health problems result in much higher sickness benefits, disability benefits and unemployment insurance benefits The direct costs of mental ill-health extend well beyond the health system; mental illness leads to substantial additional spending in many social security programmes, including paid sick leave benefits, disability benefits and unemployment insurance benefits. Expenditure on disability benefits accounts for the bulk of mental health-related social spending. It is estimated that mental health problems accounted for EUR 112 billion in disability benefits across the EU as a whole in 2015 (or 0.76% of GDP). Paid sick leave benefits related to mental health problems accounted for another EUR 28 billion (or 0.19% of GDP) in 2015, whereas unemployment insurance benefits were estimated to add another EUR 29 billion (or 0.20% of GDP). As already noted, these estimated costs of mental health problems on social spending are an under-estimation as they do not include the cost of other social programmes, such as social assistance benefits or lone-parent benefits. Indirect costs of mental health problems on employment and productivity Beyond the direct costs to health systems and social security benefits, mental ill-health also contributes to substantial indirect costs, primarily related to reduced labour market participation and productivity.These indirect costs include not only lower employment rates for people with mental health problems, but also reduced productivity due to higher absenteeism and lower productivity at work (often referred as “presenteeism”). These costs add up to over EUR 240 billion or 1.6% of GDP across EU countries in 2015. Lost income and employment due to mortality from mental health problems and suicide is estimated at EUR 22 billion per year across EU countries Over 50 000 premature deaths among the working-age population (people aged 25-64) were due to mental health problems and suicide across EU countries in 2015. Assuming that all those people who died prematurely would have been employed until age 65 at the same employment rate as the rest of the population, the associated potential loss for the economy is estimated to be about 640 300 potentially productive life years across EU countries. Assuming that these people would have earned the median income in each HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 30

33 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? country, this amounts to EUR 22 billion in potential income loss each year, or 0.15% of GDP across the EU as a whole. This loss in income as a share of GDP was particularly large in Slovenia, Belgium, Latvia and Lithuania, mainly because of higher suicide rates. Lost income due to lower employment rate of people with depression is estimated at EUR 176 billion per year across EU countries Living with mental health problems has an impact on people’s daily lives, including their ability to work. Mental health problems often impede an individual’s ability to participate in the labour market which can lead to a “vicious” circle whereby the longer people are out of work, the more damaging the consequences are for their mental health (OECD, 2014). The analysis here only focuses on the labour market impact of depression, as it is the only mental health problem considered in the last wave of the European Health Interview Survey in 2014. Figure 1.7 shows that people reporting chronic depression have much lower employment rates than the rest of the population. Only about half of the population aged 25-64 reporting chronic depression were in employment, compared with over three-quarter (77%) among those who do not report chronic depression on average across EU countries. This employment gap is particularly large in Cyprus, Croatia, Malta, Romania and Bulgaria, although this may partly be due to small sample sizes in EHIS. The cost of this lower employment rate related to chronic depression is estimated at about EUR 176 billion in 2015, representing an amount equivalent to 1.2% of GDP across EU countries as a whole. Figure 1.7. People reporting chronic depression are much less likely to work in all EU countries % of working age pop. aged 25-64 With depression Without depression % of working age population aged 25-64 100 90 80 70 60 50 40 30 20 10 0 Note: Weighted EU28 average. People with depression are identified through the question “During the past 12 months, have you had any of the following diseases?” with depression being one of these diseases. Due to missing data, the assumption has been made that the situation in Ireland is the same as the EU average. Eurostat Database, based on the European Health Interview Survey (2014). Source: 1 2 http://dx.doi.org/10.1787/888933834053 Higher absenteeism and lower productivity at work amongst people with mental health problems is estimated to cost about EUR 42 billion in EU countries Even when people with mental health problems are working, the cost of mental health problems for employees and employers in terms of greater absenteeism and lower HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 31

34 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? productivity at work is high. Reduced working hours and more days of absence from work are estimated to cost about EUR 19 billion or an amount equivalent to 0.13% of GDP across EU countries in 2015. Even when at work, people with mental health problems do not always function to their full or usual abilities and may be less productive – what is often referred as “presenteeism”. Based on the finding that workers experiencing mental ill-health are about 6% less productive than those without such problems (Hilton et al., 2008), and assuming that lower productivity is reflected in lower wages, the cost of this loss of productivity is estimated at almost EUR 23 billion in 2015. The high direct and indirect costs of mental illness should not be seen as a foregone conclusion. Greater and more effective investment in mental health promotion and treatment could help substantially reduce many of these costs and help more people realise their full potential. Actions to promote mental health and prevent mental illness in Europe The substantial costs of mental health problems make a clear case for increasing efforts to promote good mental health and prevent mental illness, as well as to identify the signs and symptoms of mental illness early, and improve the management and treatment of mental health problems when they occur. More and more European countries are ensuring they have comprehensive policies in place. Several countries (e.g. Belgium, the Czech Republic, Finland, France, Hungary, Ireland, Italy, the Netherlands, Slovenia, Spain, Portugal and the United Kingdom) have a specific plan or policy document addressing mental health promotion and prevention. Mental health promotion or prevention policies are designed to promote mental health in schools and workplaces, to prevent suicide, to improve the mental well-being of older people, or detect mental distress early on. As awareness of mental illness improves, and stigma around mental illness falls, more people may also seek help when they experience mental illness. Several international strategies have also supported a greater focus on addressing mental health issues. The 2015 Recommendation of the OECD Council on Integrated Mental Health, Skills and Work Policy (2015) (see Box 1.3) aims to foster mental well-being and improve awareness of mental health conditions by encouraging activities that promote good mental health as well as help-seeking behaviour when mental illness occurs. The European Framework for Action on Mental Health and Wellbeing (European Commission, 2016), too, focused on the effective implementation of policies and interventions contributing to promotion of mental health and the prevention and treatment of mental disorders, including through integration of mental health in all policies and multi-sectoral cooperation. The importance of including mental health promotion is echoed in the activities of the EU-Compass for Action on Mental Health and Wellbeing (see Box 1.4). The WHO Comprehensive Mental Health Action Plan 2013-2020 (WHO, 2013) emphasises integrated and coordinated prevention, promotion, care and support including via the implementation of a multi-sectoral strategy that combines universal and targeted interventions for promoting mental health and preventing mental disorders. There are more than 100 prevention and promotion actions in place across the 28 European countries and 3 EFTA countries (with counting capped at one per life course category in each country). Actions were identified across different points across the life HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 32

35 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Box 1.3. OECD Recommendation of the Council on Integrated Mental Health, Skills and Work Policy Recognising that mental ill-health demands interventions that are cross-sectoral in scope and complementary in nature, in 2015 the OECD Council published the OECD Recommendation of the Council on Integrated Mental Health, Skills and Work Policy (OECD, 2015). This recommendation is a sign that governments in OECD countries understand that good policies can make a significant difference when it comes to preventing mental illness at all ages, including in youth and adolescence, in supporting those experiencing mental illness to stay in the workplace and supporting those who have left employment to return to the labour market. The OECD Recommendation gives a series of guidelines to address the impact of mental ill-health on employment, education, health and social outcomes.These guidelines, which all OECD signatories are expected to follow, encourage countries to seek to “promote mental well-being, prevent mental health conditions, and provide appropriate and timely services which recognise the benefits of meaningful work for people living with mental health conditions”. The EU-Compass for Action on Mental Health and Wellbeing Box 1.4. The EU-Compass for Action on Mental Health and Wellbeing drove the collection, exchange and analysis of information on policy and stakeholder activities in mental health in European countries between 2015 and 2018. The Compass was a means of communicating information on the European Framework for Action on Mental Health and Wellbeing, as well as monitoring the mental health and well-being policies and activities of EU countries and non-governmental stakeholders. Main activities under the Compass included the identification and dissemination of good practices in mental health, collection of information on activities in mental health, and holding mental health workshops in each EU country and in Iceland and Norway. The EU Compass generated a series of published good practice, annual reports, and consensus paper, especially around seven priority areas: ● Preventing depression and promoting resilience (priority for 2016) Better access to mental health services (priority for 2016) ● ● Mental health at work (priority for 2017) ● Mental health in schools (priority for 2017) ● Preventing suicide (priority for 2017) Providing community-based mental health services (priority for 2018) ● Developing integrated governance approaches (priority for 2018) ● Alongside governments, the Compass also engaged with businesses, educational institutions and civil society organisations on their role in implementing positive mental health initiatives. Engaged stakeholders, and policies collected from these stakeholders, are also available on the EU Compass web platform. course including: prenatal, perinatal and infancy; children aged 2-10 years and their parents; children and young people aged 11-25 years; workplace mental health; unemployed populations; and older people. Actions were identified from an OECD survey HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 33

36 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? of mental health promotion and prevention programmes, the WHO Mental Health Atlas 2017, and actions reported to the EU Compass 2016-2018, and supplemented with a literature review. Figure 1.8 identifies countries reporting at least one action in a particular life course area. At least one prevention or promotion action was found in every European country. Targeted prevention or promotion programmes were found in all but four countries (Bulgaria, the Czech Republic, Malta and Romania). Generalised prevention and promotion programmes were also reported, for instance the Czech Republic was unable to divide programmes into target groups as programs supported by the Ministry of Health of the Czech Republic are mostly designed for all persons with mental illness. Countries reporting at least one promotion or prevention action Figure 1.8. for mental health in areas across the life course Number of countries reporting at least one promotion or prevention action, out of the 31 EU and EFTA countries Number of countries reporting at least one action 25 20 15 10 5 0 Children aged 2-10 Pre-natal period to age 2 Young people aged 11-25 Workplace mental health Mental health of the Older people unemployed McDaid, Hewlett and Park (2017); EU Compass for Action on Mental Health and Wellbeing (2017); WHO (2018); EU Compass for Source: Action on Mental Health and Wellbeing, 2018 (2018). http://dx.doi.org/10.1787/888933834072 1 2 It is clear from Figure 1.8 that the d istribution of actions to promote mental well-being and prevent mental ill-health is uneven throughout the life course. 22 of 31 countries had actions in place targeting young people aged 11-25 and the actions targeting the workplace, while 18 countries had actions targeting the prenatal to 2 years period, with the same number for children aged 2-10 years. However, actions to target the mental health of unemployed persons were reported or identified in the literature for only 9 countries, and actions targeting the mental health of older populations were reported or found in only 12 countries. Preventing deaths by suicide Though suicide remains a major cause of death, and still contributes significantly to mortality from mental illness (as discussed earlier in the chapter), longstanding national commitments to reducing suicide in European countries have helped to reduce the rate of suicide in most countries. On average, the number of deaths by suicide per 100 000 population fell from 12.5 in 2005 to 10.9 in 2015 (Eurostat, 2017). In some countries the falls were even more significant, albeit often from a higher starting rate. Between 2005 and 2015, deaths by suicide fell by more than 20% in almost half of all EU28 countries. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 34

37 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? A range of measures are recognised as effective in reducing suicide, including restricting access to lethal means, raising awareness of suicide and suicide risk, improving access to mental health treatment, signposting to sources of help and protective measures in suicide “hotspots”, and tailored efforts to reduce suicide following hospitalisation, for example psychosocial assessment and good follow-up care (Hawton et al., 2016; Zalsman et al., 2016). Such approaches have helped some countries achieve significant falls in suicide rates (for instance Finland, see Box 1.5), even as all countries still continue to seek to prevent suicide more effectively. Box 1.5. A renewed strategy to prevent suicide in Finland In Finland, the rate of suicide has fallen by over 50% over the past 30 years. A significant driver of the reduction in suicide has been the fall amongst young men aged 20-29 (Figure 1.9). Nonetheless, death by suicide amongst young Finnish males remains high in comparison with other Nordic countries (Denmark, Norway, Sweden). Mental illness and alcohol dependence or abuse are significant causal factors (Titelman et al., 2013; Wahlbeck et al., 2011), but socioeconomic conditions have also had an impact. Suicide amongst young Finns (15-29), 1980-2016 Figure 1.9. 15 - 19 20 - 29 Men Women Individuals Individuals 200 200 150 150 100 100 50 50 0 0 Source: Statistics Finland. http://dx.doi.org/10.1787/888933834091 1 2 Suicide prevention campaigns in Finland began in the 1980s, and led to a series of national suicide prevention programmes that ran during the 1990s. Finland’s strategy identified depression, access to mental health care, substance and alcohol abuse, and access to lethal means as central features.The strategy also led to the establishment of crisis phone lines for persons experiencing suicidal thoughts, and guidance to the media, for instance not reporting suicide methods (Patana, 2014; Korkeila, 2013). Recognition that suicide was particularly high amongst young men led to the development of the “Time Out! Back on the track” ( Aikalisä! ) initiative in 2004, which promoted social inclusion amongst vulnerable men. Two-thirds of Elämä raitelleen participants reported that the participation in the programme was worthwhile, while about 60% considered it had improved their life situation (Appelqvist-Schmidlechner et al., 2012). At the end of 2017, the Finnish Parliament allocated EUR 300 000 in 2018 to develop a new national strategy to prevent suicide, which will be included in Finland’s new broader National Mental Health Strategy. This work will establish a network for coordinating suicide prevention, and improve the planning, implementation, monitoring and evaluation of suicide prevention measures (Finnish Government, 2017). HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 35

38 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Some countries have also developed dedicated national suicide strategies, or included suicide prevention in their broader mental health strategy. A comparative study also found that nationwide suicide prevention programmes had a positive effect in helping to reduce suicide, especially those focused on reducing suicide amongst elderly and young populations (Matsubayashi and Ueda, 2011). For example, Austria began “Suicide Prevention Austria” (Suizidprävention Austria [SUPRA]) in January 2017, focused on national and regional coordination of suicide prevention strategies, developing media support for suicide prevention, research, and integration of suicide prevention into other health promotion activities (EU Compass Consortium, 2017). Early life interventions to promote mental well-being Efforts to ensure good mental health in the first few years of life are cost effective in terms of mental, physical, and social outcomes. Effective actions can start even before a child is born: poor maternal mental health – conditions such as anxiety, depression, post- traumatic stress and psychosis affecting some 10-20% of women in the perinatal period (Gavin et al., 2005) – have been associated with poorer physical and cognitive development (Ibanez et al., 2015), higher risk of pre-term birth, and lower birth weight (Jarde et al., 2016). Many countries have programmes that focus on maternal health, infant health, promoting mental well-being in pre-schools, or parenting support. In England, clinical guidelines by NICE suggest that primary care providers discuss mental health and well-being with women upon first contact during the early postnatal period (National Institute for Health and Care Excellence, 2018). In 2017 the Baby-Mother-Father Perinatal Mental Disorders Service in Hungary developed a new official guideline in intersectoral cooperation, providing support for treatment of perinatal and postnatal depression, which has started as a pilot programme in one hospital (EU Compass for Action on Mental Health and Wellbeing, 2018). Programmes which promote parenting skills and seek to improve parent-child relationships, often targeting vulnerable or at-risk children, can have a positive impact on the mental health of parents and children. In Germany, the “Early Help” initiative gives support to parents of children aged 0 to 3, delivered by family midwives and other professionals, and is available to all families with more intensive services available for cases requiring more support (McDaid, Hewlett and Park, 2017). Promotion of good mental health in schools Schools are an ideal setting for interventions to promote mental well-being as almost all children and young people in Europe spend a good part of their day in school settings. School-based interventions can benefit mental health, develop mental health literacy, as well as improve social and educational outcomes; long-term benefits include improved academic performance, better resilience, and better cognitive skills (Weare and Nind, 2011; Durlak et al., 2011). Investing in good mental health for school-aged children can reduce the risk of children dropping out of school or having a difficult school-to-work transition (OECD, 2015). School-based programmes often take a universal approach, covering either the full school population or a specific age group (e.g. primary school children or secondary school children). A few countries have introduced programmes that target vulnerable or at-risk children or young people – for instance Finland, Norway or the United Kingdom (McDaid, Hewlett and Park, 2017). Interventions delivered in schools can include actions targeting HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 36

39 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? teaching skills, promotion of positive mental health or well-being, programmes to improve mental health “literacy” and understanding of mental disorders, reduce stigma, or actions to prevent bullying and cyberbullying (see Box 1.6). General mental health promotion programmes are common, for example in Slovenia, the Slovenian Network of Health Promoting Schools which covers 324 schools (around 55%), adopted the theme of mental health promotion in 2015-16, developing a manual for teachers to promote mental health. Zippy’s Friends is a universal school-based programme adopted in 27 countries, including Denmark, Ireland and Lithuania, which helps young children to develop coping and social skills. An evaluation in Norway found that the programme had helped improve the classroom atmosphere, reduce bullying, and improve academic scores (Holen et al., 2013; Clarke, Bunting and Barry, 2014). Box 1.6. Understanding and preventing cyberbullying With the increasing ubiquity of the internet, social media and online platforms, the way people, and particularly young people, interact has dramatically changed. While technological developments offer children and young people new opportunities for personal development and growth, they also present challenges to health and well-being. Concern has been rising in particular about cyberbullying. Cyberbullying can include sending offensive messages or comments online, spreading rumours, excluding victims from online groups and other forms of harassment (OECD, 2017). Like bullying, exposure to cyberbullying has been related to a wide range of negative outcomes, including stress and suicidal thoughts (Kowalski et al., 2014), depression and anxiety (Fahy et al., 2016). The Health Behaviour in School-Aged Children (HBSC) survey of 42 countries asked children about their experiences of bullying on the internet, and found that on average 11% of children aged 11, 13 and 15 reported having been cyberbullied at least once by message. Just over 3% of children reported having been cyberbullied by message at least 2-3 times a month. In all countries the rate of bullying in school was found to be significantly higher than the rate of cyberbullying. Figure 1.10. Bullying and cyberbullying experienced by children aged 11, 13 and 15, 2013/1 4 Cyberbullying² Bullying in School¹ % of children aged 11, 13 and 15 60 53 49 50 39 38 36 40 34 34 34 33 32 31 30 30 29 28 28 28 26 25 30 23 22 22 21 20 18 18 17 17 16 15 20 13 22 19 17 17 16 16 10 14 14 13 12 11 11 11 11 11 10 10 8 9 9 9 7 7 6 6 8 84 8 6 7 0 1. Proportions who reported being bullied at least twice at school in the past couple of months. 2. Proportions who had experienced cyberbullying by message (instant messages, wall-postings, emails and text messages) at least once. Health Behaviour in School-aged Children (HBSC) Survey, 2013/14. Source: http://dx.doi.org/10.1787/888933834110 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 37

40 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? (cont.) Understanding and preventing cyberbullying Box 1.6. Some interventions have been found as effective ways to counter cyberbullying, including school-based anti-bullying programmes, programmes including parent meetings, parent and child education, and teaching empathy and coping skills (Hutson, Kelly and Militello, 2018; Farrington and Ttofi, 2009) although the long-term effectiveness of these programmes is not clearly evidenced (Cantone et al., 2015). Protecting and improving the mental health of the working-age population Actions around mental health in the workplace – which 20 countries report having in place – include efforts to improve mental well-being, actions to support workers experiencing mental ill-health stay in work, and actions to facilitate return-to-work after a period of sickness absence. Such interventions can contribute to reducing some of the high economic costs related to mental illness noted earlier in the chapter and contribute to maximising productivity, opportunities and fulfilment for employees. The most economically effective interventions were found to be those targeting individuals rather than organisations (McDaid and Park, 2014; Hamberg-van Reenen, Proper and van den Berg, 2012). Many European countries are using health and safety legislation and labour laws to safeguard and promote mental well-being at work. Austria, Belgium, Finland, France, Norway, and the Netherlands are using labour legislation to tackle psychosocial workplace risks. Finland and Lithuania require employers to assess and respond to mental stress and strain at work (McDaid, Hewlett and Park, 2017). Workplace programmes can focus on the individual, or on an organisation-wide approach, for instance promoting mental health awareness amongst managers, changes to the physical working environment, and improving social relations at work. In Belgium prevention advisers give guidance to workplaces on psychological well-being, and support the preparation of risk assessment plans to minimise stress and violence at work (Samele, Frew Stuart and Urquia Norman, 2013). In the Netherlands the “[email protected]: Stress Prevention at Work” aims to identify and deal with stress in the workplace through a learning network, a digital Occupational Stress platform, and roadmaps tailored to each individual company (EU Compass for Action on Mental Health and Wellbeing, 2017). Few initiatives, though, were found to focus on improving the mental health of the unemployed, with actions reported or identified in the literature for only nine countries.This is despite strong evidence that unemployment is a strong risk factor for mental illness. As noted before, lost income due to lower employment rate of people with depression alone is estimated to amount to EUR 176 billion per year across EU countries, and these estimated costs would be even higher if other mental health disorders were included. Where they exist, many programmes focus on helping to reintegrate individuals who already had mental health problems, rather than supporting the mental well-being of unemployed persons (McDaid, Hewlett and Park, 2017). A few exceptions can be found. In a suburb of Athens, a centre for psychological support of the long-term unemployed was established in 2013, supported by the European Social Fund and Ministry of Health (Center of Psychosocial Support of Long-term Unemployed, 2016). Given that unemployment is a strong risk factor for mental ill-health, it is important that policies to promote good mental health reach these more vulnerable populations. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 38

41 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Promoting good mental health among older people As the European population ages – more than 18% of the European population is now over 65, and about 5% is over 80 – promoting healthy ageing is a growing policy priority (OECD, 2017). Mental well-being should be a key part of healthy ageing alongside physical health. There are key mental health risks linked to ageing, for instance around the sometimes-difficult transition from work to retirement, or related to physical illness and frailty. Social isolation, loneliness, and lower levels of contact with friends and family can also contribute to lower levels of mental well-being. Equally, older people commonly fall outside of social structures such as schools and workplaces where mental health promotion and illness prevention interventions are more common, as Figure 1.8 shows. To promote mental well-being amongst older populations, interventions have focused on tackling some of the risk factors for mental illness, for example loneliness, and promoting activities that foster mental well-being, for instance through promoting social participation. Although evidence on the cost-effectiveness of interventions for the older population is limited, a systematic literature review including more than 10 countries found that participation in social activities, psychosocial educational interventions, intergenerational activities and volunteering, and some educational activities could help protect the mental well-being of older people (McDaid, 2015). Though far fewer actions to promote the well-being of older people are found than for other parts of the life course, a number of countries are nonetheless intervening with actions primarily to reduce loneliness and isolation. In England, efforts to tackle loneliness amongst older people entailed identifying, signposting, and in some cases funding, of local activities such as lunch clubs, dance afternoons, befriending services, and sports groups (McDaid, Hewlett and Park, 2017). In Norway, government grants are awarded to local areas to create social activities with a social participation component, while in Iceland volunteers from the Icelandic Red Cross make weekly visits to older, ill, or isolated individuals. Conclusions Mental health problems represent a huge burden in terms of morbidity and mortality, and can have devastating consequences on the lives of people experiencing mental ill-health, their friends, relatives and caregivers. More than one in six people across EU countries had a mental health problem in 2016, with an estimated 25 million people suffering from anxiety disorders, 21 million from depressive disorders, 11 million people living with drug and alcohol use disorders, almost 5 million people suffering from bipolar disorder, and schizophrenic disorders affecting an estimated 1.5 million people. For each of these individuals, mental illness will affect their daily lives, their relationships, their jobs, their physical health, their economic status and opportunities. In some cases, mental ill-health leads not just to lives lived less fully, but also to lives lost prematurely: over 84 000 people died of mental health problems and suicides across EU countries in 2015. While mortality rates – driven primarily by deaths from suicide – vary considerably by gender and by country and have been falling over time in almost all countries, each of these deaths must be seen as a tragedy, and no European country can rest easy.The experience of European countries where deaths from suicide have been reduced so substantially are heartening, and offer policy lessons for other countries to follow. The burden of mental illness, and the impact of lives lost from suicide and other causes related to mental ill-health, contribute to significant economic costs in Europe.This HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 39

42 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? chapter estimates total costs related to mental ill-health to be equivalent to more than 4% of GDP. While around one-third of these costs are direct spending on health services, most of these costs relate to social security benefits and the indirect costs of mental ill-health in the labour market, driven by lower employment rates and reduced productivity due to mental illness. Many European countries are taking action to prevent mental illness and to promote mental well-being. More than one hundred interventions to promote good mental health and protect populations from the negative impacts of mental illness were found across the EU, targeting all age groups. Measures are being adopted to promote well-being in schools and nurseries, with new parents, or in workplaces. Reducing stigma and increasing understanding of mental well-being are policy priorities. Furthermore, with improved population-level awareness and understanding of mental health, the stigma around seeking mental health care and talking about mental illness falls. Overcoming stigma and improving diagnosis rates can be expected, in turn, to contribute to more robust data on the true prevalence of mental ill-health. As this chapter shows, mental ill-health is not distributed evenly across the population, and there are important age, gender and socio-economic differences in the burden of disease. Some groups are also less likely to be the target of promotion or prevention interventions. Supporting vulnerable groups, such as older people or unemployed people, is important to build more inclusive and active societies, but at present far fewer policies reach these groups. The dialectic relationship between distance from social structures and deteriorated mental well-being should also not be underestimated. Just as mental ill-health reduces the likelihood of being in employment, unemployment increases the risk of having poor mental health. Programmes that foster good mental health – reducing loneliness, encouraging social participation, building support structures – and interventions that can identify and respond to signs of mental distress, should be priorities for European countries. The growing evidence base along with the significant burden of mental illness make clear that there is a societal case for introducing many such promotion and prevention programmes, but there is also a clear economic case for further investment in this area. Actions to prevent mental illness and promote good mental health can bring lifelong benefits to children and their families, workplace interventions can reduce absenteeism and presenteeism, and suicide prevention strategies can prevent tragic losses of life and potential. The costs of mental illness are extremely high, the potential gains from strengthening mental well-being are significant, and the opportunities for promotion and prevention are far from exhausted. This chapter lays the grounds for a clear case: much more can and must still be done to promote mental well-being and prevent mental ill-health. Note 1. These IHME estimates are lower than those previously reported by Wittchen et al. (Wittchen et al., 2011), partly because they do not include the prevalence of dementia. References Appelqvist-Schmidlechner, K. et al. (2012), “Dissemination and implementation of the Time Out! Getting Life Back on Track programme – results of an evaluation study”, International Journal of . http://dx.doi.org/10.1080/14623730.2012.703045 Mental Health Promotion , Vol. 14/2, pp. 96-108, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 40

43 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Cantone, E. et al. (2015), “Interventions on bullying and cyberbullying in schools: A systematic review”, , Vol. 11/Suppl 1 M4, pp. 58-76, http:// Clinical practice and epidemiology in mental health: CP & EMH dx.doi.org/10.2174/1745017901511010058 . Cavanagh, J. et al. (2003), “Psychological autopsy studies of suicide: A systematic review”, Psychological medicine , Vol. 33/3, pp. 395-405. . Center of Psychosocial Support of Long-term Unemployed (2016), www.menoenergos-pepsaee.gr/index.php/en Clarke, A., B. Bunting and M. Barry (2014), “Evaluating the implementation of a school-based emotional well-being programme: A cluster randomized controlled trial of Zippy’s Friends for children in disadvantaged primary schools”, , Vol. 29/5, pp. 786-798, http:// Health Education Research dx.doi.org/10.1093/her/cyu047 . Coldefy, M. and C. Gandré (2018), “Personnes suivies pour des troubles psychiques sévères : une espérance de vie fortement réduite et une mortalité prématurée quadruplée”, Questions d’economie , Vol. No. 237/September 2018, www.irdes.fr/recherche/questions-d-economie-de-la-sante/237- de la Santé personnes-suivies-pour-des-troubles-psychiques-severes-une-esperance-de-vie-fortement-reduite.pdf (accessed on 14 September 2018). Durlak, J. et al. (2011), “The Impact of Enhancing Students’ Social and Emotional Learning: A Meta- Analysis of School-Based Universal Interventions”, Child Development http:// , Vol. 82/1, pp. 405-432, . dx.doi.org/10.1111/j.1467-8624.2010.01564.x EU Compass Consortium (2017), Good Practices in Mental Health and Well-Being https://ec.europa.eu/health/ , sites/health/files/mental_health/docs/ev_20161006_co05_en.pdf . EU Compass for Action on Mental Health and Wellbeing (2018), Annual Activity Reports of Member States and Stakeholders , EU Compass. EU Compass for Action on Mental Health and Wellbeing (2017), Annual Activity Report of Member States and Stakeholders , EU Compass. European Commission (2016), EU Framework for Action on Mental Health and Well-being . Eurostat (2017), Eurostat Database , http://ec.europa.eu/eurostat/data/database . Fahy, A. et al. (2016), “Longitudinal Associations Between Cyberbullying Involvement and Adolescent http://dx.doi.org/10.1016/ Journal of Adolescent Health Mental Health”, , Vol. 59/5, pp. 502-509, . j.jadohealth.2016.06.006 Farrington, D. and M. Ttofi (2009), “School-Based Programs to Reduce Bullying and Victimization”, Campbell Systematic Reviews . , Vol. 6, http://dx.doi.org/10.4073/csr.2009.6 Ministry of Social Affairs and Health finances the Current Care Guideline for Finnish Government (2017), suicide prevention https://valtioneuvosto.fi/en/article/-/asset_publisher/1271139/stm-rahoittaa-kaypa- , hoito-suositusta-itsemurhien-ehkaisyyn . Gavin, N. et al. (2005), “Perinatal depression”, , Vol. 106/5, Part 1, pp. 1071-1083, Obstetrics & Gynecology http://dx.doi.org/10.1097/01.AOG.0000183597.31630.db . Grundy, E., T. van den Broek and K. Keenan (2017), “Number of children, partnership status, and later- life depression in eastern and western Europe”, The Journals of Gerontology: Series B , http://dx.doi.org/ 10.1093/geronb/gbx050 . Gustavsson, A. et al. (2011), “Cost of disorders of the brain in Europe 2010”, European Neuropsychopharmacology , Vol. 21/10, pp. 718-779, http://dx.doi.org/10.1016/j.euroneuro.2011.08.008 . Hamberg-van Reenen, H., K. Proper and M. van den Berg (2012), “Worksite mental health interventions: Occupational and environmental medicine , Vol. 69/11, A systematic review of economic evaluations”, pp. 837-45, http://dx.doi.org/10.1136/oemed-2012-100668 . Hawton, K. et al. (2016), “Psychosocial interventions following self-harm in adults: A systematic review and meta-analysis”, The lancet. Psychiatry , Vol. 3/8, pp. 740-750, http://dx.doi.org/10.1016/S2215- 0366(16)30070-0 . Hilton, M. et al. (2008), “Mental ill-health and the differential effect of employee type on absenteeism and presenteeism”, Journal of Occupational and Environmental Medicine , Vol. 50/11, pp. 1228-1243, http://dx.doi.org/10.1097/JOM.0b013e31818c30a8 . Holen, S. et al. (2013), “Implementing a universal stress management program for young school children: Are there classroom climate or academic effects?”, Scandinavian Journal of Educational . http://dx.doi.org/10.1080/00313831.2012.656320 Research , Vol. 57/4, pp. 420-444, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 41

44 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? Hoven, C., D. Mandell and J. Bertolote (2010), “Prevention of mental ill-health and suicide: Public health European Psychiatry http://dx.doi.org/10.1016/J.EURPSY.2010.01.011 . perspectives”, , Vol. 25/5, pp. 252-256, Hutson, E., S. Kelly and L. Militello (2018), “Systematic review of cyberbullying interventions for youth and parents with implications for evidence-based practice”, , Worldviews on Evidence-Based Nursing Vol. 15/1, pp. 72-79, http://dx.doi.org/10.1111/wvn.12257 . Ibanez, G. et al. (2015), “Effects of antenatal maternal depression and anxiety on children’s early cognitive development: A prospective cohort study”, PLOS ONE , Vol. 10/8, http://dx.doi.org/10.1371/ journal.pone.0135849 . IHME (2018), Global Health Data Exchange , www.healthdata.org/ . Jarde, A. et al. (2016), “Neonatal outcomes in women with untreated antenatal depression compared with women without depression”, JAMA Psychiatry ,Vol.73/8,pp.826, http://dx.doi.org/10.1001/ jamapsychiatry.2016.0934 . Kessler, R. et al. (2007), “Age of onset of mental disorders: A review of recent literature”, Current Opinion , Vol. 20/4, pp. 359-364, . in Psychiatry http://dx.doi.org/10.1097/YCO.0b013e32816ebc8c Kessler, R. et al. (2007), “Lifetime prevalence and age-of-onset distributions of mental disorders in the World psychiatry: Official World Health Organization’s World Mental Health Survey Initiative”, , Vol. 6/3, pp. 168-76. journal of the World Psychiatric Association (WPA) Korkeila, J. (2013), “2781 – Suicide prevention in Finland: Advances and perspectives”, European Psychiatry , Vol. 28, pp. 1, http://dx.doi.org/10.1016/S0924-9338(13)77371-1 . Kovess-Masfety, V. et al. (2016), “Comparing the prevalence of mental health problems in children 6-11 Social Psychiatry and Psychiatric Epidemiology , Vol. 51/8, pp. 1093-1103, http:// across Europe”, dx.doi.org/10.1007/s00127-016-1253-0 . Kowalski, R. et al. (2014), “Bullying in the digital age: A critical review and meta-analysis of cyberbullying research among youth”, Psychological Bulletin , Vol. 140/4, pp. 1073-1137, http:// . dx.doi.org/10.1037/a0035618 Liu, N. et al. (2017), “Excess mortality in persons with severe mental disorders: A multilevel intervention framework and priorities for clinical practice, policy and research agendas”, World , Vol. 16/1, pp. 30-40, psychiatry: Official journal of the World Psychiatric Association (WPA) http:// dx.doi.org/10.1002/wps.20384 . Matsubayashi, T. and M. Ueda (2011), “The effect of national suicide prevention programs on suicide rates in 21 OECD nations”, , Vol. 73/9, pp. 1395-1400, http://dx.doi.org/10.1016/ Social Science & Medicine j.socscimed.2011.08.022 . McDaid, D. (2015), Independence and Mental Wellbeing for older people . McDaid, D., E. Hewlett and A. Park (2017), “Understanding effective approaches to promoting mental , OECD Health Working Papers , No. 97, OECD Publishing, Paris, health and preventing mental illness” http://dx.doi.org/10.1787/bc364fb2-en . McDaid, D. and A. Park (2014), “Investing in Wellbeing in the Workplace”, in Wellbeing , John Wiley & Sons Ltd, http://dx.doi.org/10.1002/9781118539415.wbwell105 . McLoughlin, A., M. Gould and K. Malone (2015), “Global trends in teenage suicide: 2003-2014”, QJM , Vol. 108/10, pp. 765-780, http://dx.doi.org/10.1093/qjmed/hcv026 . Munizza, C. et al. (2013), “Public beliefs and attitudes towards depression in Italy: A national survey”, PLoS ONE http://dx.doi.org/10.1371/journal.pone.0063806 . , Vol. 8/5, pp. e63806, for Health and Care Excellence (2018), Antenatal and postnatal mental health: Clinical National Institute . management and service guidance/Guidance and guidelines/NICE Nock, M. et al. (2009), “Cross-National Analysis of the Associations among Mental Disorders and Suicidal Behavior: Findings from the WHO World Mental Health Surveys”, PLoS Medicine , Vol. 6/8, http://dx.doi.org/10.1371/journal.pmed.1000123 . Observatoire national du suicide (2018), Suicide : enjeux éthiques de la prévention, singularités du suicide à l’adolescence – 3e rapport/février 2018 . OECD (2018), Education indicators in focus #60: How is depression related to education? , OECD. OECD (2017), PISA 2015 Results (Volume III): Students’ Well-Being , PISA, OECD Publishing, Paris, http:// dx.doi.org/10.1787/9789264273856-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 42

45 I.1. PROMOTING MENTAL HEALTH IN EUROPE: WHY AND HOW? OECD (2015), , Mental Health and Fit Mind, Fit Job: From Evidence to Practice in Mental Health and Work http://dx.doi.org/10.1787/9789264228283-en Work, OECD Publishing, Paris, . Recommendation of the Council on Integrated Mental Health, Skills and Work Policy ,OECD/ OECD (2015), LEGAL/0420, . https://legalinstruments.oecd.org/en/instruments/334 OECD (2014), Making Mental Health Count: The Social and Economic Costs of Neglecting Mental Health Care , OECD Health Policy Studies, OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264208445-en . OECD (2012), , Mental Health and Work, Sick on the Job?: Myths and Realities about Mental Health and Work OECD Publishing, Paris, http://dx.doi.org/10.1787/9789264124523-en . Patana, P. (2014), “Mental Health Analysis Profiles (MhAPs): Finland” , OECD Health Working Papers , No. 72, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jz1591p91vg-en . Samele, C., Frew Stuart and Urquia Norman (2013), European profile of prevention and promotion of mental health (EuroPoPP-MH) . Titelman, D. et al. (2013), “Suicide mortality trends in the Nordic countries 1980-2009”, Nordic Journal of Psychiatry . http://dx.doi.org/10.3109/08039488.2012.752036 , Vol. 67/6, pp. 414-423, Wahlbeck, K. et al. (2011), “Outcomes of Nordic mental health systems: Life expectancy of patients with mental disorders”, , Vol. 199/06, pp. 453-458, http://dx.doi.org/ British Journal of Psychiatry . 10.1192/bjp.bp.110.085100 Weare, K. and M. Nind (2011), “Mental health promotion and problem prevention in schools: What Health Promotion International , Vol. 26/suppl 1, pp. i29-i69, http://dx.doi.org/ does the evidence say?”, 10.1093/heapro/dar075 . WHO (2018), Mental Health Atlas 2017 , World Health Organization, Geneva. WHO (2014), Preventing suicide suicide A global imperative , World Health Organization. WHO (2013), , World Health Organization. Mental Health Action Plan 2013-2020 WHO (2001), “Mental health: Strengthening mental health promotion” , 220 , World Health Organization. Wittchen, H. et al. (2011), “The size and burden of mental disorders and other disorders of the brain in Europe 2010”, http://dx.doi.org/10.1016/j.euroneuro.2011.07.018 . Zalsman, G. et al. (2016), “Suicide prevention strategies revisited: 10-year systematic review”, The . http://dx.doi.org/10.1016/S2215-0366(16)30030-X Lancet Psychiatry , Vol. 3/7, pp. 646-659, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 43

46

47 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART I Chapter 2 Strategies to reduce wasteful spending: Turning the lens to hospitals and pharmaceuticals Evidence suggests that as much as one-fifth of health spending is wasteful, and could be reduced or eliminated without undermining health system performance. With as much as 9.6% of European GDP directed to health care, reducing such spending is thus important not only for improving access to needed care, but also for ensuring health system resilience. This chapter points the lens at two particular areas of waste: hospitals and pharmaceuticals. Hospitals represent an integral and essential component of any functioning health system, but are often the most expensive part. In many instances, the resources consumed in hospitals can be put to more efficient use. Improved community care for chronic diseases could reduce millions of avoidable admissions and bed days across EU countries. Reducing unnecessary investigations and procedures would not compromise quality. Greater use of day surgery and reducing delays in discharging patients no longer requiring inpatient care could also free-up resources for patients with greater needs. Minimising waste and optimising the value derived from expenditure on pharmaceuticals are also critical to efficient and sustainable health systems. This chapter discusses a mix of supply and demand side policy levers that include ensuring value for money in selection and coverage, procurement and pricing of medicines; exploiting the potential of savings from generics and biosimilars; encouraging rational prescribing and use; and improving adherence to treatment. Ultimately, progress in reducing wasteful spending may be seen not only as a barometer of quality improvement, but also an ethical and financial imperative in the pursuit of more resilient and equitable health care systems. 45

48 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Introduction Reducing wasteful spending in health is an important objective in both good and bad economic times. In an economic downturn, properly targeting wasteful spending in health care can help ensure that cost-containment efforts do not compromise quality and outcomes, thus contributing to the health system’s resilience. In better times, reducing wasteful spending in health is increasingly seen as a sound quality improvement strategy. It can also release resources that can be better targeted to improving the system’s accessibility. In other words, reducing waste can contribute to improving health system performance along several dimensions. Evidence suggests that as much as one-fifth of health spending is wasteful and could be eliminated without undermining health system performance (OECD, 2017). This alarming estimate – seldom challenged by experts – is well supported by available research. For example, in 2012, a sample of physicians polled in France reported that on average they viewed 28% of interventions as not fully justified (Vanlerenberghe, 2017). A study in the Netherlands estimated that 20% of expenditure on acute care could be saved by reducing overuse, increasing the integration of care, and involving patients in care decisions (Visser et al., 2012). In Italy, a country that spends less on health than many other Western European countries, the proportion of inefficient or wasteful public spending was estimated to be around 19% in 2017 (Fondazione GIMBE, 2018). Wasteful spending can take many forms (as illustrated in Figure 2.1) and has a range of effects: ● Patients are unnecessarily harmed, or receive unnecessary or low-value care that makes little or no difference to their health outcomes. ● The same outcomes can be achieved with fewer resources. For example, some health systems have low utilisation of generic medicines; others provide care in resource- intensive places such as hospitals, when it could be provided in the community. ● A number of administrative processes add no value, and funds are lost to fraud and corruption. With up to 9.6% of Europe’s GDP devoted to health care in 2017, waste serves only to undermine the financial sustainability of health systems. Pursuing efficiency in health spending and maintaining access to services are persistent, but at times conflicting policy challenges in most European countries. Tackling wasteful spending can only work to improve value for money and support both. In this chapter, the lens is pointed squarely at two particular areas of waste: hospitals and medicines. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 46

49 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS A pragmatic approach to identifying and categorising wasteful Figure 2.1. spending on health Unnecessary duplication of tests and ices serv voidable adverse events A Patients do not r e ceive Low-value care: ineffective, inappropriate, the right care not cost-effective Discarded inputs g (e. . unused medicines) Benefits could Waste occurs t be ob ained with when ... Overpriced input fewer resources (e.g. generic vs brand medicines) High cost inputs used unnecessarily (e.g. physician instead of nurse, inpatient instead of outpatient care) Resources are e unnec ssarily taken away from patient care Administrative waste F rau d, abuse and corruption , http://dx.doi.org/10.1787/9789264266414-en Tackling Wasteful Spending on Health . Adapted from OECD (2017), Source: Addressing wasteful spending in hospitals Hospitals represent an integral and essential part of any functioning health system. Yet, as illustrated in Figure 2.2, resources consumed in hospitals could be put to more efficient use. For example, improved community care for ambulatory care-sensitive conditions could reduce avoidable admissions. Tackling the overuse of hospital services could reduce the resources used during a necessary admission without compromising quality. Other opportunities to deploy available hospital resources more efficiently include more extensive use of day surgery in place of inpatient care. This, together with other strategies directed to reducing discharge delays, can help ensure that patients leave the hospital as early as possible. These examples are discussed in turn below. Figure 2.2. Pressure points on wasteful hospital spending Deploy day-surgery Tackle hospital services overuse Curb delayed Improve community discharges care of ACSCs Increase efficiency and safety to reduce the use of hospital resources Ensure patients Reduce leave hospital as unnecessary early aspossible hospital admissions HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 47

50 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Reducing potentially avoidable admissions Potentially avoidable hospital admissions for some chronic conditions consume over 37 million bed days each year A large number of hospital admissions could be averted through better prevention and management of both acute and chronic conditions outside the hospital. Among more than 30 conditions for which hospitalisation could be reduced with better primary care (also referred as ambulatory care-sensitive conditions) (Purdy et al., 2009), five stand out as particularly relevant in European countries: 1) diabetes, 2) hypertension, 3) heart failure, 4) chronic obstructive pulmonary disease (COPD) and bronchiectasis and 5) asthma. Across the EU, over 4.6 million admissions were made for these five conditions in 2015 – 1 amounting to 5.6% of all admissions which might have been avoided (Table 2.1). The average length of stay (ALOS) for these five conditions was 8.1 days, which exceeded Table 2.1. Hospital admissions for five chronic conditions, EU countries, 2015 COPD and Total (five Heart failure Diabetes Hypertension Asthma Bronchiectasis conditions) 4 653 924 Admissions/discharges 800 303 665 396 1 749 384 1 109 865 328 976 % of all admissions 2.1% 1.3% 0.4% 5.6% 0.8% 1.0% (avg.) 6.9 9.5 8.9 6.6 8.1 Average LOS (days) 8.5 Total bed days 6 794 572 4 597 886 16 619 148 9 855 601 2 177 821 37 603 706 Proportion of all bed days 1.1% 0.7% 2.7% 1.6% 0.4% 6.5% The data on hospital admissions refer to discharges (including deaths in hospital). They include patients in all Note: age groups, but exclude outpatient and day cases (patients who do not stay overnight in hospital).The number of bed days was calculated by multiplying the number of admissions (discharges) by ALOS. The total number of admissions (discharges) excludes healthy neonates. OECD Health Statistics, https://doi.org/10.1787/health-data-en and Eurostat database. Source: Figure 2.3. Share of potentially avoidable hospital admissions due to five chronic conditions, EU countries, 2015 Share of total discharges (%) Number of discharges per 100 000 population % Per 100 000 population 3000 10 8.6 8.5 2400 8 6.4 6.3 6.3 6.2 6.1 6.1 6.0 5.8 5.7 5.7 5.5 1800 6 5.3 5.1 5.1 5.1 4.9 4.8 4.8 4.8 4.8 4.7 4.2 4.1 3.9 3.8 3.7 3.7 1200 4 3.3 600 2 0 0 Note: The data on hospital admissions refer to discharges related to five chronic conditions: diabetes, hypertension, heart failure, COPD and bronchiectasis and asthma. They include patients in all age groups, but exclude outpatient and day cases (patients not staying overnight in hospital).These potentially avoidable hospital admissions do not control for any variation in the prevalence of these five chronic conditions. Estonia and Greece are not shown due to missing data for several of these causes of hospitalisation. Data for Cyprus are not shown as they only include discharges from public hospitals, resulting in substantial under-estimation as most hospitals are private. and Eurostat Database. OECD Health Statistics, https://doi.org/10.1787/health-data-en Source: 1 2 http://dx.doi.org/10.1787/888933834129 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 48

51 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS the ALOS for all causes of hospitalisation (7.4 days). In total, admissions for these 5 conditions represented over 37 million bed days in 2015. Cross-country comparisons of potentially avoidable hospital admissions should be interpreted with caution, as many other factors, beyond better access to primary care, can influence the statistics, including data comparability and the prevalence of these chronic conditions. Nevertheless, admission rates for these five chronic conditions were particularly high in Bulgaria, Romania, Germany, Lithuania, Austria and Hungary, while as a proportion of all hospital admissions, rates were highest in Bulgaria and Romania, followed by Poland, Germany, Spain and Hungary (Figure 2.3). Reducing admissions requires meeting people’s needs outside of the hospital Recognising the need to improve access to care outside hospitals, many EU countries have taken steps to increase the availability of primary and community care, and to introduce new models of intermediate care that can serve as alternatives to hospitals. Many people present to hospitals simply because their primary care providers are unavailable. To address this, a number of countries have increased access to after-hours primary care. For example: ● In the Netherlands, after-hours care is organised at the municipal level in GP “posts”. These posts are generally situated near or within hospitals in order to provide urgent primary care overnight, and work closely with emergency departments. Nearly all GPs work for a GP post. Specially trained assistants respond to phone calls and perform triage, with GPs then determining referrals to hospital. GPs are paid at hourly rates for after-hours work and must provide at least 50 hours of after-hours care per year to maintain their GP registration. As GP care in the Netherlands is free at the point of service, and a mandatory deductible applies for (emergency) hospital care, patients have a financial incentive to choose GP posts over the Emergency Department (Wammes et al., 2017). ● In Denmark, after-hours care is organised by the regions. The first line of contact is a regional telephone service, most often answered by a physician or sometimes a nurse in Zealand and the Copenhagen region who decides whether to refer the patient for a home visit or to an after-hours clinic, usually co-located with a hospital emergency department. GPs can choose to take on more or less work within this programme and receive a higher rate of payment for after-hours work (Vrangbaek, 2017). ● In 2017, Portugal established a call centre that operates around the clock and, among other services, provides guidance to patients based on their needs. Among 800 000 callers in 2017, 26% were advised on self-care, 42% referred to physicians and 24% directed to emergency services. Some countries have also started to develop intermediate in-home care services as an alternative to hospital-based ones. For example: ● In the United Kingdom, since 2005 “virtual wards” have been set up in some parts of the country to provide care at home for people recently discharged or at high risk of hospital (re-)admission. Care is provided through multi-disciplinary care teams. Evidence suggests that these “virtual wards” have reduced unplanned hospital (re)-admissions and the length of stay in hospitals for the most at risk groups (Sonola et al., 2013). HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 49

52 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS ● In France, the “hospital at home” model, organised and funded through hospitals, is designed to offer patients the option of receiving hospital care at home for certain conditions. In 2016, more than 100 000 patients in France were treated in a “hospital at home” programme, equivalent to 175 000 admissions, an increase of 8% over 2015 (FNEHAD, 2017). Most countries recognise that in order to respond effectively to the needs of ageing populations and the growing burden of chronic disease, further efforts are needed both to strengthen access to primary care and to provide more continuous and coordinated care outside hospitals. Measuring and addressing overuse in hospitals Unfortunately, not all care received by hospitalised patients is necessary, and in some cases, may not only be futile but even cause harm. Many services that are delivered offer only very modest benefit to patients, or are of benefit only to some, and in some cases the evidence of benefit is weak or lacking altogether (Brownlee et al., 2017). In a recent effort to identify services overused in hospitals, researchers reviewed more than 800 recommendations targeting low-value services issued in the United States, Canada, Australia and the United Kingdom, and found that two-thirds of them pertained to services delivered in hospitals (Chalmers et al., 2018), including investigations and surgical procedures. Another recent study in the United Kingdom identified 71 low-value interventions performed in general surgery alone (Malik et al., 2018). One in four European countries has now systematically documented unwarranted variation in the use of hospital services using Atlases Detecting and measuring wasteful spending on low-value care has mobilised considerable effort over the years, with two main approaches currently in use. The first consists broadly of comparing utilisation rates for specific low-value services across geographic areas, adjusting for population need (for lack of better indicators, generally using age and gender as proxies). These analyses invariably display very large and unwarranted variations in utilisation that cannot be explained by differences in disease burden, standards of care, or patient preference, especially within countries. For example, in 2011 caesarean-section (C-section) rates in Italy varied by a ratio of 1 to 6 across local health units (OECD, 2014). In 2015 they varied to a similar degree across areas in Spain (on-line Spanish Atlas, see below) and by a ratio of 1 to 2 across French Départements in 2014 (Le Bail and Or, 2016). This approach, a hallmark of the US Dartmouth Atlas of Health Care, has been used in at least five European countries to generate “Atlases of variation in health care” (Table 2.2). Additionally, in 2014, in the context of the EU-funded ECHO project, Slovenia, Denmark and Portugal produced atlases of low value care. These atlases cover a similar set of services, in particular elective surgery. They help raise public awareness about the problem of overuse and may catalyse behaviour change, but their operability is limited as they do not typically identify when, for whom, and which specific providers’ services may have been over or under-provided. Nevertheless, this comparative approach can help identify areas where overuse is systemic, as overuse of various services is often correlated in a given area (Miller et al., 2018). HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 50

53 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Examples of hospital interventions adolescents,cholecystectomy,hysterectomy,prostatectomy, surgery, considered generally effective and thus less subject hysterectomy, knee replacement, thyroidectomy. Hip fracture tonsillectomy and appendectomy in children and The 2015 atlas examines 9 surgical procedures: C-section, 11 surgeries Tonsillectomy, appendectomy, C-section, bariatric surgery, prostate surgery, carpal tunnel surgery, cholecystectomy, replacement. 8 surgeries: Hip replacement, knee replacement, knee arthroscopy, carpal tunnel surgery, cataract surgery, carotid artery surgery, hysterectomy, C-section. Bariatric surgery, myringotomy, phlebotomy, medical imaging (MR,CT), appendectomy, tonsillectomy, knee and hip replacement. knee replacement, coronary bypass, implantation of a relatively few interventions known to be of variable value. differences in the coverage of appropriate interventions but The last edition emphasises quality indicators and Atlas (up from 34 in the first edition). Around 100 indicators are mapped in the 2016 compendium defibrillator. 11 atlases relate to hospital procedures including orthopaedic surgery (hip fracture, knee and hip replacement), general surgery, paediatric hospitalisations, cardiovascular procedures (including stroke management), diabetes care, cancer care, hospitalisations for mental health problems, avoidable hospitalisations for frail patients or chronic conditions and procedures considered lower-value care. to unwarranted variation was used as a benchmark. Exceptions for hospital care include tonsillectomy and hip Approach and scope elective surgical procedures. The 2015 atlas analyses geographic variations for a set of literature suggesting a proportion of procedure is low-value. Analysis of geographic variations in “priority” surgical (>20 000) and increasing volume, large variations, initiatives authorities, evidence of relevance from international in place addressing these services or interest by various Analysis of geographic variations in elective surgical procedures. recommended service) or outcome (mortality, survival). indicators assessing process (% of people who receive factor, volumes of specific services provided, or quality indicators are presented which may include morbidity, risk Variations in standardised utilisation rates are analysed for 3 categories of services: i) proven to be effective, e.g. hip fracture repair ii) services whose effectiveness is uncertain Selection criteria included high and increasing volume, iii) generally considered lower-value care, e.g. spinal fusion. The latest atlases are presented in an interactive online platform where different indicators can be explored across zones. evidence of relevance from international literature. Interventions are prioritised based on importance (spending), convenience (data available) and relevance (literature, policy debate). A standard methodology is used across procedures selected to represent all specialties. procedures. Selection criteria for procedures included: high propose topics for review. each programme, geographic variations for an ad-hoc set of budget programmes (cancer, mental health disorders). For Stakeholders in the health care system as well as citizens can (prevalence) and per international literature. Selection criteria included relevance to the general public variations for the main categories of diseases covered by NHS The “compendium” atlases highlight unwarranted geographic beyond appropriately selected groups of patients, e.g. C-section, Atlases of variations in health care in Europe Stated objective outcomes and activity. The goal is Identify unwarranted variations in Measure variations to highlight “efficiency shortcomings and policy debate. 17 decentralised regions to inform services within and across population use of specific hospital unwarranted variations in Identify systematic and quality deficits”, spark public debate and encourage the evidence-based NHS services to all improve value in the system. variation in practice. by documenting unwarranted Promote appropriate care in elective surgery. Highlight unexplained variations Reduce unwarranted variations in inappropriate care, thereby improving quality and reducing cost. to ensure provision of same quality patients. development of measures to Table 2.2. Authors Aragon INAMI (National Institute for Health and Disability Insurance) in consultation with relevant KCE (public research centre centre and experts. in collaboration with a research IRDES (research centre), in collaboration with Ministry of Health and with contribution from medical societies (independent think tank), Bertelsmann Foundation in health) Consortium of around 50 NHS and Public Health England (Department of Health agency), researchers coordinated by specialists research units in Valencia and Author’s analysis based on Atlases. variation 2016 France Country (2017) Diagnostic tests (2017) Thematic Atlases: Liver disease Germany 2015 Healthcare Fact Check: The development of regional variations Previous edition: 2011 Since 2012 Thematic fact-checks also prepared on specific interventions (e.g. C-sections, back surgery) Since 2012 Previous editions: 2010 and 2011. Healthcare (Compendium) Atlas of medical practice 2016 England updated thematic atlases Practice): A platform of regularly Atlas VPM (Variations in Medical Since 2006 Spain 100 procedures Thematic Atlases for over Under preparation surgery in Belgium Analysis of variations in elective 2006 Belgium Time period Document The NHS Atlas of Variation in Source: HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 51

54 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS The other strategy to measure the extent of overuse is more direct, and consists of identifying, by using patient-level data, those services likely to have been delivered inappropriately. Analyses of service delivery records are undertaken to identify the characteristics of those patients who should not have received a particular service. This analysis can produce estimates of the amount of resources “wasted”, but is limited to those services for which the criteria for appropriateness are sufficiently specific and can be mapped to available data. By aggregating analyses across services it is possible to build bottom-up estimates of wasteful spending. In the United Kingdom, a recent study of services in general surgery using a similar approach identified a potential EUR 153 million which could be saved annually by NHS England (Malik et al., 2018). Rates of C-sections are still growing in a third of EU countries C-sections are a prime example of a surgical procedure which can save lives when clinically indicated, but for which the benefits are disputed. At population level, C-section rates above 15% of deliveries are not associated with reductions in maternal, neonatal or infant mortality (Stordeur et al., 2016). Yet, in 2016, on average 28% of babies were born by C-section in Europe, a rate that varies more than threefold between the Netherlands (16%) and Cyprus (55%). C-section rates began increasing rapidly in the 1980s and continued to rise on average by more than 6% per annum between 2000 and 2005. The growth rate slowed to 2.6% per annum between 2005 and 2010 and further decreased to 1.2% over the 2 following 5 years (Figure 2.4). In many countries, elective C-section among low-risk women is among the first procedures for which interventions aimed at reducing overuse have been introduced. compares European country levels and trends in C-section rates over the last Figure 2.4 10 years, with the centre of the graph representing the European average for both metrics. In many countries in Central and Eastern Europe, C-section rates have risen very rapidly over the past decade, and are very high (most notably in Poland, Romania and Bulgaria), suggesting overuse may have yet to receive much attention. In contrast, many countries of Northern Europe have considerably lower C-section rates, and these have remained fairly stable over the last 10 years. Nordic countries have traditionally had low C-section rates, while a number of other countries in which rates have increased slowly or even declined have put in place specific policies to target this. Tackling overuse is likely to require multi-pronged strategies that engage patients and clinicians in particular Policies targeting patients and providers to address overuse revolve around three types of levers: ● Producing and publishing information on overuse. This can i) raise awareness; ii) enable better informed conversations between providers and patients (as illustrated by the 3 Choosing Wisely® campaign ); or iii) serve to benchmark providers against their peers. For example, all maternity units in Belgium receive confidential annual reports detailing their obstetric indicators and comparing them with other maternity units, encouraging poor performers to question their practices. Supporting behaviour change through, for example, clinical decision-making support ● tools or feedback and audits. In 2013, France offered methodological support to maternity units that volunteered to undertake practice analyses and develop change strategies. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 52

55 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS C-section rates in 2016 and their annual growth rate between 2006 and 2016 Figure 2.4. 8% Romania Poland 7% Bulgaria 6% 5% Czech Slovak Republic 4% Republic Croatia Hungary 3% Ireland Slovenia EU Average 2% Austria UK 1% Netherlands Luxembourg Estonia Belgium Germany Sweden Latvia Average annual growth rate of C-sections, past 10 years France 0% Denmark Finland Portugal Lithuania Spain Italy -1% Malta -2% 100 150 200 250 450 400 350 300 Number of C-sections per 1 000 live births Note: Cyprus is not represented due to a break in the data series, but the 2016 C-section rate is by far the highest in the EU (554 per 1 000 live births). The annual growth rate for Luxembourg only covers the period 2011 to 2016 due to a break in the series in 2011. Data are not available for Greece. Eurostat, except Netherlands: Perinatal registry ( www.perined.nl/ ). Source: http://dx.doi.org/10.1787/888933834148 1 2 Financial levers, such as payment systems limiting incentives for providers to deliver ● low-value services, or limiting service coverage to circumstances where comparative effectiveness is documented. Financial incentives are used in France, Portugal, and Italy, targeting procedure prices, hospital budgets and regional budget allocations respectively. Table 2.3 summarises the strategies used by a handful of countries to reduce C-section rates and provides additional concrete examples. Although impact evaluations are lacking, the interventions presented above are believed to have contributed to slowing the growth or reducing C-section rates in countries that have implemented them. However, to date achieving significant and sustained impact in reducing the overuse of various investigations and surgical procedures has proven elusive. Addressing overuse is complex and requires systemic effort and multi-pronged strategies; evidence of impact is often incomplete and system-dependent (OECD, 2017; Mafi and Parchman, 2018; Ellen et al., 2018; Elshaug et al., 2017). Nevertheless, reducing unwarranted use is a quality- enhancing strategy which offers the potential to free-up significant resources in the health system. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 53

56 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Table 2.3. Examples of strategies to reduce C-section rates in Europe Provide direct behaviour Leverage information Levers Observed change Financial incentives change support and raise awareness New best practice guidelines Reduction in programmed France Between 2009-12, France reduced Campaign providing methodological support to the gap between DRG prices for (2012). C-section and increase in the Started in 2010 (Haute Autorité proportion programmed after vaginal delivery and C-section User-friendly flyers for expectant volunteer maternities in their from 40% to 16% (13% in 2018) 39 weeks, with a more marked efforts to analyse and improve de Santé, 2016) mothers detailing the indications for C-sections. improvement in maternities which in public hospitals (17% for private practice (2013). C-section justification must be hospitals in 2018). joined the programme supporting behaviour change. documented in patient’s medical file. Funding of public hospitals C-section rates are 10% lower Portugal Creation of a national commission Information sessions in Started in 2010 indexed to C-section rates. in 2016 than in 2010. maternities with a C-section rate to control C-section rates. above 35% and discussions of Debates in policy and scientific (Ayres-De-Campos circles. options to reduce C-sections et al., 2015) (in the region with highest rates). Training of health professionals (in the region with highest rates). New practice guidelines (2015). Belgium All maternity units receive a yearly confidential report detailing (Stordeur et al., obstetric indicators and 2016) comparting them with other maternity units. Since 2012, C-section rates are Italy Patients can look-up the C-section one of 35 indicators for which rate in any given hospital in a regional targets are set and website run by the National Outcomes Programme of the monitored by the MoH. Good performance across these Ministry of Health (MoH). indicators and progress is rewarded by a 3% increase in the health budget to regions. Exploiting the potential of day surgery Greater use of day surgery can also reduce the utilisation of hospital resources, with the added benefit that most patients prefer day surgery as it allows them to return home the same day. The use of day surgery has increased in all EU countries over the past few decades, thanks to progress in surgical techniques and anaesthesia, but the pace of diffusion has varied, with some countries leading the way in adopting day surgery earlier 4 and faster, and others still lagging behind. The diffusion of day surgery varies widely across EU countries The trends in the adoption of day surgery presented here focus on four high-volume surgical procedures: cataract surgery, tonsillectomy, inguinal hernia repair, and laparoscopic 5 cholecystectomy. The diffusion of day surgery varies greatly both across these four surgical procedures and across countries. While almost all cataract surgery is now performed as day surgery in most EU countries, the average rate of day surgery in 2015 was 40% for inguinal hernia repairs, 32% for tonsillectomies and 13% for laparoscopic cholecystectomies. The 22 EU countries included in the analysis can be classified into three groups in terms of adoption of day surgery: advanced adopters, moderate adopters, and low adopters (Figure 2.5). The Nordic countries and the United Kingdom have led the way in adopting day surgery for a growing number of interventions, and the Netherlands has also expanded day surgery more rapidly than most other EU countries. Nearly all cataract operations in Denmark, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 54

57 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Nordic countries have led the way in adopting day surgery, Figure 2.5. whereas countries in Central and Eastern Europe have generally lagged behind n.a. Legend Low adopters Moderate adopters n.a. Advanced adopters n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. (Malta) n.a. Note: The grouping of countries is based on an analysis of the distance of the country to the EU average for each of the four selected surgical procedures in 2015. Data are not available for Bulgaria, Czech Republic, Greece, Latvia, Slovak Republic and Switzerland. Data for Cyprus are not shown as they only include discharges from public hospitals, resulting in a large bias given that most hospitals are private. Source: and Eurostat Database. https://doi.org/10.1787/health-data-en OECD Health Statistics, Finland, Sweden and the United Kingdom have been performed as day surgery for well over a decade (F anel A). Day surgery rates for inguinal hernia repair and tonsillectomy igure 2.6, P are also much higher in these countries (over 70% and over 50% respectively) than in other EU countries, and laparoscopic cholecystectomy is also increasingly performed as day surgery, and rates now reaching at least 30% in Sweden and over 50% in Denmark. Several countries in Western Europe (Belgium, France and Ireland) and in Southern Europe (Portugal, Spain, Italy and Malta) have been moderately fast adopters of day surgery. In many of these countries, day surgery has grown fairly rapidly over the past decade for some interventions, for example cataract surgery in Fr Panel B) igure 2.6, ance and Portugal (F but remains much more limited for other interventions such as inguinal hernia repair and laparoscopic cholecystectomy. However, national averages often mask large variations within countries. For example, in Belgium day surgery rates for laparoscopic cholecystectomy range from nil in many hospitals yet to adopt this practice, to 50% or 60% in those hospitals that have been leading the way (Leroy et al., 2017). This indicates that a lot of scope remains in this group of countries to expand day surgery further. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 55

58 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Figure 2.6. Diffusion of day surgery between 2005 and 2016 in selected EU countries y stectom y ic cholec p arosco p La y Tonsillectom uinal hernia g air of in p Re y er g Cataract sur Panel A. Advanced adopters Denmark Finland % % 100 100 75 75 50 50 25 25 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Sweden United Kingdom % % 100 100 75 75 50 50 25 25 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Panel B. Moderate adopters Belgium France % % 100 100 75 75 50 50 25 25 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Spain Portugal % % 100 100 75 75 50 50 25 25 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 and Eurostat Database. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: http://dx.doi.org/10.1787/888933834167 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 56

59 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS In Austria, Germany and several countries in Central and Eastern Europe (e.g. Hungary, Poland and Romania), the diffusion of day surgery for most interventions has generally been much slower. While progress has been made on cataract procedures, the use of day surgery for most other interventions remains much more limited. The indicator on “Day surgery” in Chapter 8 shows low day surgery rates in these countries for inguinal hernia repair and tonsillectomy. Further progress in day surgery could help achieve substantial savings in hospital expenditure. For example, a recent report in France estimated that an increase of 3 percentage points in day surgery could result in savings of EUR 200 million per year (CNAMTS, 2018). Enabling greater diffusion of day surgery A number of barriers and enabling factors can influence the uptake of day surgery not only across countries, but also across different hospitals within each country. The same broad types of policy levers that can be used to reduce the overuse of diagnostic tests and treatments can also be used to reduce the unnecessary hospitalisation of patients who could instead be managed with day surgery: ● Publicly reporting the use of day surgery at different levels (national, regional and hospital levels) can play an important role in monitoring progress. One good example of such regular monitoring is the release of the British Association of Day Surgery’s Directory of Procedures, which is accompanied by a national dataset identifying the best performers in the use of day surgery for up to 200 interventions (BADS, 2016).The Belgian Health Care Knowledge Centre also released a comprehensive report in 2017 reporting variations in day surgery rates between Belgium and other neighbouring countries, as well as between the three Belgian regions, and across hospitals (Leroy et al., 2017). ● Providing required support for behavioural and clinical change is also important, so that lagging hospitals or hospital units can learn from and catch up with the most innovative and best performers. Experience in many countries shows that the development of day surgery is often led by “local champions” who drive change in clinical practice. Providing proper financial incentives to ensure that health care providers (hospitals and ● surgical teams) do not lose revenue by moving towards a greater use of day surgery, and may even be financially better-off, is also key. The Best Practice Tariffs in England provide a good example of an explicit policy to incentivise moves toward day surgery (see below). These interventions are likely to be more effective if they are part of a comprehensive strategy to promote day surgery and are led by clinicians. In Sweden, one of the main factors that has contributed to the expansion of day surgery over the past few decades has been clinical leadership in the adoption of evidence-based guidelines to streamline pre- and post-operative surgical procedures, and promote safe and effective use of day surgery. Nationwide collaboration and support from national authorities have helped to set up and disseminate new standards, while leaving sufficient autonomy to enable adaptation to local circumstances. The expansion of day surgery has helped achieved substantial savings, but further progress is still possible. A 2016 review by the National Board of Health and Welfare showed that the costs of the 11 most common types of procedures would have been 14% higher if the share of day surgery had not increased between 2005 and 2013. However, the review also pointed out that the full cost saving potential has not yet been reached, as the share of day surgery still varied widely across the 21 regional health administrations. For HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 57

60 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS example, the rate of day surgery for tonsillectomy varied between 4% and 94% in 2013 (Tiainen and Lindelius, 2016). In the United Kingdom, the British Association of Day Surgery (BADS) has played an instrumental role in the development of day surgery in England by gradually expanding the list of procedures deemed suitable and safe for day surgery from 20 in 1990 to more than 200 procedures in 2016 (BADS, 2016). A national dataset also accompanies this Directory of Procedures, providing the latest data on the percentage of procedures successfully performed as day cases and for each procedure also indicating the performance of the top 5%, 25% and 50% of hospitals. Since 2009, the BADS has also worked with the Department of Health to develop Best Practice Tariffs to provide financial incentives to support the further development of day surgery. By initially paying a relatively higher price for day surgery, the Best Practice Tariffs incentivise providers to treat patients as day cases, and the incentives are gradually r educed as day surgery becomes the norm, as is the case now for cataract surgery (Table 2.4). These financial incentives have contributed to a steady increase in the share of day surgery for interventions such as inguinal hernia repair, tonsillectomy and laparoscopic cholecystectomy in England since 2009 (see Figure 2.6). Best Practice Tariffs for day surgery in England, Table 2.4. selected interventions, 2017 Inpatient reimbursement (EUR) Surgical procedure Day surgery reimbursement (EUR) 902 902 Cataract surgery 1 581 1 424 Repair of inguinal hernia 1 269 1 146 Tonsillectomy (children) 1 157 1 257 Tonsillectomy (adults) Laparoscopic cholecystectomy 2 002 2 214 Note: A Best Practice Tariff is no longer provided for cataract surgery as nearly all are now day cases. The conversion into euros is based on an exchange rate of GBP 1 = EUR 1.16. Source: National datasets for Payment by Results. France has combined financial incentives and administrative measures, over time aligning inpatient and ambulatory surgery tariffs closer to the costs of the latter. Since 2008, hospitals with relatively low ambulatory surgery rates can be required by the health insurance fund to request prior authorisation for each instance of planned inpatient admission for those surgeries (which can be justified, for example if a patient cannot be accompanied by a responsible adult upon discharge). The initial list of surgical procedures included cataract surgery; laparoscopic cholecystectomy and hernia repair were added later, but tonsillectomy has not yet been added, which in part explains the trends observed in Figure 2.6. Reducing delayed discharge from hospital 6 Delayed discharges unnecessarily increase health care costs In many cases, savings can be gained through better management of length of stay in hospital, which can be reduced through better co-ordination and planning within hospitals, and between hospitals and post-discharge care settings. Unnecessarily delayed discharges can be costly to health systems for several reasons. Patients who are clinically ready to be discharged can occupy beds that could otherwise be used to care for patients with greater needs. A recent cross-country review estimated that the cost of delayed discharge ranges from EUR 230-650 per patient per day (Rojas-García et al., 2018). In the HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 58

61 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS United Kingdom (England), the National Audit Office has estimated the cost of delayed transfers of care for people aged 65 and over to be GBP 820 million per year (~EUR 726 million) (National Audit Office, 2016). Delayed discharges from hospital also contribute to high-cost care through their effects on the health of patients. A longer stay in hospital increases the risk of health care- associated infections, and can accelerate functional decline, particularly among elderly patients (Covinsky et al., 2003; Zisberg et al., 2015). Bed days attributable to delayed transfers of care, 2016 Table 2.5. Bed days/1 000 population Number of bed days Denmark 30 844 5 Ireland 201 977 43 82 411 16 Norway 40 Sweden 393 124 2 254 821 34 United Kingdom (England) Note: Data for the United Kingdom (England) refer to April 2016-March 2017. Bed days per 1 000 population for Denmark was country-reported. Bed days per 1 000 population for all other countries are based on dividing the total number of bed days lost by the 2016 population (UN Population Prospects 2017, medium variant). Suzuki (forthcoming), “Reducing delays in hospital discharge”, OECD Health Working Papers. Source: The extent of delayed discharges differs markedly, from 5 bed days per 1 000 population in Denmark to 43 bed days per 1 000 population in Ireland, also the country with the highest bed occupancy rate (94%).The proportion of bed days occupied by patients with delayed discharge is driven by two related components: the number of patients who experience a delayed discharge, and the length of the additional stay. In the United Kingdom (England), for example, the number of patients who experienced a delay in discharge from hospital increased by 60% between 2011 and 2016, with the total number of excess bed days over 2.25 million in 2016 (NHS England, 2018). Figure 2.7. Bed days associated with delayed transfers of care, England (United Kingdom) Awaiting care facility placement Administrative and financing-related delays Patient or family choice Awaiting home-based services Number of patient days 2500000 2000000 1500000 1000000 500000 0 2012-13 2013-14 2014-15 2015-16 2016-17 2011-12 NHS England (2018). Source: http://dx.doi.org/10.1787/888933834186 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 59

62 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Similarly, the number of patients with a recorded delay in discharge doubled in Norway between 2012 and 2016 (Helsedirektoratet, 2018). It is estimated that patients over 65 make up 85% of those with delayed discharge in England (Department of Health, 2016). With population ageing, the challenge of patients experiencing a delay in discharge in European countries is of growing concern. Approaches to reducing delayed discharge from hospital The reasons behind rising rates of delayed discharge in many European countries are multifactorial, with elements from health and social care systems. Many of the key drivers are factors outside the hospital itself, including capacity shortages in intermediate, home and long-term care, as well as poor transition planning and care co-ordination. Several countries have taken steps to increase the capacity of intermediate care facilities and home care to accommodate people who no longer require acute care. Increasing the availability of intermediate care is used as a strategy to improve hospital and Sweden. Strengthening home-based transitions in the Netherlands, Norway, Scotland, care services, including hospital-at-home and outreach services following discharge, has been found to both reduce length of stay and the risk of hospital readmission (O'Connor et al., 2015). Poor management of hospital transitions and lack of co-ordination between hospitals and community-based services also contribute substantially to delays in discharge (Barker et al., 1985; Shepperd et al., 2013). Hospital discharge planning processes often begin too late in the patient’s hospital stay to ensure effective post-discharge care in time. Policies to improve co-ordination, including better integration of primary care into care co-ordination processes, and incentivising better co-ordination through pay-for-performance and pay-for-co-ordination schemes, can help to ensure patient care is better managed following discharge. Better monitoring of delayed hospital discharges enables countries to develop more finely tailored approaches to reducing them. At least eight European countries currently monitor delayed discharges in some form, of which five have developed financial incentives for reducing them. In Denmark, Norway, Sweden, and the United Kingdom (England), where municipalities play a strong role in delivering social care in the community, financial penalties have been introduced for every additional day a patient spends in hospital after they are clinically ready for discharge. In Denmark, a sharp increase in the daily penalty in January 2017 – from DKK 1976 (~EUR 265) to DKK 3952 (~EUR 530) per day, rising to DKK 5928 (~EUR 795) for the third and all subsequent days of delay – was associated with a decline in the number of delayed discharges reported by hospitals. After hospital discharge was identified through patient surveys as the least satisfying aspect of a hospital stay, Norway began re-organising the discharge process, including starting the discharge planning process at admission, communicating important information to municipalities during the admission, facilitating a discharge discussion with patients and families, and creating a discharge checklist. In addition, hospitals are required to contact municipalities within 24 hours of an admission if they believe the patient will require follow-up from health or social care services once discharged. Addressing wasteful spending on pharmaceuticals After inpatient and outpatient care, pharmaceuticals represent the third largest component of health spending (see Chapter 5). In 2016, on average medicines accounted HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 60

63 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS for 17% of total health expenditure in EU countries (excluding medicines used in hospitals), but more than 40% in Bulgaria, over 30% in Romania, and in excess of 25% in Latvia, Lithuania, Greece, Hungary, Croatia and the Slovak Republic (Figure 2.8). Trends in pharmaceutical expenditure are thus an important influence on overall health expenditure patterns. While a high level of spending does not in itself indicate waste, optimising the value derived from medicines expenditure and identifying and eliminating waste where it occurs are both critical to achieving efficient and sustainable health care systems. Pharmaceutical expenditure (retail) per capita Figure 2.8. and as a share of health expenditure, 2016 Voluntary health insurance and out-of-pocket % health expenditure Government/compulsory schemes Per capita, Euro PPP % health expenditure 45 900 40 800 35 700 30 600 25 500 20 400 15 300 10 200 5 100 0 0 Note: Pharmaceuticals used in hospitals could add another 30% of spending on top of retail spending. Source: and Eurostat Database. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, http://dx.doi.org/10.1787/888933834205 1 2 To achieve these objectives – without reducing benefits for patients or undermining the quality of care – a mix of supply and demand side levers can be considered to: i) ensure value for money in selection and coverage, procurement and pricing; ii) promote off-patent competition and exploit the potential of generics and biosimilars; iii) encourage rational use; and iv) improve adherence (Figure 2.9). These are discussed in turn in the remainder of this section. Ensuring value for money in selection and coverage, procurement and pricing Using health technology assessment (HTA) to inform the selection of covered medicines One approach to avoiding wasteful spending is to ensure that those medicines selected for procurement or reimbursement reflect good value for money. Health technology assessment (HTA) is a comparative, multi-disciplinary process used to evaluate the added benefit or impact of health technologies, and which can be used to inform decision makers’ assessment of the opportunity cost of replacing an existing standard of care with a new therapy. In this way, selection and coverage decisions can avoid displacing high value products with ones of lesser value to the health system. HTA can also be used to review the value for money offered by existing therapies, and to adjust prices to reflect a desired level of cost-effectiveness or willingness to pay. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 61

64 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Figure 2.9. Possible approaches to reducing wasteful spending on pharmaceuticals Ensure value for exploit Fully money in selection, potential coverage, pricing and from savings procurement generics and biosimilars Improve adherence to treatment, Promote ra tional optimise pre s and cribing curb packaging, and overuse reduce waste Many European countries have established, and several more are in the process of institutionalising forms of HTA to inform the selection of medicines for their public programmes. Twenty-three EU Member States have HTA mechanisms that assess medicines; 20 have HTA systems that also assess medical devices, and 17 countries include the assessment of other technologies. While cooperation between EU countries on HTA has been increasing over time, as part of its 2017 work programme, the European Commission (EC) announced an initiative to take this a step further. In January 2018, the EC issued a proposed Regulation on HTA covering new medicines and certain new medical devices, providing a basis for increased cooperation at EU level. Under the regulation, Member States would develop common HTA tools, methodologies and procedures for: 1) joint clinical assessment; 2) joint scientific consultations for developers seeking advice from HTA bodies; 3) identification of emerging health technologies. Member States are currently debating the substance of the proposed regulation, particularly whether (and the extent to which) the cooperation on clinical assessment should be mandatory (European Commission, 2018). Increasing bargaining power Intra- and international cooperation among buyers can increase bargaining power, and can improve both the information and resources available to buyers. Belgium, the Netherlands and Luxembourg established a cooperative initiative in 2015, and were joined by Austria in 2016 and Ireland in 2018. The initiative involves cooperation in informing and developing pricing and reimbursement decisions, including joint HTA, horizon scanning and exchange of information from national disease registries, as well as joint price negotiations HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 62

65 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS with industry (BeNeLuxA, 2017; Department of Health, Ireland, 2018). To date, the focus has been on high-cost and orphan drugs considered priorities in each of the countries, and for which assessment methods are deemed sufficiently similar to allow for such cooperation. Similar cooperation has been announced, but not yet implemented by Bulgaria and Romania in the procurement of high-cost drugs (Novinite.com, 2016); by Poland, Hungary, the Slovak Republic and Lithuania (Visegrad Group, 2017); and by ten Southern European countries that are signatories to the Valletta Declaration (Infarmed, 2018). Promoting off-patent competition and exploiting the potential of generics 7 and biosimilars It is widely recognised that the development of competitive generics markets are an important mechanism for reducing expenditure without compromising benefits to patients (Seeley. E, 2008). The use of a cheaper generic equivalent (or in some cases, a cheaper, therapeutically interchangeable drug from the same therapeutic class) in lieu of an originator medicine can generate significant cost savings. Moreover, the market entry of generics can also enhance patient access, particularly in lower-income countries (Elek et al., 2017). Some countries set single reimbursement amounts for groups of therapeutically equivalent drugs, known as “reference prices”, and these can substantially reduce government or other third-party payer outlays. However, they can also discourage competition and lead to higher prices for off-patent medicines than might be expected through competitive procurement mechanisms such as tendering. Rather than offer discounts to government or other third-party payers, to gain market share manufacturers may set their list prices at the reference price level, but offer discounts or other inducements to wholesalers and/or pharmacies. Where third-party payers then reimburse the full reference price, significant profits accrue to wholesalers and pharmacies without any benefits flowing to consumers or third-party payers (Seiter, 2010). In response, some countries have imposed ceilings on wholesaler and pharmacy margins or introduced profit-sharing arrangements (European Commission, 2012). Evidence also suggests that direct regulation of generics prices, for example, by imposing fixed discounts relative to originator products (or using reference prices) is less effective in reducing prices than where prices are established through competitive mechanisms such as tendering or negotiation (OECD, 2017) However, competition-inducing policy measures should be tailored to respective care settings (outpatient vs inpatient) and take into account issues of long-term supply certainty. Across Europe, prices, market shares and timing of market entry of generic medicines vary widely (Rémuzat et al., 2017; Kanavos, 2014). In 2016, generics accounted for more than 75% of the volume of medicines covered by basic health coverage in Germany and the United Kingdom, but made up less than 30% in Switzerland and Italy, and less than 15% in Luxembourg. A recent study also reported that prices of generics in Switzerland were more than six times higher than in the United Kingdom (Wouters, Kanavos and McKee, 2017). Yet, generic market entry intensity or price decline cannot be entirely explained by the size of a geographical market (Kanavos, 2014). Although some of the observed differences in uptake across countries may reflect differences in the timing of patent expiries, generic uptake depends very much on policies implemented at national level (Belloni, Morgan and Paris, 2016; EvaluatePharma®, 2018). In addition to promoting competitive procurement and pricing, these include encouraging rapid market entry of follow-on products on loss of HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 63

66 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS market exclusivity of originator medicines; promoting or mandating prescribing by 8 international non-proprietary name (INN) ; encouraging and incentivising pharmacists to substitute at the point of dispensing; and incentivising and educating patients. Generic market share by volume and value, 2016 (or latest year) Figure 2.10. Value Volume % 90 80 70 60 50 40 30 20 10 0 Data reflect the total market when available (if not, data reflect the reimbursed market or the community pharmacy market). Note: and Eurostat Database. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: http://dx.doi.org/10.1787/888933834224 1 2 Several European countries employ a range of approaches to promoting generic uptake, while others are yet to establish policy frameworks that fully exploit their potential. Over the past decade, Belgium and France have introduced financial incentives to encourage patients to choose a generic rather than an originator product. Belgium also has prescription quotas for doctors, mandatory substitution for some categories of drugs, education and information campaigns for patients, and fixed fees for pharmacies to avoid any unintended incentives to dispense either originator or generic products. However, even though the generic market share by volume doubled from 17% to 35% between 2005 and 2015 in Belgium, generic use is still low relative to many other EU countries such as the United Kingdom, Germany and the Netherlands (OECD/European Observatory on Health Systems and Policies, 2017). France (in 2009) and Hungary (in 2010) have also introduced incentives for GPs to prescribe generics through pay-for-performance (P4P) schemes. Between 2011 and 2016, the generic market share by volume in France increased from 18% to 28%, but similar to Belgium, it remains well below the EU average, in part because France restricts the categories of drugs for which generic substitution and competition are permitted. In 2015, France implemented mandatory INN prescribing, and the 2017 National Action Plan for the promotion of generics aims to increase the generic market share by a further 5 percentage points by 2018 (CNAMTS, 2018). In Italy, prescribers may indicate either the INN or the brand of a medicine, but unless a reason is provided to preclude substitution (or the patient objects) the pharmacist must dispense the cheapest version of the product. Greece has issued prescribing guidelines; set maximum prices for generics; implemented a compulsory, country-wide electronic prescription system to monitor prescribing and HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 64

67 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS dispensing; and mandated prescribing by INN, generic substitution in pharmacies, and the use of generics in public hospitals (OECD/European Observatory on Health Systems and Policies, 2017). Countries that have achieved strong or rapid improvement in penetration of generics include the Netherlands, Denmark, and Spain. In the Netherlands, competition between generics is encouraged by “preference policy”, whereby insurers only reimburse the cheapest generic (Zuidberg, 2010). Denmark introduced price controls and promotion of generics, and increased generic market share by volume from less than 40% in 2007 to over 60% in 2015 (OECD/European Observatory on Health Systems and Policies, 2017). Spain adopted a series of measures that include accelerating market entry of generics and mandatory pharmacy substitution with the cheapest generic (since 2006); the generic market share by volume increased from 14% in 2005 to 47.5% of the total reimbursed market in 2016 (OECD/European Observatory on Health Systems and Policies, 2017). Biologics represent one of the most rapidly growing segments of the pharmaceutical market, predicted to increase from 25% of global sales (by value) in 2017 to 31% in 2024 (EvaluatePharma®, 2018). Just as generic versions of small-molecule medicines generate opportunities to obtain comparable health benefits at lower prices, so do “follow-on biologics” – known as biosimilars . However, expanding biosimilar uptake presents some additional challenges; the inherent complexity of biological products means that biosimilars can be more challenging to develop and manufacture than small molecule generics, and as they are not identical to their reference products, they may not be suitable for substitution at the point of dispensing–akeydriverofg eneric uptake. Biosimilars have been available in Europe for over a decade, and as of 31 March 2018, more than 40 biosimilar products in 15 different biologic classes were approved for marketing in the EU, with 19 new biosimilars authorised between January 2017 and March 2018 (Aideed, 2018). However, despite Europe accounting for nearly 90% of global biosimilar sales (Brennan, 2018), the overall market penetration of biosimilars remains low. With many major patent expiries anticipated between 2018 and 2024, opportunities for further savings are substantial (IMS Institute for Healthcare Informatics, 2016). Across Europe, significant differences exist in policy approaches to biosimilar pricing and reimbursement, stakeholder incentives for biosimilar use, and levels of education and awareness, with consequent variations in uptake and the extent of savings . (Roediger, Freischem and Reiland, 2017; Rémuzat et al., 2017) A recent study of biosimilar policies in 24 countries (20 EU Member States, plus Iceland, Norway, Russia and Serbia) showed that many biosimilars were not uniformly accessible across Europe, with Germany the only country in which all approved biosimilars were available and funded (Moorkens et al., 2017). In most countries, biosimilar pricing in ambulatory care involves a mix of mechanisms (see Table 2.6), while in the hospital setting, tendering is used in all countries, either at national level or by individual hospitals. In the majority of countries, the reference product and biosimilar may be subject to internal reference pricing to set a common reimbursement level (Moorkens et al., 2017). Demand side measures include incentives for physicians to prescribe biosimilars. For example, France encourages physicians to prescribe at least 20% insulin glargine biosimilars in ambulatory care, while in Belgium biosimilars form part of physicians’ quotas for prescribing low-cost medicines, and they are encouraged to prescribe at least 20% biosimilars for treatment-naïve patients. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 65

68 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Table 2.6. Biosimilar policies across Europe Internal reference Incentives Biosimilar pricing in ambulatory care Country Substitution pricing to prescribe Austria 1st/2nd/3rd biosimilar prices -38%/-15%/-10% discount from Reference Product Yes Yes No (RP). RP must reduce price by 30% three months after 1st biosimilar reimbursement. After 3rd biosimilar, RP must match price of the cheapest available biosimilar. Prices of biosimilars negotiated on a case by case basis, maximum reimbursed price No No Belgium Yes cannot be > RP. RP must reduce price on market entry of biosimilar. No Ex-factory price of biosimilar cannot exceed lowest price in a set of countries Bulgaria Yes No (Bulgaria, Romania, France, Latvia, Greece, Slovak Republic, Lithuania, Portugal, Italy, Slovenia, Spain, Belgium, Czech Republic, Poland, Hungary, Denmark, Finland, or Estonia), referred to as external reference pricing (ERP). Ceiling retail price is determined using 3-levels of regressive margins. Croatia Yes No No Biosimilar price determined via ERP (Italy, Slovenia, Czech Republic, Spain, France). 1st biosimilar: -15% on RP/subsequent biosimilars: -10%. Czech Republic The price and reimbursement of 1st biosimilar -30% of the RP. List price of RP remains No No Yes the same, but reimbursement level is lowered to the price of the biosimilar. The maximum price of the biosimilar is determined via ERP of all EU countries except Bulgaria, Czech Republic, Estonia, Luxembourg, Germany, Austria, Romania, Cyprus and Malta. Free pricing, with volume based pricing scheme (rebates when expenditure exceeds Yes No No England agreed total). However biosimilars predominantly sold to hospitals, which procure them via a nationally coordinated tendering process. The price is negotiated; in ambulatory care the price must be at least 15% < RP. Yes Yes Yes Estonia Finland The price of the biosimilar must be < the price of the RP. The wholesale price of the No No Yes 1st reimbursable biosimilar must be at least 30% < wholesale price of RP. Yes Prices determined by negotiation, but typically 10-20% below the price of RP, taking France Yes Yes into account a range of factors including the price in the rest of Europe. Germany Free pricing. Yes Yes Yes/No No Yes Iceland The price of the biosimilar must not be higher than the lowest wholesale price No in Denmark, Norway, Sweden and Finland. Once a biosimilar is on the market, the price of the RP is reduced to 80% of the original ex-factory price. The price of the biosimilar is negotiated, typically 10-20% below RP. No No No Ireland In general, biosimilars are priced approximately 20% < RP. No Regional No Italy Latvia 1st biosimilar at least -30% on RP; 2nd and 3rd biosimilars at least -10% on 1st/2nd Yes No Yes biosimilars; subsequent biosimilars: -5% further decrease. Price may not be > 1/3 lowest price in Czech Republic, Romania, Slovakia, Hungary and Denmark, and no higher than in Estonia and Lithuania. Malta Maximum price is set for national procurement through ERP. Procurement No No No by centralised tendering (by INN, thus promoting competition). Netherlands The price of a biosimilar is officially the same as the price of the RP. Yes In hospitals No Norway The price of the biosimilar cannot be higher than the price of the RP. No Yes No Yes No Poland 1st biosimilar: -25% on RP; 2nd biosimilar must be < 1st, “limit groups” exist Yes where the cheapest is the limit for the whole group. Portugal ERP, with annual changes in reference countries (2017: Spain, France and Italy), No In hospitals Yes/No to establish maximum price. For reimbursement biosimilar must be < 80% of RP or < 70% of RP when biosimilar market share is  5% for the INN. Serbia No 1st biosimilar: -30% on RP, sets the reimbursement rate. 2nd biosimilar: -10% on No Yes 1st biosimilar. 3rd biosimilar: -10% on 2nd biosimilar, with maximum 90% of average price in Slovenia, Croatia, and Italy. National tendering by brand name can occur. No Slovenia Biosimilar price is either 92% of the lowest price in Austria, Germany and France, No Yes or 92% of median price in other EU/EEA countries. If the biosimilar is not in any of the reference countries or EU/EEA countries, price is 68% of RP. The price of the biosimilar is negotiated, typically 20-30% below the price of the RP. Yes In some regions No Spain A maximum price is set for national procurement. The price of the biosimilar must be same or lower than that of RP. Sweden No Regionally No ERP: External Reference Pricing, EU: European Union, EEA: European Economic Area, RP: Reference Product. * In Italy, biosimilars are considered interchangeable with their RPs, but substitution is only at the discretion of the prescriber. See www.agenziafarmaco.gov.it/sites/default/files/2_Position-Paper-AIFA-Farmaci-Biosimilari.pdf . ** In Portugal, substitution is encouraged for infliximab, rituximab and etanercept if the biosimilar is cheaper and the patient stable, but is not mandatory. Adapted from Moorkens et al. (2017). Source: HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 66

69 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Box 2.1. Current and potential future savings from the use of biosimilars In 2016, it was estimated that biosimilars could generate savings up to EUR 100 billion by 2020 in the five most populous countries in the European Union (Germany, France, Italy, Spain and the United Kingdom) plus the United States. Although thus far price reductions offered by biosimilars have not been nearly as large as those seen with small molecule generics, discounts of over 60% have been reported for selected products (see graphs). Uptake of biosimilars also varies substantially across Europe. That said, the correlation between biosimilar market shares and price reductions is weak, suggesting the existence of barriers to effective competition. Promoting biosimilar uptake is important for driving savings and ensuring the continued participation of players in the market, but it is the market entry of biosimilars that promotes price competition. The two graphs below show a) the market penetration of biosimilars as a proportion of all products within the same drug class eligible for biosimilar competition (vertical bars, left axis) and b) the price evolution across all products within the class eligible for biosimilar competition (diamonds, right axis).The first graph shows the results for the class of drugs known as erythropoietins, used in the acute care setting to stimulate red blood cell production in a number of conditions, including chronic renal failure. Erythropoietins were among the first biosimilar products to be approved in Europe.The second graph shows similar metrics for anti-TNF alfas, a class of drugs used for a range of chronic conditions such as rheumatoid arthritis and Crohn’s disease, and for which biosimilars have entered the market more recently. Figure 2.11. Market share of biosimilars and price evolution - see Note ) left axis ( Market share ( Price evolution ) - see Note ht axis g ri Erythropoietins, 2015 100% 0% 80% -20% 60% -40% 40% -60% 20% -80% 0% -100% Anti-TNF alfas, 2016 100% 0% 80% -20% 60% -40% 40% -60% 20% -80% 0% -100% Graphs show market share of biosimilars for year shown: a) biosimilar treatment days (TD) as a Note: proportion of TD of all products in the drug class eligible for biosimilar competition ( vertical bars, left axis ) and b) price evolution (change in price per TD for year shown across all products in the drug class eligible for biosimilar competition, relative to price per TD in the year prior to biosimilar market entry [ right axis ]). Source: IMS Institute for Healthcare Informatics (2016); Quintiles IMS (2017). 1 2 http://dx.doi.org/10.1787/888933834243 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 67

70 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Portugal has recently implemented financial incentives for pharmacies to encourage dispensing of lower price medicines, and defined target market shares for biosimilar versions of infliximab, etanercept and rituximab. In the Netherlands, limitations on the prescribing of reference products are often part of agreements reached between insurance companies and hospitals, though budget constraints within hospitals already provide incentives for the use of biosimilars. Substitution rules are also important in influencing biosimilar uptake. With the exception of Estonia, France, Latvia, and Poland, most countries do not permit unrestricted substitution of biologicals at the point of dispensing. In France, draft legislation permitting substitution of biosimilars was introduced in 2017 but is limited to initiating treatment in treatment-naïve patients, or to ensuring continuity for patients previously dispensed a biosimilar (ibid.). Encouraging rational use Efforts to minimise waste in expenditure on medicines can be undermined significantly by over-prescribing and inappropriate use. Over-prescribing not only wastes resources, it increases the risks of therapeutic failure, adverse events, and the development of antimicrobial resistance (AMR).This section focuses on two specific groups of medicines that are frequently subject to over-prescription, and have particular implications for public health: antibiotics and hypnotics/anxiolytics (mainly benzodiazepines). Antimicrobial resistance represents an increasingly serious social and economic burden globally, projected to be responsible for as many as 33 000 deaths per year in the EU alone between 2015 and 2050, if no effective action is put in place (OECD, 2018). In addition, overprescribing of antibiotics incurs a number of other direct and indirect costs, by medicalising conditions for which antibiotics are not useful, and by putting patients at risk of adverse effects (and the costs of treating them). Primary care accounts for 80-90% of all antibiotic prescriptions in Europe, with most prescribed for respiratory tract infections (van der Velden et al., 2013). However, rates of antibiotic prescribing differ significantly across Europe, despite little evidence of differences in the prevalence of infectious diseases (Llor and Bjerrum, 2014). In 2016 the population- weighted average consumption of antibiotics for systemic use in the community was 9 22 defined daily doses (DDD) per 1 000 population per day, and ranged from 10 DDD (the Netherlands) to 36 DDD per 1 000 population per day (Greece), a 3.5-fold difference (Figure 2.12). Prescribing influences have been shown to be multifactorial and include cultural and socioeconomic elements, diagnostic uncertainty, the way health care is funded or reimbursed, the percentage of generic drugs in the market, economic incentives and pharmaceutical industry influences, attitudes and beliefs about the therapeutic value of antibiotics among patients, as well as differences in prescriber and patient expectations of consultations for respiratory tract infections (Llor and Bjerrum, 2014). A 2014 survey of over 1 000 GPs in the United Kingdom reported that 55% felt under pressure, mainly from patients, to prescribe antibiotics, and 44% admitted to prescribing antibiotics to get a patient to leave the surgery (Cole, 2014). There is a clear need to improve health literacy, in particular to raise awareness about antibiotic use and resistance among European populations, while the increasing prevalence of antibiotic-resistant bacteria could be addressed, at least in part, by promoting more limited and appropriate antibiotic use in primary care and in the community (European Centre for Disease Prevention and Control, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 68

71 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Consumption of antibiotics in the community, EU/EEA countries, Figure 2.12. 2016 (DDDs per 1 000 population per day) 21 16 15 12 12 13 16 17 DDD per 1 000 pop. per day 24 10.45 - 14.95 20 24 10 14 14.96 - 20.28 28 20.29 - 25.60 19 25 25.61 - 30.93 24 30.94 - 36.26 13 15 n.a. 30 14 30 21 n.a. n.a. 27 20 n.a. n.a. n.a. 22 36 22 n.a. 33 16 (Malta) These data are mainly drawn from sales of antibiotics in the country, or a combination of sales and reimbursement data. Cyprus and Note: Romania provide data on overall consumption (including the hospital sector). Spain provides data only on reimbursed antibiotics (i.e. not including consumption without prescription or not reimbursed). European Centre for Disease Prevention and Control (ECDC) (2017). Source: 2014). Findings from a recent OECD publication investigating the effectiveness and cost effectiveness of public health policies to promote prudent use of antimicrobials support an upscaling of national actions in this direction (OECD, 2018). In addition to differences in antimicrobial use, patterns of resistance, and the extent to which effective national policies to deal with AMR have been implemented vary within the EU. In June 2017, the European Commission adopted the EU One Health Action Plan against AMR to i) make the EU a best practice region; ii) boost research, development and innovation; and iii) shape the global agenda on AMR (European Commission, 2017). The European Commission has also published guidelines for the prudent use of antimicrobials in human health (European Commission, 2017). Levels of prescribing of hypnotics and anxiolytics, especially among the elderly, are another important public health issue. Benzodiazepines (BZDs) and related drugs are frequently prescribed for older adults for anxiety and sleep disorders, despite HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 69

72 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Box 2.2. Approaches to reducing AMR in Belgium AMR has been recognised as an important public health issue in Belgium for several years. The Belgian Antibiotic Policy Coordination Committee, established in 1999, is responsible for fostering more appropriate use of antibiotics in humans and animals and for promoting infection control and hospital hygiene, with the overall aim of reducing AMR. Recent measures to reduce antibiotic consumption have targeted patients (e.g. through public awareness campaigns and increased co-payments for some antibiotics) and prescribers (e.g. through organised feedback), and have contributed to a reduction in hospital-acquired antibiotic-resistant staphylococcus infections. Although Belgium performs relatively well in terms of levels of resistance, it now faces challenges in preventing and controlling infections by carbapenem resistant isolates (CREs). OECD/European Observatory on Health Systems and Policies (2017), European Centre for Disease Source: Prevention and Control (2018). well-documented risks of adverse effects including fatigue, dizziness and confusion. Long- term use of BZDs can also lead to falls, accidents and overdose, as well as tolerance, dose escalation and dependence, long-term cognitive impairment and pseudo-dementia (Ford and Law, 2014). Apart from the associated mortality and morbidity, these impose substantial additional and potentially avoidable costs on health systems. In addition to issues arising from prolonged use, there is also concern about the types of BZDs being prescribed in the older age groups, with long-acting products not recommended in older adults (OECD, 2017). While data are available for only a few countries (F wide variations in prescribing igure 2.13), 10 long-term BZD prescribing in the over 65s highest in rates are apparent, with the rate of Ireland, and nearly 13 times that of Estonia. Conversely, prescribing of BZDs in the long-acting over 65s was highest in Estonia, with a rate more than 17 times that of Finland. Figure 2.13. Elderly patients with prescriptions for benzodiazepines or related drugs, number per 1 000 patients aged 65 and over, 2015 or nearest year Elderly patients with prescriptions for long-acting benzodiazepines or related drugs Elderly patients with long-term prescriptions for benzodiazepines or related drugs Per 1 000 persons aged 65 years and over 160 150 140 113 120 102 100 85 82 80 67 57 54 60 44 36 34 31 40 27 20 18 18 18 20 9 5 n.a. 0 Finland Estonia Slovenia Spain Portugal Ireland Netherlands Denmark Sweden Norway OECD Health Statistics 2018, . Source: https://doi.org/10.1787/health-data-en http://dx.doi.org/10.1787/888933834262 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 70

73 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Improving adherence and other avenues for reducing waste Apart from contributing to an estimated 200 000 premature deaths, poor adherence to prescribed medication is thought to cost as much as EUR 125 billion in Europe each year in avoidable hospitalisations, emergency care, and adult outpatient visits (OECD, 2017). Three prevalent chronic conditions – diabetes, hypertension, and hyperlipidaemia – appear to give rise to the highest avoidable costs. Among patients with these three conditions, it has been estimated that between 4 and 31% do not fill their first prescription; of those who fill their first prescription only 50 to 70% take their medications regularly (i.e. at least 80% of the time); and more than half discontinue taking them within two years (Khan and Socha-Dietrich, 2018). Modelled over a 10-year period in five European countries (Italy, Germany, France, Spain, and England), the potential savings from increasing adherence to antihypertensive treatment to 70% have been estimated at EUR 332 million (Mennini et al., 2015). Research undertaken in the United Kingdom also identified potential savings of over GBP 100 million (EUR 111 million) annually if 80% of patients with hypertension were adherent to treatment (Trueman et al., 2010). The same report estimated the annual cost of medicine wastage in primary care to be as high as GBP 300 million (EUR 333 million), of which GBP 100-150 million (EUR 111-166 million) was identified as avoidable. However the authors also found that while patient non-adherence contributes to wastage, a range of other factors are also implicated, some of which are unavoidable, such as treatment changes due to lack of efficacy or the emergence of adverse effects. Those that can be addressed included inappropriate repeat prescribing and dispensing processes, which, independently of any patient action, may cause excessive volumes of medicines to be supplied (Trueman et al., 2010). A study examining waste samples in Vienna in 2015-16 found significant quantities of prescription medicines discarded in household garbage. By extrapolation the authors estimated the value of the discarded medicines to correspond to approximately 6% of public pharmaceutical expenditure nationally in the year of survey, or at least EUR 21 per person to Austrian social health insurance (Vogler and de Rooij, 2018). Box 2.3. Reducing waste in the United Kingdom The National Health Service’s MedicineWaste campaign provides information about common reasons for discarding medicines, describes simple steps for patients to follow to enhance adherence, and proposes a short checklist for clinicians to evaluate repeat prescriptions (NHS Business Services Authority, 2015). In addition, across the UK, pharmacists work alongside GPs to improve outcomes by undertaking patient-facing clinical medication reviews, and improving the management of long-term conditions (Mann et al., 2018). In September 2017, the Department of Health & Social Care established a Short Life Working Group (SLWG) to provide advice on a programme of work to improve medication safety. Recommendations of the SLWG included the rollout of primary care interventions such as PINCER (pharmacist-led information technology intervention) which have been shown to be effective in reducing a range of medication errors in general practices with computerised clinical records. Other efficiency initiatives introduced in the United Kingdom in recent years include the Hospitals Pharmacy & Medicines Optimisation (HoPMOp) project, which helps NHS acute hospital trusts to implement the recommendations of the review of NHS productivity and efficiency by Lord Carter of Coles, (GIRFT) project, which aims to reduce unwarranted and the Getting It Right First Time variation in clinical practice across the NHS. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 71

74 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS In hospitals, medicines may be discarded because of inappropriate pack sizes, often the case with drugs requiring weight-based dosing (common in oncology), or that are supplied in single-dose units that must either be administered or discarded once opened (OECD, 2017). The latter issue requires an audit of the extent to which regulation – or a lack of it – contributes to unnecessary waste. For instance, regulatory agencies could require manufacturers to provide drugs in a variety of pack sizes to ensure that an amount of drug more closely corresponding to a patient’s body weight or size can be drawn up without waste, and could develop or revise existing guidelines on vial sharing. Alternatively, payers could determine reimbursement amounts that correspond to the actual dose administered (i.e. no reimbursement for leftover drug) (ibid.). Policies aimed at tackling poor adherence and unnecessary waste of medicines by patients are aimed at encouraging improved communication between clinicians and patients and enhancing patient understanding of the importance of completing prescribed courses of treatment. Clinical trials conducted in the United Kingdom and Sweden suggest that wastage could be reduced by up to 30% if patients starting new courses of treatment were offered additional opportunities to discuss medication-related issues over and above the initial instructions given at the time of prescribing (OECD, 2017). Targeted medication reviews can be used to monitor patients’ consumption of medication and establish the need for (or lack of) prescription renewal (Trueman et al., 2010) (Box 2.3). Conclusions Progress in reducing wasteful spending in health is not only a barometer of quality improvement; it is both an ethical and financial imperative in the pursuit of resilient and equitable health care systems. While the estimate that as much as one-fifth of health spending could be eliminated is sobering, the many avenues for saving money and streamlining services, without undermining access or quality of care, are cause for optimism. Pointing the lens at two major areas of expenditure – hospitals and medicines – reveals a range of options for improving efficiency and reducing waste, but significant variation across Europe in the extent to which these options are being deployed. For hospitals, reducing or eliminating unnecessary investigations and procedures, many of which expose patients to unnecessary risks without the prospect of clinical benefit, is an obvious target for direct intervention. Expanding the use of day surgery can also be instigated at hospital level. However minimising avoidable admissions, particularly for ambulatory care-sensitive conditions, reducing unnecessary length of stay, and improving discharge processes require broader perspectives. Enhanced primary care services, expanded post- acute care facilities, post-discharge care coordination, and in-home care services all require health system reforms that cannot be initiated by hospitals alone. For pharmaceuticals, creating and supporting competitive markets and promoting the uptake of generics and biosimilars can generate substantial savings. That said, reducing waste does not necessarily mean spending less; it may equally be achieved by gaining better value for money from existing expenditure. Both supply and demand side levers offer scope for better value. Using health technology assessment to inform selection, pricing and procurement of new medicines facilitates an understanding of the true opportunity costs of therapies and helps avoid the displacement of high value interventions with ones of lesser value. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 72

75 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS In all approaches to reducing waste, stakeholder engagement and effective communication are critical. Prescribers and patients need to understand the value offered by generics and biosimilars, and be adequately reassured as to their equivalence and safety. Both need to appreciate the risks of overprescribing antibiotics and the circumstances in which they are of low or no benefit. In hospitals, patients and providers need to recognise that not only will certain investigations and procedures provide no benefit, they may even be harmful. Financial incentives for patients and providers must also be calibrated to reinforce appropriate behaviours. Above all, the development and promulgation of guidelines and protocols that provide both a basis for discussion and engagement and support for rational clinical decision-making, are critical to the waste-reducing armamentarium. Notes 1. This analysis captures only five of thirty conditions for which hospitalisations may be avoidable through better primary care, and is therefore conservative.That said, not all hospitalisations related to these five conditions would be avoidable. Some analysts argue that only admissions involving a short stay in hospital – as a proxy for severity – should be counted (Swerissen, Duckett and Wright, 2016). 2. Analyses which group women according to obstetric criteria (for instance number of foetuses, presentation of foetus, previous C-section) provide finer analyses of the drivers behind these trends and differences in C-sections rates (Betrán et al., 2014). 3. A campaign, established in 2012 by the American Board of Internal Medicine and since emulated in a growing number of countries, has sought to promote a dialogue around appropriate care. One of its core strategies has been to encourage medical societies to draw up shortlists of services known to be used inappropriately, and issue “do-not-do” recommendations to guide providers and patients in reducing their utilisation. 4. Day surgery is defined as the release of a patient who was admitted to a hospital for a planned surgical procedure and discharged the same day. The analysis covers 22 EU countries only due to data gaps in the other six: Greece and Latvia do not report data on day surgery; Cyprus only reports data for public hospitals (which account for less than half of hospital activities); and Bulgaria, the Czech Republic and the Slovak Republic only report data for one or two of the procedures considered here. The main limitation in data comparability is that many countries do not include outpatient surgery, defined as situations where patients are not formally admitted to or discharged from hospitals (see the indicator “Day Surgery” in Chapter 8 for more information). 5. Tonsillectomy is mainly performed in children. Inguinal hernia repair is a procedure to repair a defect in the abdominal wall that allows abdominal contents to slip into a narrow tube called the inguinal canal and is commonly performed laparoscopically (using minimally invasive keyhole surgery, allowing patients to return home more quickly). Cholecystectomy is the removal of the gallbladder, also commonly performed laparoscopically. 6. Delayed discharges from hospital are defined here as cases in which a hospital patient remains in hospital, despite being clinically ready to be discharged. 7. A generic medicine is defined as a pharmaceutical product with the same qualitative and quantitative composition in active substances, and the same pharmaceutical form as the reference product, and to which bioequivalence has been demonstrated. A biosimilar is a biological medicinal product that contains a “follow-on” version of an already-authorised biological reference product and has no clinically meaningful differences in terms of safety and effectiveness from the reference product. However, although biosimilars are conceptually similar to generic versions of chemically derived small molecule medicines, because of the complexity and inherent heterogeneity of biotechnological products, and of the manufacturing processes used to produce them, a follow-on biologic is referred to as “biosimilar” rather than “biogeneric”. 8. International Non-proprietary Names (INN) are unique and globally recognised names used to identify pharmaceutical substances. All pharmaceutical products are assigned an INN; most will also carry a brand or trade name which, unlike the INN, may differ between countries. 9. The DDD is the assumed average maintenance dose per day for a drug used for its main indication in adults. The DDD is a unit of measurement and does not necessarily reflect the recommended or prescribed daily dose. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 73

76 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS 10. “ refers to prolonged duration of use; “long-acting” refers to a drug that has slow absorption Long-term” and maintains its effects over an extended period. References http:// Aideed, H. (2018), “The European Biosimilars Landscape: What To Expect In The Year Ahead”, investors.abbvie.com/phoenix.zhtml?c=251551&p=irol . Ayres-De-Campos, D. et al. (2015), “Lowered national cesarean section rates after a concerted action”, , Vol. 94, pp. 391-398, http://dx.doi.org/10.1111/aogs.12582 . Acta Obstetricia et Gynecologica Scandinavica BADS Directory of procedures , https://daysurgeryuk.net/en/shop/publications/bads-directory-of- BADS (2016), . procedures-5th-edition/ Barker, W. et al. (1985), “Geriatric consultation teams in acute hospitals: Impact on backup of elderly patients”, , http://dx.doi.org/10.1111/j.1532-5415.1985. Journal of the American Geriatrics Society tb07153.x . Belloni, A., D. Morgan and V. Paris (2016), “Pharmaceutical expenditure and policies: Past trends and future challenges” , OECD Health Working Papers , No. 87, OECD Publishing, Paris, http://dx.doi.org/ 10.1787/5jm0q1f4cdq7-en . BeNeLuxA (2017), The BeNeLuxA Initiative – Collaboration on pharmaceutical policy – Terms of Reference , http:// beneluxa.org/sites/beneluxa.org/files/2017-07/BeneluxA_Terms_of_References_final_0.pdf . Betrán, A. et al. (2014), A systematic review of the Robson classification for caesarean section: What works, doesn’t work and how to improve it , http://dx.doi.org/10.1371/journal.pone.0097769 . Brennan, Z. (2018), , www.raps.org/ Unpacking an NEJM Perspective on the Lackluster Biosimilar Uptake in the US news-and-articles/news-articles/2018/3/unpacking-an-nejm-perspective-on-the-lackluster-bi . Brownlee, S. et al. (2017), “Evidence for overuse of medical services around the world”, Lancet , Vol. 390/ 10090, pp. 156-168, http://dx.doi.org/10.1016/S0140-6736(16)32585-5 . Chalmers, K. et al. (2018), “Developing indicators for measuring low-value care: Mapping Choosing http://dx.doi.org/10.1186/s13104- , Wisely recommendations to hospital data”, BMC Research Notes 018-3270-4 . Améliorer la qualité du système de santé et maîtriser les dépenses Propositions de l’Assurance CNAMTS (2018), , www.ameli.fr/fileadmin/user_upload/documents/180702_CP19_rapport_vdef.pdf . Maladie pour 2019 Cole, A. (2014), “GPs feel pressurised to prescribe unnecessary antibiotics, survey finds”, BMJ (Clinical research ed.) , Vol. 349, pp. g5238, http://dx.doi.org/10.1136/BMJ.G5238 . Covinsky, K. et al. (2003), “Loss of independence in activities of daily living in older adults hospitalized with medical illnesses: Increased vulnerability with age”, Journal of the American Geriatrics Society , http://dx.doi.org/10.1046/j.1532-5415.2003.51152.x . Department of Health (2016), Discharging older patients from hospital: Report by the Comptroller and Auditor General , National Audit Office, London. Department of Health, Ireland (2018), Ireland to open negotiations with Belgium, the Netherlands, Luxembourg and Austria on drug pricing and supply – Minister Harris/Department of Health , https:// health.gov.ie/blog/press-release/ireland-to-open-negotiations-with-belgium-the-netherlands-luxembourg- and-austria-on-drug-pricing-and-supply-minister-harris/ . Elek, P. et al. (2017), “Policy objective of generic medicines from the investment perspective: The case of clopidogrel”, , Vol. 121/5, pp. 558-565, http://dx.doi.org/10.1016/J.HEALTHPOL.2017.02.015 . Health Policy Ellen, M. et al. (2018), “Addressing overuse of health services in health systems: A critical interpretive Health Research Policy and Systems , Vol. 16/1, pp. 48, http://dx.doi.org/10.1186/s12961-018-0325-x . synthesis”, Elshaug, A. et al. (2017), “Levers for addressing medical underuse and overuse: Achieving high-value health care”, The Lancet , Vol. 390/10090, pp. 191-202, http://dx.doi.org/10.1016/S0140-6736(16)32586-7 . European Centre for Disease Prevention and Control (2018), ECDC country visit to Belgium to discuss antimicrobial resistance issues , http://dx.doi.org/10.2900/065952 . European Centre for Disease Prevention and Control (2017), Summary of the latest data on antibiotic consumption in the European Union ESAC-Net surveillance data November 2017 , https://ecdc.europa.eu/en/ . publications-data/summary-latest-data-antibiotic-consumption-eu-2017 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 74

77 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS European Centre for Disease Prevention and Control (2014), , Key messages for primary care prescribers . https://antibiotic.ecdc.europa.eu/en/get-informed/key-messages/key-messages-primary-care-prescribers Press release – Assessing health technology in the EU: Commission proposes to European Commission (2018), reinforce cooperation amongst Member States , http://europa.eu/rapid/press-release_IP-18-486_en.htm . European Commission (2017), A European One Health Action Plan against Antimicrobial Resistance (AMR) , www.who.int/entity/drugresistance/documents/surveillancereport/en/index.html . CONTENTS EU Guidelines for the prudent use of antimicrobials in human health European Commission (2017), http:// , . dx.doi.org/10.1111/j.1469-0691.2011.03570.x European Commission (2012), Executive Agency for Health and Consumers-EAHC-European Commission EU Pharmaceutical expenditure forecast Final report 26 November 2012 . EvaluatePharma® (2018), “World Preview 2018, Outlook to 2024. 11th Edition”, www.evaluate.com/ PharmaWorldPreview2017 . FNEHAD (2017), Ensemble Construisons l’Hospitalisation de Demain – Rapport d’activité 2016-2017 . 3° Rapporto sulla sostenibilità del Servizio Sanitario Nazionale www.rapportogimbe.it/ Fondazione GIMBE (2018), , . 3_Rapporto_GIMBE.pdf Ford, C. and F. Law (2014), Guidance for the use and reduction of misuse of benzodiazepines and other hypnotics and anxiolytics in general practice , www.emcdda.europa.eu/attachements.cfm/att_248926_EN_UK59_benzos.pdf . Haute Autorité de Santé (2016), Césariennes programmées à terme: Évolution des taux de césariennes à terme en France entre 2011 et 2014 et évaluation de l’impact du programme d’amélioration des pratiques . Helsedirektoratet (2018), Norwegian Health Directorate, Number of patients ready for discharge, https:// statistikk.helsedirektoratet.no/bi/Dashboard/7051023e-6dff-42bf-a8ea-28c689d117f8?e=false&vo= viewonly . IMS Institute for Healthcare Informatics (2016), Delivering on the Potential of Biosimilar Medicines The Role of Functioning Competitive Markets , www.medicinesforeurope.com/wp-content/uploads/2016/03/IMS-Institute- Biosimilar-Report-March-2016-FINAL.pdf . , Press release: The La Valletta Technical Committee met for the fourth time in Lisbon Infarmed (2018), www.infarmed.pt/documents/281/2183009/Press+release+-+La+Valletta+Declaration+Committee+meeting+ in+Lisbon/10694f9c-211c-4315-81d6-fafbb0abfeeb . Kanavos, P. (2014), “Measuring performance in off-patent drug markets: A methodological framework and empirical evidence from twelve EU Member States”, Health Policy , http://dx.doi.org/10.1016/ j.healthpol.2014.08.005 . Khan, R. and K. Socha-Dietrich (2018), “Investing in medication adherence improves health outcomes and health system efficiency: Adherence to medicines for diabetes, hypertension, and hyperlipidaemia” , OECD Health Working Papers , No. 105, OECD Publishing, Paris, http://dx.doi.org/10.1787/8178962c-en . Le Bail, M. and Z. Or (2016), Atlas des variations de pratiques médicales – Recours à dix interventions chirurgicales , IRDES. Leroy, R. et al. (2017), Proposals for a further expansion of day surgery in Belgium , KCE, https://kce.fgov.be/ sites/default/files/atoms/files/KCE_282_Day_surgery_Report_0_2.pdf . Llor, C. and L. Bjerrum (2014), “Antimicrobial resistance: Risk associated with antibiotic overuse and initiatives to reduce the problem”, Therapeutic Advances in Drug Safety , Vol. 5/6, pp. 229-241, http:// dx.doi.org/10.1177/2042098614554919 . Mafi, J. and M. Parchman (2018), “Low-value care: An intractable global problem with no quick fix”, BMJ quality & safety http://dx.doi.org/10.1136/bmjqs-2017-007477 . , Vol. 27/5, pp. 333-336, al. (2018), Savings from reducing low-value general surgical interventions Malik, H. et http://dx.doi.org/ , 10.1002/bjs.10719 . Mann, C. et al. (2018), Clinical Pharmacists in General Practice: Pilot scheme Independent Evaluation Report: Full Report National Evaluation of Clinical Pharmacists in GP Practices (Pilot Phase)/1 , www.nottingham.ac. uk/pharmacy/documents/generalpracticeyearfwdrev/clinical-pharmacists-in-general-practice-pilot-scheme- . full-report.pdf Mennini, F. et al. (2015), “Cost of poor adherence to anti-hypertensive therapy in five European countries”, Eur J Health Econ , Vol. 16, pp. 65-72, http://dx.doi.org/10.1007/s10198-013-0554-4 . Miller, G. et al. (2018), “A Framework for Measuring Low-Value Care”, Value in Health , http://dx.doi.org/ 10.1016/j.jval.2017.10.017 . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 75

78 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Moorkens, E. et al. (2017), “Policies for biosimilar uptake in Europe: An overview”, , Vol. 12/12, PLOS ONE http://dx.doi.org/10.1371/journal.pone.0190147 . pp. e0190147, , www.england.nhs.uk/ NHS Business Services Authority (2015), Pharmaceutical waste reduction in the NHS wp-content/uploads/2015/06/pharmaceutical-waste-reduction.pdf . NHS England (2018), Statistical work areas: Delayed transfers of care , www.england.nhs.uk/statistics/ . statistical-work-areas/delayed-transfers-of-care Bulgaria, Romania Sign Agreement on Negotiating Expensive Medicines Discounts Novinite.com (2016), , www.novinite.com/articles/177343/Bulgaria%2C+Romania+Sign+Agreement+on+Negotiating+Expensive+ Medicines+Discounts . O’Connor, M. et al. (2015), “The impact of home health length of stay and number of skilled nursing Research visits on hospitalization among medicare-reimbursed skilled home health beneficiaries”, in Nursing & Health , http://dx.doi.org/10.1002/nur.21665 . OECD (2018), Stemming The Superbug Tide: Just A Few Dollars More , OECD Publishing, Paris. OECD (2017), Tackling Wasteful Spending on Health http://dx.doi.org/10.1787/ , OECD Publishing, Paris, . 9789264266414-en Geographic variations in health care: What do we know and what can be done to improve health OECD (2014), , https://doi.org/10.1787/2074319x system performance? . OECD/European Observatory on Health Systems and Policies (2017), Belgium: Country Health Profile 2017 , State of Health in the EU, OECD Publishing, Paris/European Observatory on Health Systems and Policies, Brussels, http://dx.doi.org/10.1787/9789264283299-en . OECD/European Observatory on Health Systems and Policies (2017), Denmark: Country Health Profile 2017 , State of Health in the EU, OECD Publishing, Paris/European Observatory on Health Systems and Policies, Brussels, http://dx.doi.org/10.1787/9789264283343-en . OECD/European Observatory on Health Systems and Policies (2017), Greece: Country Health Profile 2017 , State of Health in the EU, OECD Publishing, Paris/European Observatory on Health Systems and Policies, Brussels, http://dx.doi.org/10.1787/9789264283404-en . , OECD/European Observatory on Health Systems and Policies (2017), Spain: Country Health Profile 2017 State of Health in the EU, OECD Publishing, Paris/European Observatory on Health Systems and Policies, Brussels, . http://dx.doi.org/10.1787/9789264283565-en Purdy, S. et al. (2009), “Ambulatory care sensitive conditions: Terminology and disease coding need to be more specific to aid policy makers and clinicians”, Public Health http://dx.doi.org/10.1016/ , j.puhe.2008.11.001 . Rémuzat, C. et al. (2017), “Key drivers for market penetration of biosimilars in Europe”, Journal of market access & health policy , Vol. 5/1, pp. 1272308, http://dx.doi.org/10.1080/20016689.2016.1272308 . Roediger, A., B. Freischem and J. Reiland (2017), “What pricing and reimbursement policies to use for off- Generics patent biologicals in Europe? – results from the second EBE biological medicines policy survey”, and Biosimilars Initiative Journal , Vol. 6/2, pp. 61-78, http://dx.doi.org/10.5639/gabij.2017. 0602.014 . Rojas-García, A. et al. (2018), Impact and experiences of delayed discharge: A mixed-studies systematic review , http://dx.doi.org/10.1111/hex.12619 . Seeley, E.K. (2008), “Generic medicines from a societal perspective: Savings for health care systems?”, , Vol. 14, No. 2, Eurohealth http://apps.who.int/medicinedocs/documents/s20970en/s20970en.pdf . Seiter, A. (2010), A Practical Approach to Pharmaceutical Policy www.openknowledge.worldbank.or g/ , . bitstream/handle/10986/2468/552030PUB0Phar10Box349442B01PUBLIC1.pdf?sequence=4&isAllowed=y Shepperd, S. et al. (2013), “Discharge planning from hospital to home”, The Cochrane database of systematic reviews , http://dx.doi.org/10.1002/14651858.CD000313.pub4 . Sonola, L. et al. (2013), South Devon and Torbay: Proactive case management using the community virtual ward and the Devon Predictive Model , The King’s Fund, www.kingsfund.org.uk . Stordeur, S. et al. (2016), Elective caesarean section in low-risk women at term: Consequences for mother and offspring – Synthesis ,KCE, https://kce.fgov.be/sites/default/files/atoms/files/KCE_275Cs_Elective_ caesarean_section_synthesis.pdf . Suzuki, E. (forthcoming), “Reducing delays in hospital discharges”, OECD Health Working Papers , OECD Publishing, Paris. Chronic failure in primary medical care , Grattan Institute. Swerissen, H., S. Duckett and J. Wright (2016), HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 76

79 I.2. STRATEGIES TO REDUCE WASTEFUL SPENDING: TURNING THE LENS TO HOSPITALS AND PHARMACEUTICALS Tiainen, A. and B. Lindelius (2016), , www.lakartidningen.se/Opinion/Debatt/ Fortsätt utveckla dagkirurgin . 2016/01/Stor-utvecklingspotential-med-dagkirurgi/#comments Trueman, P. et al. (2010), “Evaluation of the Scale, Causes and Costs of Waste Medicines”, http:// discovery.ucl.ac.uk/1350234/1/Evaluation_of_NHS_Medicines_Waste__web_publication_version.pdf . van der Velden, A. et al. (2013), “Prescriber and Patient Responsibilities in Treatment of Acute Respiratory Tract Infections – Essential for Conservation of Antibiotics”, Antibiotics ,Vol.2/2, pp. 316-327, http://dx.doi.org/10.3390/antibiotics2020316 . Vanlerenberghe, J. (2017), , Améliorer la pertinence des soins : un enjeu majeur pour notre système de santé Sénat, www.senat.fr/rap/r16-668/r16-6681.pdf . Visegrad Group (2017), “The Visegrad Group: The Czech Republic, Hungary, Poland and Slovakia/VISEGRAD BULLETIN 5 (2/2017)”, www.visegradgroup.eu/visegrad-bulletin-5-2 . Visser, S. et al. (2012), Kwaliteit als medicijn, aanpak voor betere zorg en lagere kosten , Booz & Company. Vogler, S. and R. de Rooij (2018), “Medication wasted – Contents and costs of medicines ending up in http://dx.doi.org/10.1016/ household garbage”, , Research in Social and Administrative Pharmacy j.sapharm.2018.02.002 . Vrangbaek, K. (2017), “The Danish Health Care System”, in Mossialos, E. et al. (eds.), International Profiles of Health Care Systems , Commonwealth Fund. Wammes, J. et al. (2017), “The Dutch Health Care System”, in Mossialos, E. et al. (eds.), International Profiles of Health Care Systems , Commonwealth Fund. Wouters, O., P. Kanavos and M. McKee (2017), “Comparing Generic Drug Markets in Europe and the , Vol. 95/3, pp. 554-601, United States: Prices, Volumes, and Spending”, The Milbank Quarterly http:// dx.doi.org/10.1111/1468-0009.12279 . Zisberg, A. et al. (2015), “Hospital-Associated Functional Decline: The Role of Hospitalization Processes Beyond Individual Risk Factors”, Journal of the American Geriatrics Society , http://dx.doi.org/10.1111/ jgs.13193 . Zuidberg, C. (2010), The pharmaceutical system of the Netherlands , https://ppri.goeg.at/Downloads/ Publications/The%20pharmaceutical%20system%20of%20the%20Netherlands_FINAL.pdf . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 77

80

81 P ART II Overview of health indicators HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018

82

83 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART II Chapter 3 Health status This chapter describes the health status of EU citizens, including recent trends in life expectancy, the main causes of death, health inequalities by gender and socioeconomic status, and the occurrence of communicable and chronic diseases. Life expectancy now reaches 81 years on average across EU countries, but the gains have slowed down markedly in several Western European countries in recent years, with even some reductions in certain years. This appears to have been driven by a slowdown in the rate of reduction of deaths from circulatory diseases and periodical increases in mortality rates among elderly people due partly to bad flu seasons in some years. The main causes of deaths across EU countries remain circulatory diseases (over 1 900 000 deaths in 2015) and cancers (1 320 000 deaths), which together account for over 60% of all deaths. Large inequalities in life expectancy persist not only by gender (women still live nearly 5½ years more than men on average), but also by socioeconomic status. On average across EU countries, 30-year-old men with a low education level can expect to live about 8 years less than those with a university degree or the equivalent. The “education gap” among women is smaller, at about 4 years. Large inequalities also exist in how people rate their health: nearly 80% of adults in the highest income group report to be in good health across EU countries, compared with about 60% of people in the lowest income group. Communicable diseases, such as measles, hepatitis B and many others, pose major threats to the health of European citizens, although vaccination can efficiently prevent these diseases. 13 475 cases of measles were reported across the 30 EU/EEA countries from May 2017 to May 2018, up by nearly 60% over the preceding 12-month period. But in most countries where vaccination coverage is high, very few cases of measles were reported. 81

84 II.3. HEALTH STATUS TRENDS IN LIFE EXPECTANCY Life expectancy has increased in EU countries life expectancy in some countries. Another important over the past decades, but this rise has slowed down factor that has contributed to the recent slowdown in since 2010 in many countries, particularly in Western life expectancy gains in many EU countries is the Europe. slowdown in the reduction in death rates from circulatory diseases, which was previously the main Life expectancy at birth reached 81 years across factor driving life expectancy gains. the 28 EU member states in 2016. Spain and Italy have the highest life expectancy among EU countries, with In the United Kingdom, the recent stalling in life life expectancy reaching over 83 years in 2016. Life expectancy gains has prompted comments about the expectancy at birth now exceeds 80 years in causes, including the possible effects of austerity two-thirds of EU countries, but still remains at only measures on health and other public spending (Hiam around 75 years in Bulgaria, Latvia, Lithuania and et al., 2018). In Europe, some countries that have Romania (Figure 3.1). implemented more severe austerity measures, such as Greece and Spain, have continued to experience rising As is the case around the world, women live life expectancy since 2011, with the notable exception of longer than men in EU countries – on average nearly 5½ 2015 when life expectancy also came down in these two longer – although this gap has narrowed by one year countries. Further research is needed to understand since 2000 as life expectancy among men increased better the recent slowdown in life expectancy gains in more rapidly in most countries.The current gender gap many European countries (Raleigh, 2018). is particularly large in Latvia and Lithuania where women live more than 10 years longer than men, and is also quite large in Bulgaria and Romania. These gender gaps are partly due to greater exposure to risk Definition and comparability factors among men, particularly greater tobacco Life expectancy at birth measures the average consumption, excessive alcohol consumption and less number of years that a person can expect to live healthy diet, resulting in higher death rates from heart based on current mortality rates (age-specific diseases, various types of cancer and other diseases. death rates). However, the actual age-specific Until recently, life expectancy was rising fairly death rates of any particular birth cohort cannot rapidly and steadily across EU countries, by about be known in advance. If age-specific death rates 2½ years per decade on average. However, since 2011, the are falling, actual life spans will on average be gains in life expectancy have slowed down markedly, higher than life expectancy calculated with particularly in some Western European countries, with current death rates. less than half a year gained between 2011 and 2016 in countries like France, Germany, the Netherlands and the United Kingdom. Life expectancy actually decreased in 8 EU countries in 2012 and in 19 countries in 2015, including in France, Germany, Italy and the References United Kingdom, particularly among people aged EuroMOMO (European Monitoring of Excess Mortality for over 75, before recovering in 2016 (Figure 3.2). Public Health Action) (2018), European Mortality . Bulletin, www.euromomo.eu The marked reduction in 2015 was due at least partly to excess mortality in the winter months, Hiam, L., D. Harrison, M. McKee and D. Dorling (2018), Why especially among older people, related to a bad flu is life expectancy in England and Wales “stalling”?, season and increased mortality from cardiovascular , Journal of Epidemiology and Community Health http:// . dx.doi.org/10.1136/jech-2018-210580 diseases. Excess mortality among older people has also been observed during the winter 2017-18 Raleigh, V. (2018), Stalling life expectancy in the UK, (EuroMOMO, 2018), which may impact negatively on , 362, 27 September 2018. British Medical Journal HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 82

85 II.3. HEALTH STATUS 3.1. Life expectancy at birth, by gender, 2016 Women Total Men Years 90 85 80 75 70 65 75.7 75.4 83.5 83.4 82.7 82.5 82.4 82.3 82.2 81.8 81.8 81.7 81.5 81.5 81.5 81.3 81.2 81.2 81.0 81.0 80.9 79.1 78.2 78.0 78.0 77.3 76.2 75.3 74.9 74.9 74.9 83.7 82.5 82.5 78.5 78.1 76.5 60 1. Three-year average (2014-16). Source: Eurostat Database. http://dx.doi.org/10.1787/888933834281 1 2 3.2. Trends in life expectancy, 2005-16 y German EU28 g Ital France United Kin y dom Life expectancy at 75 Life expectancy at birth Years Years 15 85 14 83 13 81 79 12 77 11 75 10 2009 2005 2007 2009 2011 2013 2015 2013 2015 2007 2005 2011 Source: Eurostat Database. http://dx.doi.org/10.1787/888933834300 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 83

86 II.3. HEALTH STATUS INEQUALITIES IN LIFE EXPECTANCY Looking beyond the gap by education level, some Large inequalities in life expectancy exist not countries regularly monitor inequalities in life only by gender, but also by socioeconomic status, no expectancy by income or deprivation level. In France, matter how it is measured – by education level, the results for the period 2012-16 show a gap of income or occupational group. This section focuses 8 years in life expectancy at age 35 between men in mainly on inequalities by education level since this is the top income quartile and those in the bottom the socioeconomic indicator with the most widely income quartile. This gap is slightly smaller (5 years) available data. among women (INSEE, 2018). Inequalities in life expectancy by education level Reducing inequalities in life expectancy across are generally larger among men than among women, socioeconomic groups requires coordinated actions and are particularly large in Central and Eastern Europe involving not only health ministries but also other countries, 30-year-old (Figure 3.3). On average across EU ministries responsible for education, labour, social men with less than upper secondary education can protection and housing (James et al., 2017). expect to live about 8 years less than those with a tertiary education (a university degree or the equivalent). The education gap among women is smaller, at about 4 years. In the Slovak Republic, Hungary, Poland, the Definition and comparability Czech Republic and Latvia, 30-year-old men with a low level of education can expect to live more than 10 years Life expectancy measures the average number less than those with a high level of education. of remaining years of life for people at a specific age based on current mortality conditions. This education gap in life expectancy is due to Education level is based on the ISCED 2011 higher mortality rates among the least educated at classification. The lowest education level refers to different ages. Figure 3.4 shows the difference in the people who have not completed their secondary (age-standardised) mortality rate for some of the main education (ISCED 0-2). The highest education level causes of death between low-educated and high- refers to people who have completed a tertiary educated men and women for two age groups (25-64 education (ISCED 6-8). Data on life expectancy by and 65-89 years) across 10 European countries. The education level have been extracted from the education gap is particularly large among men in both Eurostat database for most countries, with the age groups. While the mortality rate among prime-age exception of Austria, Belgium, France, Latvia, the men (25-64 years) is much lower than among older Netherlands and the United Kingdom which have men (65-89 years), the gap in mortality rate between provided data directly to the OECD. low-educated and high-educated prime-age men is wider – an almost four-fold difference. This gap is due Not all countries have information on education to much higher mortality rates from all the main as part of their deaths statistics. In such cases, causes of death among low-educated prime-age men. data linkage to another source (e.g. a census) Half of the gap in mortality rate among men in that age containing information on education is required. group is due to higher death rates from circulatory Data disaggregated by education level are only diseases and cancer, and another 20% is due to available for a subset of the population for the external causes of death (e.g. accidents and suicides). Czech Republic and Norway. In these two countries, An important gap in mortality rates by education level the large share of the deceased population with also exists among older men and women, driven missing information about their education level mainly by higher death rates from circulatory diseases can affect the accuracy of the data. and cancer (Murtin et al., 2017). Smoking remains a very important risk factor for both circulatory diseases and different types of cancer (notably lung cancer). A substantial part of the education gap in mortality is due to higher smoking References rates among people with a lower level of education James, C. et al. (2017), “Inclusive growth and health”, Chapter 4). (see indicator “Smoking among adults” in , No. 103, OECD Publishing, OECD Health Working Papers A greater prevalence of other risk factors such as Paris, . http://dx.doi.org/10.1787/93d52bcd-en excessive alcohol consumption, particularly among low-educated men, also contribute to higher INSEE (2018), “L’espérance de vie par niveau de vie” [Life mortality rates from circulatory diseases, different expectancy by living standards], Insee Première types of cancer and external (violent) causes of death. No. 1687, February 2018. Gaps in life expectancy at age 30 have remained Murtin, F. et al. (2017), “Inequalities in longevity by relatively stable over the past decade, as life education in OECD countries: Insights from new OECD expectancy increased at about the same rate for estimates”, , 2017/02, OECD Statistics Working Papers lower-educated and higher-educated people in the OECD Publishing, Paris, http://dx.doi.org/10.1787/ group of countries with time series. 6b64d9cf-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 84

87 II.3. HEALTH STATUS 3.3. Gap in life expectancy at age 30 between people with the lowest and highest level of education, 2016 (or nearest year) Men Women Slovak Republic 14.4 6.9 Hungary 12.6 6.4 Poland 12.0 5.1 Czech Republic 11.1 3.0 Latvia 11.0 8.0 Romania 9.7 3.8 Estonia 8.5 5.4 EU21 7.7 4.1 Bulgaria 6.9 4.5 France 6.5 2.6 Slovenia 6.2 2.8 Austria 6.2 3.0 Greece 6.0 2.4 Netherlands 5.8 4.6 Belgium 5.8 4.4 Finland 5.6 3.5 Denmark 5.6 3.9 Portugal 5.6 2.8 Croatia 5.2 1.6 Italy 4.5 2.9 United Kingdom 4.4 4.0 Sweden 4.1 2.9 Norway 5.0 3.4 20 0 5 10 15 20 0 5 10 15 Years Years Note: Data refer to 2012 for France and Austria and to 2011 for Latvia, Belgium and the United Kingdom (England). EU average is unweighted. Eurostat Database; national sources or OECD calculations using national data for Austria, Belgium, France, Latvia, the Source: Netherlands and the United Kingdom (England). 1 2 http://dx.doi.org/10.1787/888933834319 3.4. Mortality rates by education level and causes, 10 European countries, 2011 (or nearest year) Cancer External y Circulator Other Women Men Deaths ulation p p er 10 000 p Deaths per 10 000 population o 600 600 500 500 400 400 300 300 200 200 100 100 0 0 25-64 25-64 65-89 65-89 65-89 25-64 65-89 25-64 Low education High education Low education High education Countries covered are Belgium, the Czech Republic, Denmark, Finland, Hungary, Latvia, Norway, Poland, Slovenia and the Note: United Kingdom (England). Source: Murtin, F. et al. (2017). 1 2 http://dx.doi.org/10.1787/888933834338 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 85

88 II.3. HEALTH STATUS HEALTHY LIFE EXPECTANCY AT BIRTH AND AT AGE 65 Healthy life expectancy is an important indicator of expect to live also about 10 years free of disability on population health. It indicates whether any gains in life average. The number of healthy life years for men at expectancy are lived in good health or with some health age 65 is greater than for women in about half of EU problems and disabilities. A greater number of healthy countries. life years generally means a healthier workforce, fewer Inequalities in healthy life years by socioeconomic early retirements due to health problems, and reduced status are even greater than inequalities in life long-term care needs. expectancy,becausewomenandmenwithlower The main indicator of healthy life years used in education or income are much more likely to report the European Union is the number of years lived free of some activity limitations throughout their lives than activity limitations due to health problems (in other those with higher level of education or income (see words, disability-free life expectancy). On average indicator “Self-reported health and disability”). across EU countries, people can expect to live about A wide range of policies is required to increase 80% of their lives free of disability (Figure 3.5). This healthy life expectancy and reduce inequalities. proportion of healthy life years is lower among women These include greater efforts to prevent health than men (77% vs 81%) because women generally problems starting early in life, promote equal access report more activity limitations due to health problems to care for the whole population, and better manage at any given age and also because women live longer. chronic health problems when they occur to reduce Whereas the gender gap in life expectancy at birth is their disabling effects (OECD, 2017). about 5.5 years on average across EU countries, there is virtually no gap in healthy life expectancy (64.2 years for women compared with 63.5 years for men). Women Definition and comparability in EU countries can expect to live over 19 years of their lives with some disabilities compared with less than Healthy life years (HLY) are defined as the 15 years for men. number of years spent free of long-term activity limitation (this is equivalent to disability-free life In 2016, Malta and Sweden were the two countries expectancy). Healthy life years are calculated with the highest healthy life expectancy among both annually by Eurostat based on life table data and women and men. In these two countries, women can age-specific prevalence data on long-term expect to live more than 85% of their life expectancy activity limitations. The disability measure is the free of disability, and this share reaches around 90% for Global Activity Limitation Indicator (GALI), which men. Latvia, Estonia and the Slovak Republic had measures limitation in usual activities, coming among the lowest healthy life expectancy, reflecting from the EU-SILC survey. both relatively low life expectancy and a substantial share of life lived with some disability. The comparability of the data on healthy life years is limited by the fact that the indicator is As people get older, the share of the remaining derived from self-reported data which can be years of life that they can expect to live free of disability affected by people’s subjective assessment of falls. At age 65, people can only expect to live about their activity limitation (disability) and by social 50% of their remaining years of live free of disability and cultural factors. There are also differences across EU countries (Figure 3.6). Again, this proportion across countries in the formulation of the is substantially smaller among women (47% only) than question on disability in national languages in EU- men (54%), because women report more disability at SILC, limiting data comparability (Eurostat, 2017). any specific age and because they live longer. Women can expect to live another 21.6 years when they reach age 65 across the EU, but only about 10 of these years can be expected to be free of activity limitation, with Reference the other 11.5 years lived with some disabilities. For men, the remaining life expectancy at age 65 is more OECD (2017), Preventing Ageing Unequally , OECD Publishing, than three years shorter (18.2 years), but they can http://dx.doi.org/10.1787/9789264279087-en Paris, . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 86

89 II.3. HEALTH STATUS 3.5. Life expectancy and healthy life years at birth, by gender, 2016 (or nearest year) Healthy life years Life expectancy with activity limitation Women Men 86.3 23% S ain p 80.5 18% 85.7 25% France 79.5 21% 85.6 21% Ital y 81.0 17% 85.1 28% Luxembour g ¹ 79.8 21% 84.4 22% C yp rus ¹ 80.4 18% 84.4 32% Finland 78.6 25% 84.3 32% Portu g al 78.1 23% 84.3 31% Slovenia 78.2 25% 84.2 12% Malta ¹ 80.0 10% 84.1 32% Austria 79.3 28% 84.1 13% Sweden 80.6 9% 84.0 24% Bel ium g 79.0 19% 84.0 23% Greece 78.9 19% 83.6 23% EU28 78.2 19% 83.6 17% 79.9 16% Ireland 83.5 19% 78.6 17% German y 83.2 31% 80.0 22% Netherlands 83.0 24% 79.4 21% United Kin dom g 82.8 27% 79.0 24% Denmark 82.2 28% 73.3 26% Estonia 82.1 22% 76.1 18% Czech Re p . 82.0 21% 73.9 17% Poland 81.3 28% 75.0 24% Croatia 80.7 29% 73.8 24% Slovak Re p . 80.1 26% 69.5 19% Lithuania 79.7 24% 72.6 18% Hun ar y g 79.6 31% 69.8 25% Latvia 79.1 71.7 25% 17% Romania 78.5 71.3 14% 10% Bul aria g 81.7 25% 85.6 33% Switzerland 80.7 11% 84.2 19% Norwa y 84.0 81.0 21% 12% Iceland ¹ 51 02 55 07 00 0 25 100 75 50 Years Years 1. Three-year average (2014-16 except for Iceland: 2013-15). Data comparability is limited because of cultural factors and different formulations of question in EU-SILC. Note: Source: Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933834357 3.6. Life expectancy and healthy life years at 65, by gender, 2016 (or nearest year) Healthy life years Life expectancy with activity limitation Women Men 19.6 52% 23.7 55% France 19.4 46% 23.6 56% S p ain 19.4 46% 22.9 56% Ital y 18.7 44% 22.4 59% g ¹ Luxembour 18.4 44% 21.9 48% Bel ium g 18.2 48% 21.9 59% Finland 19.0 31% 21.8 38% Malta ¹ 18.0 57% 21.8 71% Portu al g 17.9 53% 21.8 62% Slovenia 18.5 56% 21.7 66% Austria 18.9 58% 21.7 64% Greece 18.2 46% 21.6 53% EU28 19.1 21% 21.5 23% Sweden 18.8 47% 21.3 59% C rus ¹ yp 18.1 37% 21.3 42% German y 18.6 35% 21.1 37% Ireland 18.5 44% 21.1 53% Netherlands 18.8 45% 21.1 47% United Kin dom g 15.6 65% 20.9 67% Estonia 18.2 37% 20.8 43% Denmark 16.0 49% 20.5 57% Poland 16.2 48% 20.0 56% Czech Re ublic p 14.2 61% 19.4 71% Lithuania 15.6 67% 19.2 74% Croatia 15.3 71% 19.2 78% Slovak Re . p 14.0 69% 19.0 76% Latvia 14.6 54% 18.7 66% Hun g ar y 18.3 14.7 69% 58% Romania 17.9 14.2 44% 35% Bul g aria 22.9 20.0 57% 50% Switzerland 21.6 19.1 30% 19% Norwa y 21.6 19.3 30% 20% Iceland ¹ 0 03 0 1 0 10 02 20 30 Years Years 1. Three-year average (2014-16 except for Iceland: 2013-15). Note: Data comparability is limited because of cultural factors and different formulations of question in EU-SILC. Source: Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933834376 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 87

90 II.3. HEALTH STATUS MAIN CAUSES OF MORTALITY Over 5 200 000 people died in EU countries in 2015 diseases, followed by pneumonia (see indicator (Figure 3.7). An unusual large increase in the number of “Mortality from respiratory diseases”). deaths in 2015 explains the reduction in life External causes of death, which include accidents, expectancy in many countries compared with 2014 (see suicides, homicides and other violent causes of death, indicator “Trends in life expectancy”). The higher were responsible for 3% of all deaths among women and number of deaths in 2015 across EU countries was 6% of deaths among men in EU countries in 2015. The concentrated mainly among people aged 75 and over, most important causes of violent deaths are road traffic and was attributed mainly to higher mortality from accidents and other accidental deaths, and suicides. influenza and pneumonia triggering cardiorespiratory Road traffic accidents are a particularly important cause events, Alzheimer’s disease and other dementias, and of death among young people (aged 18-25), whereas heart diseases. suicide rates generally increase with age. Slightly more women than men died across EU More than 80% of all deaths in EU countries occur countries in 2015, as there are more women in the after the age of 65. While the main cause of death population, particularly in older age groups. Once the among people aged over 65 is circulatory diseases, the population structure is adjusted by age, the age- main cause for people under 65 is cancer, particularly standardised mortality rate was about 50% higher among women (Eurostat, 2018). among men across the EU as a whole (1 287 per 100 000 Overall mortality rates vary widely across men compared with 849 per 100 000 women). countries. France, Spain and Italy have the lowest The main causes of death in EU countries are death rates, with age-standardised rates between circulatory diseases and various types of cancer, 850 and 900 deaths per 100 000 population in 2015 followed by respiratory diseases and external causes of (Figure 3.8). This was mainly due to relatively low death. mortality rates from circulatory diseases. Mortality Circulatory diseases continue to be the leading rates are highest in Bulgaria, Romania and Hungary, cause of death across the EU, accounting for over with age-standardised rates at least 50% higher than 1 900 000 deaths in 2015. Ischaemic heart diseases, which the EU average in 2015. The main reason for this much include heart attack and other diseases, and stroke are higher mortality rate in Bulgaria and Romania is higher the most common causes of death from circulatory mortality rates from circulatory diseases. In Hungary, diseases (see indicator “Mortality from circulatory higher mortality rates from cancer explain a large part diseases”). The age-standardised mortality rate from of the difference with the EU average. circulatory diseases is much higher among men than women (about 40% higher), but nonetheless diseases of the circulatory system account for a greater share of Definition and comparability deaths among women than men across EU countries. Deaths from all causes are classified to ICD-10 Some 1 320 000 people died of cancer in 2015, codes A00-Y89, excluding S00-T98. Mortality rates accounting for 22% of all deaths among women and are based on the number of deaths registered in a 29% of all deaths among men. Breast cancer and lung country in a year divided by the population. The cancer are the leading causes of cancer death among rates have been age-standardised to the revised women, whereas lung cancer and colorectal cancer European standard population adopted by are the two main causes of cancer death for men (see Eurostat in 2012 to remove variations arising from indicator “Mortality from cancer”). differences in age structures across countries and After circulatory diseases and cancer, respiratory over time. diseases are the third leading cause of death in EU countries, causing some 440 000 deaths in 2015, with the vast majority of these deaths occurring among people aged over 65. This group of diseases accounted Reference for 8% of all death among women and 9% among men. Chronic obstructive pulmonary disease (COPD) is the Eurostat (2018), “Causes of Death Statistics – People Over 65”, most common cause of mortality among respiratory Statistics Explained , European Commission, April. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 88

91 II.3. HEALTH STATUS 3.7. Main causes of mortality among women and men in EU countries, 2015 Respiratory diseases Cancer Circulatory diseases External causes Other causes Wom Me en n 2 639 135 d hs (2015) eaths (2015) 2 577 961 d eat 1 287 deaths per 100 000 population (age-standardised) 849 deaths per 100 000 population (age-standardised) 583 039 22% 714 249 27% 874 866 34% 1 041 776 40% 150 224 6% 87 690 3% 230 686 9% 211 434 8% 739 146 583 986 29% 22% External causes of death include accidents, suicides, homicides and other causes. Note: Eurostat Database. Source: http://dx.doi.org/10.1787/888933834395 1 2 3.8. Main causes of mortality by country, 2015 Other External causes Respiratory system Cancer Circulatory system Age-standardised rates per 100 000 population 1800 1660 1500 1648 1530 1490 1489 1600 1430 1390 1280 1400 1274 1237 1168 1200 1057 1041 1036 1033 1005 1003 1002 1002 993 992 989 990 969 956 930 927 913 907 901 1000 873 868 859 800 600 400 200 0 1. Three-year average (2013-15). External causes of death include accidents, suicides, homicides and other causes. Note: Eurostat Database. Source: 1 2 http://dx.doi.org/10.1787/888933834414 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 89

92 II.3. HEALTH STATUS MORTALITY FROM CIRCULATORY DISEASES Circulatory diseases remain the main cause of mortality, the disability burden from stroke is mortality in nearly all EU member states, accounting substantial. The gender gap in (age-standardised) for some 1 910 000 deaths and 37% of all deaths across mortality rates from stroke is not as large as for IHD EU countries in 2015. Circulatory diseases comprise a (less than 20%). range of illnesses related to the circulatory system, As with IHD, there are wide variations in stroke including ischaemic heart diseases (notably heart mortality rates across countries. The rates are three attacks) and cerebrovascular diseases (such as strokes). times higher than the EU average in Bulgaria, Latvia Ischaemic heart diseases and strokes alone account for and Romania. They are the lowest in France, over 55% of all deaths from circulatory diseases, and igure 3.10). Luxembourg and Spain (F caused more than one-fifth of all deaths in EU member Since 2000, stroke mortality rates have decreased states in 2015. by nearly 50% across the EU, although the gains have Ischaemic heart diseases (IHD) are caused by the slowed down over the past five years. The reduction accumulation of fatty deposits lining the inner wall of since 2000 has been much slower in some countries a coronary artery, restricting blood flow to the heart. like Bulgaria and Lithuania (only a 10% to 15% Death rates for IHD are over 80% higher for men than reduction) compared with a reduction of between 40% for women across EU countries, because of a greater to 50% in Finland, France and Germany (Figure 3.12). prevalence of risk factors among men, such as As with IHD, the reduction in stroke mortality can be smoking, hypertension and high cholesterol. attributed at least partly to both a reduction in risk Mortality rates from IHD are highest in Lithuania, factors and improvements in medical treatments (see Latvia, the Slovak Republic and Hungary, with age- indicator “Mortality following stroke” in Chapter 6). standardised rates more than three times greater Looking ahead, further progress in reducing than the EU average. The countries with the lowest mortality rates from IHD, strokes and other circulatory IHD mortality rates are France, the Netherlands, diseases may be hampered by a rise in certain risk Portugal and Spain, with death rates about two times factors such as obesity and diabetes (OECD, 2015). lower than the EU average (F igure 3.9). Since 2000, age-standardised mortality rates from IHD have declined in all countries, with an Definition and comparability overall reduction of over 40% on average across the Mortality rates are based on the number of EU, although the reduction has slowed down in recent deaths registered in a country in a year divided years (Figure 3.11). The decrease since 2000 has been by the population. The rates have been age- quite modest in some countries like Lithuania (only a standardised to the revised European standard 4% reduction), whereas it has been more rapid in population adopted by Eurostat in 2012 to Finland (a 44% reduction). Reductions in risk factors remove variations arising from differences in such as tobacco consumption have contributed to age structures across countries and over time. reducing the incidence of IHD and consequently mortality rates (see indicator “Smoking among Deaths from ischaemic heart diseases relate to adults” in Chapter 4). Improvements in medical care ICD-10 codes I20-I25, and stroke (or cerebrovascular have also played an important role (see indicator diseases) to I60-I69. “Mortality following acute myocardial infarction” in Chapter 6). Strokes (or cerebrovascular diseases) were Reference responsible for some 430 000 deaths across the EU in 2015, accounting for about 8% of all deaths. Strokes OECD (2015), Cardiovascular Disease and Diabetes: Policies for are caused by the disruption of the blood supply to the Better Health and Quality of Care , OECD Publishing, brain. In addition to being an important cause of . http://dx.doi.org/10.1787/9789264233010-en Paris, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 90

93 II.3. HEALTH STATUS 3.10. Stroke mortality, 2015 3.9. Ischaemic heart disease mortality, 2015 Total Women Total Women Men Men France France Spain Netherlands Luxembourg ¹ Portugal Belgium Spain Austria Belgium Sweden Denmark Germany Luxembourg ¹ Ireland Italy United Kingdom Greece Netherlands Cyprus ¹ Denmark Slovenia Estonia United Kingdom Cyprus ¹ Sweden Finland EU28 Malta ¹ Poland Italy Germany EU28 Ireland Poland Austria Slovenia Finland Portugal Bulgaria Greece Malta ¹ Czech Rep. Estonia Hungary Croatia Slovak Rep. Romania Croatia Czech Rep. Lithuania Hungary Romania Slovak Rep. Latvia Latvia Bulgaria Lithuania Switzerland Norway Norway Switzerland Iceland ¹ Iceland ¹ Turkey Serbia Serbia Turkey 0 100 200 300 400 200 0 600 800 400 Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population 1. Three-year average (2013-15). 1. Three-year average (2013-15). Eurostat Database. Source: Eurostat Database. Source: http://dx.doi.org/10.1787/888933834433 1 2 1 2 http://dx.doi.org/10.1787/888933834452 3.12. Trends in stroke mortality, selected EU 3.11. Trends in ischaemic heart disease mortality, selected EU countries, 2000-15 countries, 2000-15 Germany Spain EU28 Bulgaria Germany EU28 Finland France Finland Lithuania France Lithuania Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population 400 700 350 600 300 500 250 400 200 300 150 200 100 100 50 0 0 Note: OECD estimates of EU28 average for 2000 and 2001. OECD estimates of EU28 average for 2000 and 2001. Note: Eurostat Database. Source: Eurostat Database. Source: http://dx.doi.org/10.1787/888933834490 1 2 http://dx.doi.org/10.1787/888933834471 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 91

94 II.3. HEALTH STATUS MORTALITY FROM CANCER Cancer caused some 1 320 000 deaths in the (see indicator “Survival and mortality from colorectal igure 3.13). It is the second European Union in 2015 (F Chapter 6). cancer” in leading cause of mortality after cardiovascular Breast cancer is the leading cause of cancer diseases, accounting for 25% of all deaths in 2015. death among women, causing 94 300 deaths in 2015 Mortality rates from cancer are lowest in Cyprus, and accounting for 16% of all female cancer deaths. Finland, Malta, Spain and Sweden, with rates at least While incidence rates of breast cancer have increased 10% lower than the EU average. They are highest in over the past decade, death rates have declined or Hungary, Croatia, the Slovak Republic, Slovenia and stabilised, indicating increases in survival rates due to Poland, with rates more than 15% higher than the EU earlier diagnosis and better treatment (see indicator average (Figure 3.14). “Screening, survival and mortality for breast cancer” in Chapter 6). In all countries, mortality rates from cancer are greater among men than women. Overall, some Prostate cancer is the third most common cause 584 000 women and 739 000 men died from various of cancer deaths among men across EU countries types of cancer in EU countries in 2015. The aged- (particularly among men aged over 65), resulting in standardised mortality rates from cancer was 70% 75 300 deaths in 2015 and accounting for 10% of all higher among men than women on average in the EU male cancer deaths. (346 deaths per 100 000 men, compared with Death rates from all types of cancer combined 201 deaths per 100 000 women). This gender gap is among men and women have declined at least slightly particularly wide in Latvia, Lithuania, Estonia, Spain in most EU member states since 2000, although the and Portugal, with mortality rates more than two decline has been more modest than for circulatory times greater among men than among women. It can diseases, explaining why cancer now accounts for a be explained by the greater prevalence of risk factors larger share of all deaths. among men (e.g. smoking and alcohol consumption), as well as the more limited availability or use of screening programmes for cancers affecting men, Definition and comparability leading to lower survival rates after diagnosis. Mortality rates are based on the number of Lung cancer remains by far the most common deaths registered in a country in a year divided cause of death from cancer among men (25% of all by the population. The rates have been age- cancer deaths across the EU) and the second most standardised to the revised European standard common among women (after breast cancer). Some population adopted by Eurostat in 2012 to 184 000 men and 89 000 women died from lung cancer remove variations arising from differences in in EU countries in 2015. Smoking is the main risk factor age structures across countries and over time. for lung cancer. Over the past 10 years, the mortality rate from lung cancer increased by almost 20% across Deaths from all cancers relate to ICD-10 codes EU countries, driven mainly by a large increase in C00-C97, lung cancer to C33-C34. The international deaths among women in many countries. This reflects comparability of cancer mortality data can be the fact that many women started to smoke several affected by differences in medical training and decades later than men (Torre et al., 2014). practices as well as in death certification procedures across countries. Colorectal cancer is the second most common cause of cancer death, killing some 154 200 men and women in EU countries in 2015. The mortality rate from colorectal cancer is about 75% higher among References men than among women across EU countries. There are several risk factors for colorectal cancer besides Torre et al. (2014), International variation in lung cancer genetic factors and age, including a diet high in fat Cancer mortality rates and trends among women, and low in fibre, alcohol consumption, smoking and , Vol. 23, No. 6, pp. 1025-36. Epidemiology Biomarkers Prev obesity. The mortality rate has declined over the past OECD (2013), Cancer Care: Assuring Quality to Improve decade in most countries, due to a large extent to , OECD Health Policy Studies, OECD Publishing, Survival earlier detection and higher survival after diagnosis Paris, http://dx.doi.org/10.1787/9789264181052-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 92

95 II.3. HEALTH STATUS 3.13. Main causes of cancer mortality among men and women in EU countries, 2015 Men Women 739 146 deaths 583 986 deaths Stomach Stomach 22 270 34 920 4% 5% Others Colorectal Li ver Others 281 453 69 199 Colorectal 17 967 219 204 38% 12% 85 138 Li ver 3% 38% 11% 34 998 5% Panc reas 42 340 reas Panc 7% 42 676 6% Lung 88 773 15% Lung Prostate 184 605 Ovary 75 356 Breast 25% 29 798 10% 94 435 5% 16% Eurostat Database. Source: 1 2 http://dx.doi.org/10.1787/888933834509 3.14. Cancer mortality, 2015 Men Women Total Age-standardised rates per 100 000 population 500 450 400 350 300 250 200 150 100 50 0 1. Three-year average (2013-15). Source: Eurostat Database. http://dx.doi.org/10.1787/888933834528 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 93

96 II.3. HEALTH STATUS MORTALITY FROM RESPIRATORY DISEASES Mortality from respiratory diseases is the third Nearly 6 000 deaths were directly attributed to main cause of death in EU countries, accounting for influenza, with most of these deaths concentrated 8% of all deaths in 2015. More than 440 000 people among people aged over 65. But influenza also died from respiratory diseases in 2015, an increase of contributed to many more deaths among frail elderly 15% over the previous year. Most of these deaths (90%) people with chronic diseases. The European Monitoring were among people aged 65 and over. The main of Excess Mortality network estimated that up to causes of death from respiratory diseases are chronic 217 000 deaths were related to influenza among elderly obstructive pulmonary disease, pneumonia, asthma people across EU countries during the winter 2015 and influenza. (EuroMoMo, 2016). In 2015, the United Kingdom and Ireland had the The prevalence and mortality from respiratory highest age-standardised death rates from respiratory diseases are likely to increase in the coming years as diseases among EU countries (Figure 3.15). Finland, the population ages and presently unreported cases of Latvia, Estonia and Lithuania had the lowest rates, COPD begin to manifest, whether alone or in with rates only about half the EU average. co-morbidity with other chronic diseases. Death rates from respiratory diseases are on Many deaths from respiratory diseases could be average 85% higher among men than among women prevented by tackling some of the main risk factors, in all EU countries. This is partly due to higher notably smoking, and by increasing vaccination coverage smoking rates among men. Smoking is an important for influenza and pneumonia, particularly among elderly risk factor for chronic obstructive pulmonary disease people and other vulnerable groups. Better management and other respiratory diseases. of both asthma and COPD in primary care could also help reduce health complications. Chronic obstructive pulmonary disease (COPD) (or chronic lower respiratory diseases), which includes chronic bronchitis and emphysema, caused over Definition and comparability 180 000 deaths in EU countries in 2015 and accounted for over 40% of all respiratory disease mortality. Mortality rates are based on the number of Mortality from COPD varies widely across countries. deaths registered in a country in a year divided Hungary, Denmark and the United Kingdom have the by the population. The rates have been age- highest rate of mortality from COPD, with age- standardised to the revised European standard standardised rates at least two-thirds higher than the population adopted by Eurostat in 2012 to EU average (Figure 3.16). The main risk factor for COPD remove variations arising from differences in is tobacco smoking (both active and passive smoking), age structures across countries and over time. but other risk factors include occupational exposure to Deaths from respiratory diseases relate to dusts, fumes and chemicals, and air pollution more ICD-10 codes J00-J99, with pneumonia relating generally. A large number of people with COPD are only to J12-J18, chronic obstructive pulmonary diagnosed at a late stage, contributing to higher disease (or chronic lower respiratory diseases) mortality. People with COPD are also more susceptible relating to J40-J47 and asthma to J45-J46. The to influenza and pneumonia. international comparability of data on mortality Pneumonia was responsible for nearly from respiratory diseases can be affected by 140 000 deaths in EU countries in 2015, accounting for differences in medical training and coding over 30% of all respiratory disease mortality. As with practices for causes of death. Finland revised COPD, there are large variations in mortality rates some coding practices in 2005-06, leading across EU countries: Portugal, the Slovak Republic and especially to a decrease of recorded deaths the United Kingdom have the highest rates of caused by pneumonia. pneumonia mortality, whereas Finland, Greece and Austria have the lowest rates (Figure 3.17). The main risk factors for pneumonia are age, smoking and alcohol abuse, and having COPD or HIV infection References (Torres et al., 2013). Torres, A. et al. (2013), “Risk Factors for Community- More than 7 000 people died from asthma in EU acquired Pneumonia in Adults in Europe: A Literature Review”, Thorax , Vol. 68, pp. 1057-1065. countries in 2015. Mortality rates from asthma are highest in Estonia, Ireland and the United Kingdom, EuroMoMo (2016), “Excess mortality in Europe in the but remain much lower than for COPD and winter season 2014/15, in particular amongst the pneumonia. elderly”, Winter season summary 2015. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 94

97 II.3. HEALTH STATUS 3.15. Respiratory diseases mortality, 2015 Men Women Total Age-standardised rates per 100 000 population 220 200 180 160 140 120 100 80 60 40 20 0 1. Three-year average (2013-15). Source: Eurostat Database. http://dx.doi.org/10.1787/888933834547 1 2 3.16. COPD mortality, 2015 3.17. Pneumonia mortality, 2015 Latvia 14.4 Finland 2.8 France 16.8 Greece 9.5 Estonia 18.1 Austria 11.8 Bulgaria 21.2 Hungary 13.6 Slovenia 21.8 Cyprus ¹ 13.8 Poland 23.2 Croatia 15.2 Finland 23.7 Italy 16.0 Lithuania Latvia 24.7 18.1 Malta ¹ France 26.0 18.5 Cyprus ¹ Luxembourg ¹ 26.1 19.4 Slovak Rep. Bulgaria 26.4 19.7 Portugal Lithuania 27.6 19.8 Estonia Greece 20.4 28.4 Sweden Italy 20.6 30.6 Spain Sweden 20.6 31.0 Germany Spain 23.2 34.5 Netherlands Austria 24.7 35.7 EU28 28.1 EU28 36.3 Czech Rep. Luxembourg ¹ 33.1 36.5 Denmark 36.5 Romania 36.9 Slovenia 37.7 Germany 39.3 Malta ¹ Czech Rep. 37.7 40.8 Belgium 37.7 Belgium 42.9 Romania 38.5 Netherlands 47.6 Ireland 42.1 Croatia 49.7 Poland 49.8 Ireland 58.2 United Kingdom 53.7 United Kingdom 60.9 Slovak Rep. 57.2 Denmark 69.2 Portugal Hungary 72.5 57.7 Switzerland 19.2 Switzerland 27.5 Serbia 21.7 Iceland ¹ 37.9 Iceland ¹ 28.1 Serbia 41.9 Norway 33.7 Norway 51.4 Turkey 41.5 Turkey 86.8 0 102030405060 0 20406080100 Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population 1. Three-year average (2013-15). 1. Three-year average (2013-15). Source: Eurostat Database. Source: Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933834585 1 2 http://dx.doi.org/10.1787/888933834566 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 95

98 II.3. HEALTH STATUS INFANT HEALTH The Baltic countries (Estonia, Latvia and Lithuania) Infant mortality reflects the effect of socioeconomic and the Nordic countries (Finland, Sweden and conditions on the health of mothers and newborns, as Denmark) have the lowest proportion of low birth well as the effectiveness of health systems, particularly in weight babies, whereas some countries in Southern addressing any life-threatening problem during the Europe (Cyprus, Greece, Bulgaria and Portugal) have the neonatal period (i.e. during the first four weeks). highest proportion. While this proportion has decreased Infant mortality rates are low in most EU countries, slightly over the past decade in Cyprus, it has increased with an average of less than 4 deaths per 1 000 live slightly in Greece. Some suggest that the peak of 10% of births across EU countries in 2016 (F igure 3.18). However, low birth weight infants in 2010 in Greece, a sharp a small group of countries – Romania, Bulgaria, Malta increase compared with 2008, may be due to the impact and the Slovak Republic – still have infant mortality of the economic crisis on household’s access to health rates above 5 deaths per 1 000 live births. These rates, care (Kentikelenis, 2014). In Portugal, the proportion of though, have declined steadily over the past 25 years. In low birth weight babies also increased over the past Malta, infant mortality rates may be higher because decade, from 7.6% of all live births in 2006 to 8.7% in induced abortions following the detection of congenital 2013, with the rate broadly stable since then. anomalies are illegal, whereas this is possible in other countries in cases of severe and/or lethal anomalies. Around two-thirds of the deaths during the first year of life occur during the first month (i.e. neonatal Definition and comparability mortality). The main causes of death during the first Infant mortality rate is the number of deaths of month are congenital anomalies, prematurity and children under one year of age per 1 000 live births. other conditions arising during pregnancy. For deaths Some of the international variation in infant and beyond one month (post neonatal mortality), there neonatal mortality rates may be due to variations tends to be a greater range of causes – the most among countries in registering practices of common being Sudden Infant Death Syndrome (SIDS), premature infants. While some countries have no birth defects, infections and accidents. gestational age or weight limits for mortality All European countries have achieved notable registration, several countries apply a minimum progress in reducing infant mortality rates over the gestational age of 22 weeks (or a birth weight past few decades.The EU average went down from over threshold of 500 grams) for babies to be registered 10 deaths per 1 000 live births in 1990 to 3.6 deaths in as live births (Euro-Peristat, 2013). 2016. Reductions in infant mortality rates have been Low birth weight is defined by the World Health particularly rapid in Bulgaria, Poland and Romania, Organization as the weight of an infant at birth of converging towards the EU average (Figure 3.19). less than 2 500 grams (5.5 pounds) irrespective of However, the downward trend in infant mortality has the gestational age of the infant. This threshold is halted in recent years in a number of Western based on epidemiological observations regarding European countries, at least partly because of the increased risk of death of the infant. Despite increasing numbers of low birth weight infants. the widespread use of this 2 500 grams limit, Across EU countries, 1 in 14 babies (7.0%) weighed physiological variations in size occur across less than 2 500 grams at birth in 2016 (Figure 3.20). This different countries and population groups, and is up slightly from 1 in 15 babies (6.7%) in 2000. Low these need to be taken into account when birth weight can occur as a result of restricted foetal interpreting differences (Euro-Peristat, 2013). The growth or from pre-term birth. Low birth weight number of low weight births is expressed as a infantshaveagreaterriskofpoorhealthordeath, percentage of total live births. require a longer period of hospitalisation after birth, and are more likely to have health problems and disabilities later in life. Some of the main risk factors for low birth weight include maternal smoking, alcohol consumption and poor nutrition during pregnancy, low References body mass index, lower socio-economic status, having Euro-Peristat (2013), “European Perinatal Health Report: had in-vitro fertilisation treatment and multiple births, The Health and Care of Pregnant Women and their and a higher maternal age. The increased use of Babies in 2010”, Luxembourg. delivery management techniques such as induction of labour and caesarean delivery, which have increased Kentikelenis, A. (2014), “Greece’s health crisis: From the survival rates of low birth weight babies, also partly , Vol. 383, Issue 9918, austerity to denialism”, The Lancet explain the small rise in low birth weight infants. pp. 748-753. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 96

99 II.3. HEALTH STATUS 3.18. Infant mortality, 2016 Deaths per 1 000 live births 14 11.9 12 10 10 8.7 7.0 8 6.5 6.1 5.4 5.4 6 4.5 4.3 4.2 4.0 3.9 3.8 3.7 3.7 3.6 3.6 3.5 3.5 3.4 3.2 3.2 3.1 3.1 3.1 3.0 2.8 2.8 4 2.7 2.5 2.5 2.3 2.2 2.0 1.9 1.7 2 0 1. Three-year average (2014-16). Source: Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933834604 3.19. Trends in infant mortality, 1990-2016 Poland EU28 Bulgaria Romania Deaths per 1 000 live births 30 25 20 15 10 5 0 1996 2014 2012 2010 2016 2008 1990 1992 1994 2006 1998 2000 2002 2004 Source: Eurostat Database. http://dx.doi.org/10.1787/888933834623 1 2 3.20. Low birthweight, 2016 (or nearest year) % of live births 12 10.2 9.4 9.3 10 8.7 8.5 8.3 7.8 7.7 7.6 7.5 7.5 7.5 7.4 7.0 8 6.9 6.9 6.8 6.6 6.6 6.6 6.4 6.0 5.8 5.6 5.1 5.0 6 4.7 4.5 4.4 4.4 4.3 4.1 4.1 4 2 0 Source: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en , Eurostat Database and national source for Cyprus. http://dx.doi.org/10.1787/888933834642 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 97

100 II.3. HEALTH STATUS SELF-REPORTED HEALTH AND DISABILITY The health module in the EU Statistics on Income their usual daily activities because of a health and Living Conditions survey (EU-SILC) allows problem (Figure 3.23). This proportion is highest in respondents to report on their health status, whether Latvia, Austria, Portugal and Finland (with one-third they are generally in good health, have a chronic or more of respondents reporting such limitations). disease and are limited in their usual activities because Women report more often to be limited in their daily of a health problem (a common definition of disability). activities than men (26% versus 22% on average across EU member states). As expected, such activity Cross-country differences in perceived health limitations increase greatly with age: about 60% of status can be difficult to interpret because social and people aged 75 years and over report to be limited in cultural factors may affect responses. Further, since their daily activities. As with other indicators of older people generally report poorer health and more health, there are also disparities in this indicator of chronic diseases than younger people, countries with disability by income group: on average across EU a larger proportion of elderly people may have a lower countries, about 16% of people in the highest income proportion of people reporting to be in good health group report such activity limitations compared with and without any chronic disease or disability. 30% for people in the lowest income group. With these limitations in mind, most adults in the It is likely that there is also a reverse causal link European Union rate their health quite positively: between health and income inequalities, with poor two-thirds of people aged 16 and over report to be in good health status leading to lower employment and lower health in 2016 (Figure 3.21). Ireland, Cyprus, the income. Netherlands and Sweden have the highest share of adults rating their health to be good, with at least three-quarters doing so. In contrast, less than half of adults in Lithuania, Latvia and Portugal report to be in good health. Definition and comparability Menaremorelikelythanwomentoratetheir The questions used in the EU-SILC survey to health as good. There are also disparities in self- measure health status generally and the reported health across different socio-economic groups. prevalence of any chronic disease and disability On average across EU countries, nearly 80% of people in are: i) “How is your health in general? Is it very the highest income quintile report to be in good health, compared with about 60% for people in the lowest good, good, fair, bad, very bad?”, ii) “Do you have income quintile. These disparities are particularly large any long-standing illness or health problem in Baltic countries (Estonia, Latvia and Lithuania). In which has lasted or is expected to last for these three countries, at least two-thirds of people in the 6 months or more”; and iii) “For at least the past highest income group report to be in good health (which 6 months, to what extent have you been limited is equal to the EU average for all the population), but this because of a health problem in activities people proportion goes down to about one-third only for people usually do? Would you say you have been in the lowest income group. These disparities can be severely limited, limited but not severely, or not explained by differences in living and working limited at all?” (Data reported here include both conditions, as well as differences in lifestyles (e.g. people who say that they are limited severely or smoking, harmful alcohol drinking, physical inactivity, not severely). People living in institutions are not and obesity). surveyed. One-third of adults in EU member states reports The income level is reported for the lowest having a chronic disease or health problem (F igure 3.22). income quintile (people in the bottom 20% of the Adults in Finland and Estonia are more likely to report income distribution) and the highest income having some chronic illnesses or health problems, while such chronic conditions are less commonly reported in quintile (the top 20%). The income may relate Italy, Romania and Bulgaria. Women report some long- either to the individual or the household (in which standing illnesses or health problems more often than case the income is equivalised to take into account men (35% versus 31% across EU member states). There the number of persons in the household). are also some disparities in reporting chronic illnesses Caution is required in making cross-country by income group: on average, less than 30% of people in comparisons of perceived health status, since the highest income group report some chronic diseases people’s assessment of their health is subjective or health problems, compared with less than 40% for and can be affected by social and cultural factors. people in the lowest income group.These disparities are There are also differences in the formulation of particularly large again in the Baltic countries (Estonia, the question on disability across countries, Latvia and Lithuania). limiting the comparability of the data. Almost one-quarter of adults on average across EU member states reports that they are limited in HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 98

101 II.3. HEALTH STATUS 3.21. Health status perceived as good or very good, by income quintile, 2016 (or nearest year) Low income High income Total population % of population aged 16 years and over 100 90 80 70 60 50 40 30 20 83 79 76 75 74 74 73 73 71 71 71 70 70 69 69 68 67 66 66 65 65 60 60 59 59 53 48 47 43 78 78 77 76 70 66 57 10 0 Source: Eurostat Database, based on EU-SILC. http://dx.doi.org/10.1787/888933834661 1 2 3.22. Self-reported chronic condition, by income quintile, 2016 (or nearest year) Low income Total population High income % of population aged 16 years and over 70 60 50 40 30 20 10 47 44 42 41 40 39 38 38 37 36 36 35 35 35 35 33 33 33 33 31 29 29 27 26 26 24 21 19 15 38 36 34 31 30 17 15 0 Source: Eurostat Database, based on EU-SILC. 1 2 http://dx.doi.org/10.1787/888933834680 3.23. Self-reported disability, by income quintile, 2016 (or nearest year) Total population High income Low income % of population aged 16 years and over 60 50 40 30 20 10 24 23 22 22 21 31 27 23 18 37 34 33 33 32 32 31 30 29 29 29 27 27 26 26 25 25 25 24 12 17 16 17 16 16 13 13 0 Eurostat Database, based on EU-SILC. Source: 1 2 http://dx.doi.org/10.1787/888933834699 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 99

102 II.3. HEALTH STATUS NOTIFIED CASES OF VACCINE-PREVENTABLE DISEASES Communicable diseases, such as measles, Norway. The higher number of reported cases in these hepatitis B and many others, pose major threats to countries is due at least partly to a more comprehensive the health of European citizens, although vaccination surveillance and reporting system that includes both could efficiently prevent these diseases (EC, 2018). acute and chronic cases. The vast majority of cases Measles, a highly infectious disease of the respiratory reported in these countries are chronic cases. Many system, is caused by a virus. Symptoms include fever, countries with low rates such as France, Greece and cough, runny nose, red eyes and skin rash. It can lead Lithuania do not report such chronic cases. to severe health complications, including pneumonia, Reported cases of hepatitis B are higher in men encephalitis, diarrhoea and blindness. than in women. About one-third of all reported 13 475 cases of measles were reported to the hepatitis B cases occurs among people aged 25-34. For European Surveillance System by the 30 EU/EEA acute infections, heterosexual transmission is the countries from May 2017 to May 2018, up from 8 523 most common route of transmission, followed by cases for the preceding 12-month period. The average nosocomial transmission, transmission among men rate in the EU in 2017 was 2.2 cases per 100 000 who have sex with men, injuries and drug injection. population, but with wide variations across countries Mother-to-child transmission is the most common (Figure 3.24). Romania reported the highest number of route for chronic cases (ECDC, 2018b). The most new cases and highest rate (28.4 cases per 100 000 effective prevention is vaccination (see indicators on population). Greece and Italy followed with rates childhood vaccination in Chapter 6). higher than 8 per 100 000 population. An outbreak of measles started in 2016 in Romania and smaller outbreaks, amplified by low vaccination coverage, Definition and comparability stemmed partly from it in a few other countries. In Mandatory notification systems for most countries where vaccination coverage is high, communicable diseases, including measles and very few cases of measles were reported in 2017 (see hepatitis B, exist in most European countries, indicator on vaccination in Chapter 6). although case definitions, laboratory confirmation Vaccination against measles is very effective: the requirements and reporting systems may differ. vast majority of newly diagnosed people were not Measles and hepatitis B notification is mandatory vaccinated. Although 45% of measles cases occurred in all EU member states. Caution is required in among people aged 15 and older, most cases are interpreting the data because of the diversity in among infants under one year old, as they are often surveillance systems, case definitions and stilltooyoungtohavereceivedthefirstdoseof reporting practices (for example, several countries vaccine. Unvaccinated infants are generally protected only collect data on acute cases, not chronic against measles when at least 95% of population have cases). Variation between countries also likely received the second dose of vaccine (ECDC, 2018a). reflects differences in testing as well as differences Hepatitis B is a liver infection caused by a virus in immunisation and screening programmes. transmitted by contact with blood or body fluids of an infected person. People who are infected can go on to develop a chronic infection, especially those who are infected at younger ages. People with chronic References hepatitis B are more likely to suffer from liver cirrhosis and liver cancer. EC (2018), Proposal for a council recommendation on strengthened cooperation against vaccine preventable More than 29 300 hepatitis B cases were reported diseases, European Commission, Brussels. in EU/EEA countries in 2016 (ECDC, 2018b). This equals a rate of 6 cases of hepatitis B per 100 000 population ECDC (2018a), Measles outbreaks still ongoing in 2018 and fatalities reported from four countries, accessed across EU countries in 2016. Sweden, the United Kingdom 27 June 2018. and Latvia had the highest notification rates, with igure 3.25). more than 18 cases per 100 000 population (F Annual epidemiological report for 2016 ECDC (2018b), The rates are also high in Austria, Ireland, Iceland and , Stockholm. Hepatitis B HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 100

103 II.3. HEALTH STATUS 3.24. Notification rate of measles, 2017 (or nearest year) 0.9 0.2 0 0.4 0.1 0 0.1 0.1 Notification rate 0.5 per 100 000 pop. 0.4 0.1 0.2 0 - 0.5 1.1 3.2 0.5 - 5 1.4 0.7 5 + 0.1 1.1 0.4 0.4 0.8 0.3 28.4 0.2 n.a. n.a. 8.9 2.3 n.a. n.a. n.a. 0.3 0.3 9 n.a. 0 (Malta) 0.4 Note: Data refer to 2015 for Switzerland. Source: ECDC Surveillance Atlas of Infectious Diseases. http://dx.doi.org/10.1787/888933834718 1 2 3.25. Notification rate of hepatitis B, 2016 17.7 6.3 14.6 20.7 1.6 18.7 4.8 1.1 Notification rate 10.2 per 100 000 pop. 19.6 10 6.7 0 - 2 3.6 n.a 2 - 10 2.4 11.5 10 + 0.9 15 0.6 n.a. 0.5 0.1 2 1 n.a. n.a. n.a 1.1 n.a n.a. n.a. n.a. 1.1 1.5 0.2 1.8 7.6 (Malta) 0.0 The comparability of data is limited due to differences in surveillance and reporting system (many countries with low rates only Note: report acute cases, not chronic cases). Source: ECDC Surveillance Atlas of Infectious Diseases. 1 2 http://dx.doi.org/10.1787/888933834737 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 101

104 II.3. HEALTH STATUS NEW REPORTED CASES OF HIV AND TUBERCULOSIS The rates also declined sharply in Lithuania and Latvia. HIV remains a major public health issue in Europe, However, the rates have increased in Germany (from with more than 610 000 diagnosed people living with 5.2 to 7.2 per 100 000 population) and Sweden (6.6 to HIV infection in EU countries in 2016 (ECDC/WHO 7.4 per 100 000 population) between 2012 and 2016. Regional Office for Europe, 2017). In addition, another estimated 200 000 people are undiagnosed and unaware Among people with tuberculosis for whom that they are living with HIV infection (Pharris et al., information was available on HIV, about 4.5% were 2016). Nearly 30 000 people across EU countries were co-infected by HIV. newly diagnosed with HIV in 2016. This equals about six Antimicrobial resistance to tuberculosis threatens new cases of HIV infection per 100 000 population on effective treatment and control. On average, about 4% of average. Latvia had the highest rate of new cases cases of tuberculosis with drug susceptibility testing (18.5 per 100 000 population), followed by Estonia and were multi-drug resistant. These rates were much Malta. The lowest rates were in the Slovak Republic and higher in Lithuania, Estonia and Latvia than in other Hungary, with rates lower than 2.5 new cases per countries (ECDC/WHO Regional Office for Europe, 2018). 100 000 population (F igure 3.26). Despite progress, further efforts are needed to The number and rate of newly diagnosed HIV eliminate tuberculosis in EU countries in the coming cases have declined slightly overall over the past years. Countries can take a series of actions to reduce decade. However, the trend has evolved differently tuberculosis infections, including by addressing the across countries. In Estonia and Portugal, infection needs of vulnerable groups such as migrants, and by rates have decreased rapidly, although the infection optimising the prevention and care of drug-resistant rates remain above the EU average. In Latvia and Malta, tuberculosis (EC, 2018). infection rates have increased at least slightly since 2007 (ECDC/WHO Regional Office for Europe, 2017). Men are about three times more likely to be Definition and comparability diagnosed with HIV than women. About 40% of new HIV transmission is through men having sex with The rates of reported HIV are the number of men and 32% by heterosexual contact, while 4% of new cases per 100 000 population at year of new cases are through drug injection. Nearly 30% of diagnosis. Under-reporting and under-diagnosis new cases in 2016 were diagnosed at an advanced affect the reported rates, and may represent as stage of HIV and almost 50% had already been much as 40% of cases in some countries (ECDC/ infected for several years. People who were diagnosed WHO Regional Office for Europe, 2017). several years after being infected were more likely to A new reported case of tuberculosis is defined be older, infected by heterosexual sex or by drug as a patient in whom tuberculosis has been injection, and to be women. confirmed by bacteriology or diagnosed by a Sustained efforts are needed to reduce new HIV clinician. The rates are expressed per 100 000 infections through effective prevention campaigns, population (ECDC/WHO Regional Office for and more frequent HIV testing and education Europe, 2018). campaigns targeting high risk groups (EC, 2018). Tuberculosis is still an important public health issue in several EU countries, despite notable progress in most countries in reducing the number of cases References over the past few years. Nearly 59 000 new cases of tuberculosis were reported across EU countries in EC (2018), Commission Staff Working Document on 2016, down from about 70 000 cases in 2012. Romania Combatting HIV/AIDS, Viral Hepatitis and Tuberculosis in the European Union and Neighbouring Countries – had the highest rate of reported cases of tuberculosis State of play, policy instruments and good practices, in 2016, with 68.9 per 100 000 population, followed by European Commission, Brussels. Lithuania and Latvia, with rates above 30 per 100 000 population (Figure 3.27). Greece, Finland and the ECDC/WHO Regional Office for Europe (2018), Tuberculosis Czech Republic had the lowest rates, with rates below Surveillance and Monitoring in Europe 2018 – 2016 data, Stockholm. 5 per 100 000 population in 2016. Men are much more likely to be infected by tuberculosis than women in all ECDC/WHO Regional Office for Europe (2017), HIV/AIDS EU countries. Surveillance in Europe 2017 – 2016 data, Stockholm. Although the number and rate of tuberculosis Pharris et al. (2016), “Estimating HIV incidence and cases have decreased in nearly all countries since 2012, number of undiagnosed individuals living with HIV in the pace of decline has varied by country. In Romania, the European Union/European Economic Area, 2015”, the rate has decreased by more than 20% since 2012. Euro Surveillance Journal , Vol. 21, No. 48. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 102

105 II.3. HEALTH STATUS 3.26. New reported cases of HIV, 2016 8.4 3.3 4.2 4.4 17.4 18.5 4.3 New reported cases 7.4 per 100 000 pop. 10.5 < 3 7.9 4.4 3.3 3 - 5.9 4.2 8.1 6 - 9.9 2.7 11.5 > 10 1.6 2.9 2.3 6.4 7.8 2.8 3.2 2.6 2 5.7 2.8 5.4 1.4 4.4 6.8 5.7 10 3.1 9.4 14.5 (Malta) ECDC/WHO Regional Office for Europe (2017), HIV/AIDS Surveillance in Europe 2017. Source: http://dx.doi.org/10.1787/888933834756 1 2 3.27. New reported cases of tuberculosis, 2016 1.8 4.3 7.2 7.4 14.6 33.5 5.8 New reported cases 49.9 per 100 000 pop. < 6 6.7 9.4 6 - 7.9 17.0 5.2 7.2 8 - 14.9 9.3 > 15 4.9 5.0 5.5 7.3 8.0 7.2 7.4 5.7 68.9 11.0 16.9 6.6 22.4 13.8 12.8 14.2 10.5 17.8 4.1 15.6 7.1 11.5 (Malta) Source: ECDC/WHO Regional Office for Europe (2018), Tuberculosis Surveillance and Monitoring in Europe 2018. 1 2 http://dx.doi.org/10.1787/888933834775 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 103

106 II.3. HEALTH STATUS CANCER INCIDENCE In 2018, 3 million new cases of cancer are in breast cancer incidence across EU member states expected to be diagnosed in the 28 EU member states can be partly attributed to variation in the extent and (Joint Research Centre, 2018). Slightly more than half type of screening activities. Mortality rates from breast of these cancers (53% or around 1.6 million) are cancer have declined in most EU countries since the expected to be diagnosed in men. 1990s due to earlier detection and improvements in treatments, but still breast cancer continues to be one The most common cancer sites are breast cancer of the leading causes of cancer death among women (with more than 400 000 women expected to be (see indicator “Mortality from cancer” in this chapter diagnosed in 2018, accounting for 13.5% of all new and the indicator on “Screening, survival and mortality cancer cases), followed by prostate cancer (376 000 men from breast cancer” in Chapter 6). or 12.5% of new cancer cases), colon and rectum cancers (368 000 men and women for these two cancer Among men, prostate cancer is expected to sites combined or 12.3% of new cancer cases) and lung account for almost one quarter (23%) of all new cancers cancer (365 000 men and women or 12.2% of new diagnosed in 2018. The incidence of prostate cancer cases). These five cancers represent half of all the has increased in most European countries since the cancers that are expected to be diagnosed in EU late 1990s, partly because the greater use of prostate countries in 2018. Following these five cancers, the specific antigen (PSA) tests is leading to greater most common cancer sites are bladder cancer, skin detection. Lung cancer (14% of new cancer cases), and melanoma cancer, uterus cancer (corpus uteri and colon and rectum cancers (13%) also account for a large cervical), pancreas cancer and kidney cancer.These five number of new cancers detected in men. other cancers are expected to account for another 20% of all new cancer cases in the European Union in 2018 (Figure 3.28). Definition and comparability Large variations exist in cancer incidence across Cancer incidence rates are based on numbers EU countries. Hungary, Ireland, Denmark, Belgium of new cases of cancer registered in a country in and France are expected to have the highest age- a year divided by the population. Differences in standardised incidence rates in 2018 (all cancers the quality of cancer surveillance and reporting combined), with rates more than 10% higher than the across countries may affect the comparability of EU average (Figure 3.29). The incidence of lung cancer the data. Rates have been age-standardised and colon and rectal cancer is particularly high in based on the new European Standard Population Hungary (more than 50% higher than the EU average), to remove variations arising from differences in contributing largely to the overall high incidence rate. age structures across countries and over time. The high incidence of lung cancer is related to high The data come from the European Cancer smoking rates (see the indicator “Smoking among Information System (ECIS). The estimates for Chapter 4). adults” in 2018 may differ from national estimates due to These variations in incidence rates reflect not only differences in methods. variations in the real number of new cancers occurring The incidence of all cancers is classified to each year, but also differences in national policies ICD-10 codes C00-C97 (excluding non- regarding cancer screening to detect different types of melanoma skin cancer C44). cancer as early as possible (see indicators “Screening, survival and mortality from breast cancer and cervical cancer” in Chapter 6), as well as differences in the quality of cancer surveillance and reporting. Among women, breast cancer accounts for 29% of Reference all new cancers across EU countries. Colon and rectal Joint Research Centre (2018), Dataset Collection: European cancers (12% of cancer cases), lung cancer (10%) and Cancer Information System, https://ec.europa.eu/jrc/en/ uterus and cervical cancer (8%) are the next more publication/dataset-collection-european-cancer-information- common cancers diagnosed in women. The variation system . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 104

107 II.3. HEALTH STATUS 3.28. Estimated number of new cancer cases, all EU countries, 2018 Lung 364 601 Others Thyroid 243 722 58 346 Larynx, pharynx, etc Skin 150 436 120 505 Breast Respiratory 404 920 Non- 515 037 Hodgkin Others lymphoma 422 573 97 391 Leukae- Hematopoetic mia 172 005 All cancers 74 614 Bladder 3 001 956 164 450 Uterus and cervical Colon 111 595 242 987 Gastro- intestinal Reproductive 651 102 organs Kidney Ovary 572 655 Rectum 99 214 44 784 125 260 Stomach, Prostate Pancreas liver, anus, 375 842 100 005 etc Testis, vulva, etc 182 850 40 434 Note: Non-melanoma skin cancer is excluded. Source: JRC (European Cancer Information System). http://dx.doi.org/10.1787/888933834794 1 2 3.29. Estimated incidence rate for all cancers, by country, 2018 Age-standardised incidence rate per 100 000 population 800 677 638 685 662 630 668 626 615 599 594 591 588 583 576 569 566 564 556 552 547 538 539 534 547 600 515 514 510 501 495 492 470 433 448 410 428 400 200 0 All cancers are included except non-melanoma skin cancer. Numbers are age-standardised based on the European Standard Note: Population. JRC (European Cancer Information System). Source: http://dx.doi.org/10.1787/888933834813 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 105

108 II.3. HEALTH STATUS DIABETES PREVALENCE Diabetes is a chronic disease characterised by high may partly be due to a higher proportion of low- levels of glucose in the blood. It occurs either because educated people in older population groups and to the the pancreas stops producing the insulin hormone risk of diabetes increasing with age. But the prevalence (Type 1 diabetes), or because the cells of the body do of important risk factors for diabetes including obesity not respond properly to the insulin produced (Type 2 is much higher among people with a lower level of diabetes). People with diabetes are at greater risk of education (see the indicator “Obesity among adults” in developing cardiovascular diseases such as heart Chapter 4). attack and stroke if the disease is left undiagnosed or The economic burden of diabetes is substantial. poorly controlled. They also have higher risks of sight The health expenditure allocated to treat diabetes loss, foot and leg amputation, and renal failure. and prevent complications are estimated at about About 32.7 million adults were diabetics in the EUR 150 billion in 2017 in the European Union, with European Union in 2017, up from an estimated the average expenditure per diabetic adult estimated 18.2 million adults in 2000. In addition, some at about EUR 4 600 per year (IDF, 2017). 12.8 million people were estimated to have Type2diabetesislargelypreventable.Anumber undiagnosed diabetes in 2017.The number of men with of risk factors, such as overweight and obesity, diagnosed diabetes has increased particularly rapidly nutrition and physical inactivity, are modifiable. since 2000, more than doubling from around 8 million However, the prevalence of overweight and obesity is in 2000 to 17.1 million in 2017. But the number of increasing in most countries (see the indicator women with diabetes has also gone up substantially, “Overweight and obesity among adults” in Chapter 4). rising from 10.3 million in 2000 to 15.6 million in 2017, These reinforce the need for effective preventive an increase of over 50% (Figure 3.30). strategies. Diabetes is more common among older people: 19.3 million people aged 60-79 have diabetes across EU countries, compared with 11.7 million people aged Definition and comparability 40-59 and only 1.8 million aged 20-39 (Figure 3.31). The sources and methods used by the While more men than women have diabetes in International Diabetes Federation are outlined in middle-age (between 40 and 59 years old), a greater the Diabetes Atlas, 8th edition (IDF, 2017). The IDF number of women have diabetes after age 70 mainly produced estimations based on a variety of because they live longer. sources of which the majority was peer-reviewed The age-standardised prevalence rate of diabetes articles and national health surveys. In addition, among adults was 6% on average in EU countries in sources were only included if they met several 2017. The rates varied from 9% or more in Portugal, criteria for reliability. Age-standardised rates Romania and Malta to 4% or less in Ireland, Lithuania were calculated using the world population based igure 3.32). and Estonia (F on the distribution provided by the World Health Age-standardised rates of diabetes prevalence Organization. Adult population covers those aged have stabilised in many European countries in recent between 18 and 99 years old with Type 1 or Type 2 years, especially in Nordic countries, but they have diagnosed diabetes. gone up slightly in Southern Europe countries and in Central and Eastern Europe countries. These upward trends are partly due to the rise in obesity and physical inactivity, and their interactions with population References ageing (NCD Risk Factor Collaboration, 2016). Diabetes Atlas, 8th edition , International IDF (2017), Based on self-reported data on the prevalence of Diabetes Federation, Brussels. diabetes from the second wave of the European Health NCD Risk Factor Collaboration (2016), “Worldwide Trends Interview Survey conducted in 2014, adults with the in Diabetes Since 1980: A Pooled Analysis of 751 lowest level of education are more than twice as likely Population-based Studies with 4.4 Million Participants”, to report having diabetes than those with the highest http://dx.doi.org/ , Vol. 387, pp. 1513-1530, The Lancet level of education, on average across EU countries. This . 10.1016/S0140-6736(16)00618-8 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 106

109 II.3. HEALTH STATUS 3.30. Number of people with diabetes in EU28, 2000 and 2017 Women Men illion 2000: 18.3 m illion 2017: 32.7 m 8.0 17.1 10.3 15.6 million million million million (44%) (56%) (52%) (48%) Data include people aged 20-79 with Type 1 or Type 2 diagnosed diabetes. The number of people with diabetes in 2000 has been Note: estimated for some countries due to data gaps. Source: IDF Atlas, 8th Edition, 2017 and OECD estimates. 1 2 http://dx.doi.org/10.1787/888933834832 3.31. People with diabetes in EU28, by gender and age group, 2017 Men Women Million 3.0 2.5 2.0 1.5 1.0 0.5 0.0 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 25-29 20-24 30-34 Population with Type 1 or Type 2 diagnosed diabetes. Data are only available up to 79 years old. Note: Source: IDF Atlas, 8th Edition, 2017. 1 2 http://dx.doi.org/10.1787/888933834851 3.32. Share of adults with diabetes, 2017 % 14 11.9 12 9.9 9.9 9.9 9.9 9.9 9.5 9.0 8.9 10 8.4 7.6 7.5 7.2 7.1 8 6.7 6.5 6.5 6.0 5.8 5.8 5.8 5.6 5.5 5.4 5.4 4.9 4.9 4.8 6 4.6 4.6 4.5 4.4 4.4 4.2 4.0 3.7 3.4 4 2 0 Note: Age-standardised prevalence of population aged 18-99 with Type 1 or Type 2 diagnosed diabetes. IDF Atlas, 8th Edition, 2017. Source: 1 2 http://dx.doi.org/10.1787/888933834870 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 107

110 II.3. HEALTH STATUS DEMENTIA PREVALENCE Nonetheless, dementia will undoubtedly pose a Dementia describes a variety of brain disorders growing challenge to all EU countries. There has been a which progressively lead to brain damage and cause a renewed international focus on supporting countries gradual deterioration of the individual’s functional to improve the lives of people living with dementia, capacity and social relations. Alzheimer’s disease is the their families and carers. This includes an increased most common form of dementia, representing about focus on ensuring patients have access to a timely and 60% to 80% of cases. There is currently no cure or accurate diagnosis and adequate post-diagnostic disease-modifying treatment, but better policies can support. The growth of dementia-friendly initiatives improve the lives of people with dementia by helping across many EU countries – including training social them and their families adjust to living with the services, businesses, and volunteers to recognise signs condition and ensuring that they have access to high of dementia and respond appropriately – may help quality health and social care. reduce the stigma around the disease, particularly for In 2018, an estimated 9.1 million people aged over those living at home. Nevertheless, further efforts are 60 are living with dementia in EU member states, up also needed to improve care coordination to help from 5.9 million in 2000. If the age-specific prevalence patients and their families navigate complex health of dementia remains the same, ageing populations and social systems, to develop residential care models mean that this number will continue to grow adapted to the needs of people with dementia, and to substantially in the future. The overall number of improve the quality of care for people with dementia in people living with dementia in EU countries is hospitals and at the end of life (OECD, 2018). expected to rise by about 60% over the next two decades to reach 14.3 million in 2040, with the oldest people (those aged over 90) accounting for a growing share (Figure 3.33). Definition and comparability The prevalence of dementia increases rapidly with Estimates are taken from the World Alzheimer age. While only around 1% of people aged 60-64 have Report 2015, which includes a systematic review of dementia, this proportion goes up to nearly 40% among studies of dementia prevalence around the world those aged over 90 across EU countries. More women over the past few decades, assuming that age- thanmenalsolivewithdementiaatanyagegroup, specific prevalence has been constant over time. with the gap increasing at older ages (Figure 3.34). The prevalence by country has been estimated by Overall, around 7% of the population aged over 60 in applying the age-specific prevalence rates for each EU countries have dementia in 2018. This proportion is region of the world to the population structure expected to grow to over 8% by 2040 because of estimates from the United Nations (World population ageing. Countries that have high shares of Population Prospects: 2017 Revision). very elderly people now generally have a greater proportion of people with dementia. Italy, France, Greece and Spain have around 8% of their population aged over 60 living with dementia now, while this proportion References is only around 4% or less in Croatia, the Slovak Republic, Bulgaria, Romania, Poland, the Czech Republic and Matthews, F.E. et al. (2013), “A two-decade comparison of Hungary. Over the next two decades, the prevalence of prevalence of dementia in individuals aged 65 years dementia will rise particularly quickly in those countries and older from three geographical areas of England: where the share of people aged over 80 and 90 years will Results of the cognitive function and ageing Study I , Vol. 382, No. 9902. and II”, The Lancet igure 3.35). grow more rapidly (F However, there is some evidence that the age- Ngandu, T. et al. (2015), “A 2 year multidomain intervention of diet, exercise, cognitive training, and specific prevalence of dementia may be falling in some vascular risk monitoring versus control to prevent countries (Matthews et al., 2013) and it may be possible cognitive decline in at-risk elderly people (FINGER): A to reduce the risk of dementia through healthier http:// randomised controlled trial”, The Lancet , lifestyles and preventive interventions. A recent dx.doi.org/10.1016/S0140-6736(15)60461-5 . randomised-controlled trial of a multi-domain OECD (2018), Care Needed: Improving the Lives of People with intervention, including diet, physical exercise, and Dementia , OECD Health Policy Studies, OECD Publishing, cognitive training, found such lifestyle interventions to . https://doi.org/10.1787/9789264085107-en Paris, have a positive effect on cognition (Ngandu et al., 2015). If such efforts are successful, the rise in prevalence World Alzheimer Report 2015: The Global Prince, M. et al. (2015), may be less dramatic than these numbers suggest. , Alzheimer’s Disease International. Impact of Dementia HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 108

111 II.3. HEALTH STATUS 3.33. Estimated number of people with dementia in EU countries, by age group, 2000, 2018 and 2040 2000 2040 2018 Million 5 4 Total (60 to 90+): 3 14.3 million 2 Total (60 to 90+): 9.1 million 1 Total (60 to 90+): 5.9 million 0 85-89 80-84 65-69 70-74 90+ 60-64 75-79 OECD analysis of data covering 28 EU countries from the World Alzheimer Report 2015 and the United Nations. Source: http://dx.doi.org/10.1787/888933834889 1 2 3.34. Estimated prevalence of dementia among people aged 60 and over, by gender and age group, 2018 Women Men % among people aged 60 and over 50% 40% 30% 20% 10% 0% 60-64 90+ 75-79 85-89 70-74 65-69 80-84 OECD analysis of data covering 28 EU countries from the World Alzheimer Report 2015 and the United Nations. Source: http://dx.doi.org/10.1787/888933834908 1 2 3.35. Estimated prevalence of dementia among people aged 60 and over, 2018 and 2040 2040 2018 % among people aged 60 and over 12% 10% 8% 6% 4% 2% 0% OECD analysis of data from the World Alzheimer Report 2015 and the United Nations. Source: 1 2 http://dx.doi.org/10.1787/888933834927 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 109

112

113 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART II Chapter 4 Risk factors This chapter focuses mainly on modifiable risk factors to health among children and adults, including smoking, alcohol consumption and obesity. It ends with a new indicator of mortality related to environmental factors such as air pollution and extreme weather conditions. Recentestimatesindicatethatsome790000peopleinEUcountriesdiedprematurelyin2016becauseof tobacco smoking, alcohol consumption, unhealthy diets and lack of physical activity. Smoking among both children and adults has decreased in most EU countries, yet about one-fifth of adults still smoke every day, and the proportion still exceeds one in four adults in some countries that are lagging behind. Alcohol control policies have achieved progress in reducing overall alcohol consumption in several countries, with overall consumption dropping by over 10% over the past decade, but heavy alcohol consumption remains an issue among adolescents and adults. Nearly two out of five adolescent boys and girls report at least one “binge drinking” event in the past month, and more than two out of five young men aged 20-29 also report heavy episodic drinking across EU countries. The use of illicit drugs remains an important public health issue in Europe. While the use of some drugs has declined, cannabis remains frequently used among young people and the use of cocaine is on the rise in several countries. Obesity continues to spread among adults in most EU countries, while there are some signs of plateauing among children. Inequality in obesity remains marked: 12% of people with higher education level are obese compared to 20% of those with lower education level. Exposure to serious air pollutants is estimated to have caused the death of nearly 240 000 people across EU countries in 2016. Extreme weather conditions such as heat waves or cold waves are also becoming more frequent, and some episodes in the past have led to the deaths of many thousands of people, particularly among frail elderly people. These findings suggest that a much stronger focus on health promotion and disease prevention could help reduce the burden of many diseases and avoid a large number of premature deaths. 111

114 II.4. RISK FACTORS SMOKING AMONG CHILDREN Smoking in childhood and adolescence has both legislation, ban on the sale of e-cigarettes to children, immediate and long-term health consequences. The or the prohibition of proxy purchasing by adults on immediate adverse health consequences of smoking behalf of children. include addiction, reduced physical fitness and In addition to direct smoking, many children are endurance, and asthma, while early onset of smoking also exposed to second-hand smoking at home, at habits increase children’s long-term risk of school or in cars. Second-hand smoking also cardiovascular diseases, respiratory illnesses and increases greatly the risk of many respiratory diseases cancer. Children who smoke are also more likely to or other illnesses (WHO Europe, 2018). In response, experiment with alcohol and illicit drugs. many countries have taken measures to protect On average in EU countries, 25% of 15-16 year olds children from such second-hand smoking in public reported smoking in the past month in 2015 places but also in some cases by banning smoking in (Figure 4.1). More than 30% of them smoked in the past cars when children are present. month in Bulgaria, Croatia, Germany, Italy and the Slovak Republic, whereas less than 15% did so in Belgium (Flanders), Ireland, Malta and Sweden. Definition and comparability Smoking rates among 15-16 year olds have decreased The data refer to the proportion of children since 2007 in most EU countries, except in Poland and aged 15-16 year olds who report smoking in the Romania where they have increased. The largest past 30 days. The data come from the European decreases have occurred in Austria, Denmark, Ireland, School Survey Project on Alcohol and Other Drugs Latvia, Malta, and Sweden. (ESPAD). The ESPAD survey has been collecting The gap in smoking between 15-16 year old boys comparable data on smoking and other and girls is fairly small in most countries. On average, substance use among 15-16 year old students in a slightly greater proportion of 15-16 year old girls European countries every four years since 1995. reported smoking in 2015 (26% compared with 24% for Data for Spain (a non-ESPAD country) come boys). Smoking rates among 15-16 year olds have from the Spanish national school survey (2014-15). decreased since 1999, slightly more rapidly among Data from Latvia need to be interpreted with boys than girls (F igure 4.2). caution due to small sample size. A mix of policies including excise taxes to increase For more information, please see http://espad.org/ prices, clean indoor-air laws, restrictions on youth access . report/home/ to tobacco, and greater education about the effects of tobacco on health has contributed to reducing smoking rates among children and adolescents. In May 2016, the new Tobacco Products Directive became effective in all EU Member States. This directive particularly targets References adolescents and young adults, as 25% of 15-24 year olds Pötschke-Langer, M. (2016), “The Tobacco Products in the European Union are smokers (Pötschke-Langer, Health-EU Directive – Implementation in the EU”, 2016). It bans flavoured cigarettes, makes larger health Newsletter No. 174 – Focus , http://ec.europa.eu/health/ warnings (image and text) on packages mandatory, and newsletter/174/focus_newsletter_en.htm . introduces safety, quality and packaging regulations ESPAD Group (2016), ESPAD Report 2015: Results from the pertaining to e-cigarettes. European School Survey Project on Alcohol and Other Drugs , . EMCDDA and ESPAD, https://doi.org/10.2810/022073 Specific measures to reduce smoking among adolescents implemented in some countries include: WHO Europe (2018), Fact Sheet Tobacco and Oral Health, plain packaging of tobacco products, price hike, www.euro.who.int/__data/assets/pdf_file/0005/369653/ advertising restrictions, smoke-free environments . Fact-Sheet-on-Tobacco-and-Oral-Health-2018-eng.pdf?ua=1 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 112

115 II.4. RISK FACTORS 4.1. Changes in smoking rates among 15-16 year olds, 2007 to 2015 2015 2007 % reporting to smoke in the past 30 days 50 45 40 35 30 37 30 33 33 25 31 30 25 29 28 26 20 24 24 24 24 15 22 22 22 21 21 15 19 19 19 18 15 15 10 13 13 11 10 5 6 0 Note: The EU average is not weighted by country population size. The data for Belgium refers to the Flanders region only. Source: ESPAD, 2007 and 2015. 1 2 http://dx.doi.org/10.1787/888933834946 4.2. Gender gap in smoking rates among 15-16 year olds, average across EU countries and Norway, 1995 to 2015 % reporting to smoke in the past 30 days Boys Girls % reporting to smoke in the past 30 days 40 35 Boys 30 Girls 25 20 15 10 5 0 1995 1999 2003 2007 2011 2015 Source: ESPAD. 1 2 http://dx.doi.org/10.1787/888933834965 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 113

116 II.4. RISK FACTORS SMOKING AMONG ADULTS excise duties on cigarettes (European Commission, Tobacco consumption is the largest avoidable 2011), has contributed a lot to the success of tax health risk in the European Union and the most measures in EU members. In 2012, the large majority of significant cause of premature death, with over 300 000 EU countries (22 countries) were complying with the tax deaths per year according to IHME estimates (IHME, share minimum level recommended by WHO (tax share 2018). Around half of smokers die prematurely, dying representing more than 75% of the retail price of the 14 years earlier on average. It is a major risk factor for most popular brand of cigarettes) (WHO, 2014). at least two of the leading causes of mortality, circulatory diseases and cancer, and an important risk factor for many serious respiratory diseases. The proportion of adults who smoke daily varies Definition and comparability more than two-fold across EU countries (Figure 4.3). It The proportion of daily smokers is defined as the is the lowest in Nordic countries (Sweden, Finland, as percentage of the population aged 15 years and well as Iceland and Norway) and the highest in over who report tobacco smoking every day. Other Bulgaria, Greece, Hungary and Cyprus. On average, the forms of smokeless tobacco products, such as snuff proportion of adults smoking daily has decreased from in Sweden, are not taken into account. The 24% in 2006 to 20% in 2016, with large reductions in comparability of data is limited to some extent due Nordic countries, the Netherlands, Latvia and Greece. to the lack of standardisation in the measurement Men smoke more than women in all European of smoking habits in health interview surveys countries, except in Sweden and Iceland where the across EU Member States. Variations remain in the rate is virtually equal (F four men and igure 4.4). One in age groups surveyed, wording of questions, one in six women smoke daily on average in EU response categories and survey methodologies. countries. The gender gap is particularly large in Cyprus, Latvia, Lithuania and Romania. The Eurobarometer survey reports higher smoking rates among both men and women as it References includes people smoking daily or occasionally. The Tobacco Products Directive , European Commission (2014), results from the latest Eurobarometer survey http://ec.europa.eu/ European Commission, Brussels, conducted in 2017 indicate that 30% of men and 22% . health/tobacco/products/index_en.htm of women are daily or occasional smokers on average across EU countries (TNS Opinion & Social, 2017). European Commission (2011), Directive on the structure and , rates of excise duty applied to manufactured tobacco Accordingtoatobaccocontrolscalefromthe European Commission, Brussels, https://ec.europa.eu/ Association of European Cancer Leagues, the taxation_customs/business/excise-duties-alcohol-tobacco- United Kingdom, Ireland, Iceland, France and Norway energy/excise-duties-tobacco_en . are the top five European countries with the most comprehensive tobacco control policies in 2016 Feliu et al. (2018), Impact of tobacco control policies on (Joossens and Raw, 2017). Countries that have stricter smoking prevalence and quit ratios in 27 European Union tobacco control policies generally have higher Tobacco Control, countries from 2006 to 2014, 10:1-9. reductions in smoking rates and higher quit ratios http://ghdx. IHME (2018), Global Health Data Exchange, (Feliu et al., 2018), although there are exceptions. . healthdata.org/gbd-results-tool The EU Tobacco Products Directive (2014/40/EU), Joossens, L. and M. Raw (2017), The Tobacco Control Scale adopted in February 2014, requires that health warnings 2016 in Europe, Association of European Cancer appear on packages of tobacco and related products, Leagues, www.tobaccocontrolscale.org/wp-content/uploads/ bans all promotional and misleading elements on . 2017/03/TCS-2016-in-Europe-COMPLETE-LoRes.pdf tobacco products, and sets out safety and quality TNS Opinion & Social (2017), Special Eurobarometer 458 – requirements for electronic cigarettes (European Attitudes of Europeans towards tobacco and electronic Commission, 2014). One important step forward in cigarettes, http://ec.europa.eu/commfrontoffice/public health warnings is plain packaging for tobacco products opinion/index.cfm/Survey/getSurveyDetail/instruments/ aiming to restrict branding. Following the lead from . SPECIAL/surveyKy/2146 Australia, plain packaging has been adopted by an increasing number of European countries (e.g. France, WHO (2014), “Tobacco Taxation” Fact Sheet, WHO Regional Hungary, Ireland, the United Kingdom, Norway, and will www.euro.who.int/__ Office for Europe, Copenhagen, be implemented in Slovenia in 2020). data/assets/pdf_file/0007/250738/140379_Fact-sheet- . Tobacco-Taxation-Eng-ver2.pdf Among tobacco control measures, rising taxes on tobacco is the most effective way to reduce tobacco use WHO report on the global tobacco epidemic, WHO (2017), and to encourage users to quit (WHO, 2017). The EU , 2017: Monitoring tobacco use and prevention policies Directive on excise duty on tobacco (2011/64/EU), which http://apps.who.int/iris/bitstream/handle/ WHO, Geneva, requires Member States to levy a minimum rate of 10665/255874/9789241512824-eng.pdf?sequence=1 . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 114

117 II.4. RISK FACTORS 4.3. Changes in daily smoking rates among adults, 2006 and 2016 (or latest year) 2006 2016 ortin p % re g to smoke dail y 45 40 35 30 38 25 31 28 27 29 26 26 25 25 24 20 27 23 23 22 23 21 21 20 20 20 20 20 15 20 19 19 18 18 18 17 17 16 16 15 10 11 11 10 5 0 OECD Health Statistics 2018, (based on national health interview surveys), complemented https://doi.org/10.1787/health-data-en Source: with Eurostat (EHIS 2014) for Bulgaria, Croatia, Cyprus, Malta, and Romania, and with WHO Europe Health for All database for Albania, Serbia and Montenegro. 1 2 http://dx.doi.org/10.1787/888933834984 4.4. Gender gap in daily smoking rates among adults, 2016 (or latest year) Men Women % reporting to smoke daily 45 40 35 30 25 20 15 10 5 0 Source: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en (based on national health interview surveys) complemented with Eurostat (EHIS 2014) for Bulgaria, Croatia, Cyprus, Malta, and Romania, and with WHO Europe Health for All database for Albania, Serbia and Montenegro. 1 2 http://dx.doi.org/10.1787/888933835003 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 115

118 II.4. RISK FACTORS ALCOHOL CONSUMPTION AMONG CHILDREN Alcohol use in adolescence continues to be very Approximately a third of European adolescents common in Europe, with beer being by far the most report negative experiences while under the influence popular alcoholic beverage, even though the of alcohol. These include accidents or injuries (9% of percentage of 15-16 year olds reporting heavy episodic boys and girls) and unprotected sex (8% of boys and drinking has come down at least slightly in recent 5% of girls). years in several countries (ESPAD, 2016). A number of policies have proven to be effective to Two adolescent drinking patterns are specifically reduce alcohol drinking among adolescents, such as linked to negative health, education and social limiting accessibility (e.g. through restrictions on outcomes – early initiation of alcohol consumption and location and hours of sales, and raising the minimum binge drinking. About half of European adolescents age to drink alcohol), increase prices, and advertising started drinking alcohol at the age of 13 or even regulations. In January 2018, Lithuania, which has one of younger, and almost 10% have been drunk at least once the highest level of alcohol consumption among by the age of 13 (ESPAD, 2016). Children who report adolescents based on another children and adolescent early initiation to alcohol and having been drunk on survey (Inchley et al., 2016), introduced a new legislation several occasions are more likely to develop alcohol on alcohol control particularly targeting young people. dependence later in life (Spear, 2015). This legislation raised the legal drinking age from 18 to 20, restricted opening hours for sales in retail stores, and By age 15-16, over 80% of adolescents report banned all advertising for beers, wines and spirits. having tried alcohol at least once in their life, and half say that they have consumed alcohol in the past month (ESPAD, 2016). More than two-thirds of 15-16 year olds in Denmark, Austria, Cyprus, the Definition and comparability Czech Republic and Greece report having consumed Heavy episodic drinking is defined as drinking alcohol over the past month, compared with less than five or more drinks in a single occasion in the one-third in Sweden and Finland. Frequent alcohol past 30 days. National examples are given so that use is linked to how easy it is for adolescents to a “drink” is understood to contain roughly the purchase alcohol. More than 90% of adolescents in equal amount of pure alcohol as a glass of wine. countries where they report drinking regularly say The data source is the European School Survey that it is easy to obtain alcohol. Project on Alcohol and Other Drugs (ESPAD). The Heavy episodic drinking (also known as “binge ESPAD survey has been collecting comparable drinking”) is a frequent behaviour among many data on alcohol use and other substance use European adolescents – 38% of 15-16 year old boys among 15-16 year old students in European and girls reported at least one binge drinking session countries every four years since 1995. in the past month on average. Binge drinking is For more information, please see http://espad. particularly popular among adolescents in Denmark org/report/home/ . and Cyprus, with half of 15-16 year olds reporting heavy drinking in the past month. This proportion was much lower in Portugal, Norway and Iceland (Figure 4.5) References On the positive side, the proportion of adolescents who report regular binge drinking has decreased ESPAD Group (2016), ESPAD Report 2015: Results from the , European School Survey Project on Alcohol and Other Drugs significantly from 2011 to 2015 in most countries, while EMCDDA and ESPAD, https://doi.org/10.2810/022073 . it has remained stable in several other countries. This proportion has increased significantly in only two Inchley, J. et al. (eds.) (2016), “Growing Up Unequal: Gender and Socioeconomic Differences inYoung People’s Health countries (Cyprus and Montenegro). Health Behaviour in School-aged Children and Well-being”, In most countries, binge drinking is slightly more (HBSC) Study: International Report from the 2013/2014 frequent among boys than girls, although the gap has , WHO Regional Office for Europe, Copenhagen, Survey narrowed recently (Figure 4.6). In 2015, 39% of www.euro.who.int/__data/assets/pdf_file/0003/303438/ 15-16 year old boys reported heavy alcohol drinking in HSBC-No.7-Growing-up-unequal-Full-Report.pdf . the past month compared with 35% of girls. This Spear, L.P. (2015), “Adolescent alcohol exposure: Are there gender gap remains particularly large in Romania, separable vulnerable periods within adolescence?”, Cyprus, the Czech Republic and Greece. 148, pp. 122-130. Physiology & Behavior, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 116

119 II.4. RISK FACTORS 4.5. Changes in the proportion of heavy episodic drinking in the past 30 days among 15-16 year olds, 2011 to 2015 Significant decrease Significant increase No change % 60 Denmark Cyprus 50 Croatia Bulgaria Malta Slovenia Slovak Rep. Czech Rep. Hungary 40 EU21 Greece Romania Estonia Belgium Italy Poland 2015 Montenegro France 30 Ireland Sweden Finland Albania 20 Portugal Norway 10 Iceland % 0 05 0 1 03 02 0 06 04 2011 The grey line represents “no change” between 2011 and 2015. Decreases of 3 or more percentage points between successive surveys Note: are indicated with a square, increases of 3 or more percentage points with a triangle, and unchanged situations with a losange (less than ± 3 percentage points). Source: ESPAD 2011 and 2015. The data for Belgium refers to the Flanders region only. http://dx.doi.org/10.1787/888933835022 1 2 4.6. Changes in heavy episodic drinking in the past 30 days among 15-16 year old boys and girls, average across EU countries and Norway, 1995 to 2015 % Girls Boys % 60 Boys 50 40 Girls 30 20 10 0 1995 2015 2011 1999 2003 2007 The average is not weighted by country population size. Note: Source: ESPAD. 1 2 http://dx.doi.org/10.1787/888933835041 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 117

120 II.4. RISK FACTORS ALCOHOL CONSUMPTION AMONG ADULTS alcohol in public places (e.g. by requiring stores to Alcohol-related harm is a major public health reduce the public display of alcoholic products) and to concern in the European Union, both in terms of change advertising practices so as to avoid linking morbidity and mortality. Alcohol was the third leading alcohol consumption with good times such as risk factor for disease and mortality after tobacco and holidays. In addition, alcohol advertisement by high blood pressure in Europe in 2012, and accounted alcohol operators will be prohibited from social for an estimated 7.6% of all men’s deaths and 4.0% of media, except on their own websites. all women’s deaths (WHO, 2014). High alcohol intake is associated with increased risk of heart diseases and All EU countries have set maximum levels of blood stroke, as well as liver cirrhosis and certain cancers, alcohol concentration for drivers in their legislation, but but even moderate alcohol consumption increases the these regulations are not always enforced rigorously. long-term risk of developing such diseases. Foetal Less stringent policies include health promotion exposure to alcohol increases the risk of birth defects messages, school-based and worksite interventions, and intellectual impairments. Alcohol also contributes and greater counselling by family doctors or other to death and disability through accidents and injuries, primary care providers. Comprehensive policy packages assault, violence, homicide, and suicide, particularly including fiscal measures, regulations and less stringent among young people. policies are shown to be the most effective to reduce harmful use of alcohol (OECD, 2015). Measured through sales data, overall alcohol consumption stood at 9.8 litres of pure alcohol per adult on average across EU Member States in 2016, down from 11 litres in 2006 (Figure 4.7). Lithuania Definition and comparability reported the highest consumption of alcohol, with 13.2 litres per adult, followed by France, the Overall alcohol consumption is defined as Czech Republic, Bulgaria, Austria, Luxembourg, annual sales of pure alcohol in litres per person Ireland, and Latvia with more than 11 litres per adult. aged 15 years and over. The methodology to At the other end of the scale, Greece, Italy and Sweden convert alcohol drinks to pure alcohol may differ have relatively low levels of consumption, below across countries. Official statistics do not include 8 litres of pure alcohol per adult. unrecorded alcohol consumption, such as home production. In some countries (e.g. Luxembourg), Although overall alcohol consumption per capita national sales do not accurately reflect actual is a useful measure to assess long-term trends, it does consumption by residents, since purchases by not identify sub-populations at risk from harmful non-residents may create a significant gap drinking patterns. Heavy episodic drinking, also between national sales and consumption. known as “binge drinking”, is more common among Alcohol consumption in Luxembourg is thus men, and particularly young men aged 20-29. More estimated as the average alcohol consumption in than 40% of men aged 20-29 report heavy episodic France and Germany. drinking on average across the EU. Nonetheless, a sizeable proportion of both men and women at older Regular heavy episodic drinking (or binge igure 4.8). ages also report regular heavy drinking (F drinking) is derived from self-reported information as part of the European Health A number of countries have taken initiatives to Interview Survey in 2014. Regular binge drinking limit harmful use of alcohol in recent years. Some is defined as having 60 grams of pure alcohol per interventions target heavy drinkers only, other target single occasion at least once a month over the young drinkers (e.g. sales restrictions to young people past 12 months. The figures represent the below a certain age, tighter alcohol consumption limit proportion of adults who reported binge drinking for young drivers), while others are more broadly at least once a month over the past 12 months as based. In 2018, Scotland introduced minimum pricing a percentage of adults who drank alcohol in the per unit of alcohol at 50 pence, which set a minimum past 12 months. price of GBP 1 for a 500 ml can of beer and GBP 4.69 for a bottle of wine. Wales also plans to introduce such a minimum price for alcohol with the Public Health (Minimum Price for Alcohol) (Wales) Act. Minimum pricing is likely to reduce the consumption of cheap References alcohol, in particular in harmful patterns such as binge drinking and alcohol-dependent use. Tackling Harmful Alcohol Use: Economics and OECD (2015), Public Health Policy, http:// OECD Publishing, Paris, Regulations on advertising alcoholic products dx.doi.org/10.1787/9789264181069-en . havetakendifferentformsondifferentmedia(e.g. printed newspapers, billboards, the internet). For , WHO (2014), Global Status Report on Alcohol and Health 2014 example, Estonia recently passed a law restricting WHO, Geneva, http://apps.who.int/iris/bitstream/handle/ alcohol sales and alcohol marketing practices. This 10665/112736/9789240692763_eng.pdf;jsessionid=3F94D new law specifically aims to restrict the visibility of . 8649E27F3B91FBC499D73207030?sequence=1 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 118

121 II.4. RISK FACTORS 4.7. Overall alcohol consumption among adults, 2006 and 2016 (or nearest years) 2016 2006 Litres per adult per year 16 14 12 13.2 10 11.7 11.7 11.5 11.4 11.3 11.2 11.2 10.9 10.8 10.7 10.5 10.4 10.4 10.4 9.9 9.9 8 10.3 9.8 9.6 9.5 9.4 9.3 9.2 8.6 8.4 8.3 8.0 6 7.5 7.2 7.1 6.5 6.4 6.0 4 5.1 4.8 1.3 2 0 ; Global Information System on Alcohol and Health for non- OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en Source: OECD countries and Austria, Belgium, Germany, Greece, Italy, Latvia, Luxembourg, Portugal, and Spain. 1 2 http://dx.doi.org/10.1787/888933835060 4.8. Gender gap in regular heavy episodic drinking by age, EU average, 2014 % 45 40 Men 35 30 Women 25 20 15 10 5 0 15-19 20-29 30-39 40-49 50-64 65+ Age group The EU average is not weighted by country population size. Note: Eurostat, EHIS 2014. Source: 1 2 http://dx.doi.org/10.1787/888933835079 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 119

122 II.4. RISK FACTORS ILLICIT DRUG CONSUMPTION AMONG CHILDREN Adolescence is a period of experimentation The overall trend in illicit drug use among sometimes linked to engagement in risky behaviour, 15-16 year olds seems fairly stable over the past including the use of illicit substances such as decade in the case of cannabis or even showing a cannabis being by far the most “popular” drug. slight decrease in the case of other illicit drugs, Frequent and heavy cannabis use during adolescence igure 4.11 and following a large increase in the 1990s (F is linked to an increased risk of dependence and Figure 4.12). In recent years, the gender gap in the use neurobiological problems (WHO, 2016). of different types of illicit drugs between boys and girls has narrowed slightly. Close to one in five 15-16 year olds (16%) in EU countries report having consumed cannabis at least once during their lifetime, and 7% say that they have consumed cannabis in the past month. The proportion Definition and comparability of 15-16 year olds reporting to have consumed cannabis The use of illicit drugs other than cannabis the past month is highest in France (17%) and Italy (15%), includes use of amphetamines, cocaine, crack, and the lowest in Finland and Sweden (2% only) ecstasy, LSD or other hallucinogens, heroin and (Figure 4.9). In all countries, boys are more likely than GHB. girls to report having consumed cannabis in the past The data source is the European School Survey month, although the gap is almost nil in some countries Project on Alcohol and Other Drugs (ESPAD). The (e.g. Malta, Portugal and the Slovak Republic). About 1% ESPAD survey repeatedly collects comparable data of 15-16 year olds consume cannabis almost every day. on the use of illicit drugs and other substance The lifetime use of at least one illicit drug other among 15-16 year olds students in European than cannabis at age 15-16 is 6% on average across EU countries. The ESPAD survey data have been highest rates are observed igure 4.10). The countries (F collected every four years since 1995. in Bulgaria and Poland, while the lowest rates are in Data for Spain (a non-ESPAD country) come Denmark, Finland and Sweden. Boys are more likely from the Spanish national school survey than girls to report having consumed illicit drug other (2014-15), only including some indicators where than cannabis in a majority of countries. At least once comparability is high. Data from Latvia need to in a lifetime consumption of ecstasy, amphetamines, be interpreted with caution due to low sample cocaine, LSD and other hallucinogens are reported on size. average by 2% of 15-16 year olds. Crack and heroin use is less common, with only about 1% of 15-16 year olds http://espad. For more information, please see reporting use at least once during their life. org/report/home/ . The use of new psychoactive substances is an important concern in many European countries and has been identified as a priority for monitoring under Early Warning Systems. About 4% of adolescents aged References 15-16 years old report to have used such new Connor, J.P. et al. (2014), “Polysubstance use: Diagnostic psychoactive substances at least once during their challenges, patterns of use and health”, Current lifetime across EU countries, with the proportion opinion in psychiatry, 27(4), pp. 269-275. being the highest in Poland and Estonia (10%). ESPAD Group (2016), ESPAD Report 2015: Results from the The use of illegal drugs together with alcohol and European School Survey Project on Alcohol and Other other substances increases the risks of accidents and Drugs, EMCDDA and ESPAD, European Monitoring injuries for adolescents and mental health problems Centre on Drugs and Drug Addiction, Lisbon, https:// doi.org/10.2810/022073 . later in life (Connor et al., 2014). The vast majority of 15-16 year olds in EU countries (more than 90%) who WHO (2016), The health and social effects of nonmedical ever smoked cannabis have also consumed alcohol www.who.int/ cannabis use, World Health Organization, and tobacco. . substance_abuse/publications/msbcannabis.pdf HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 120

123 II.4. RISK FACTORS 4.9. Prevalence of cannabis use in the last 30 days among 15-16 year olds, 2015 Total Boys Girls % 20 18 16 14 12 10 8 6 4 5 8 17 15 14 13 12 12 12 10 9 9 9 8 8 8 8 2 5 5 43 4 4 4 4 3 2 222 2 0 Note: The EU average is not weighted by country population size. The data for Belgium refers to the Flanders region only. ESPAD, 2015. Spanish national school survey 2014-15 for Spain. Source: http://dx.doi.org/10.1787/888933835098 1 2 4.10. At least once in a lifetime use of illicit drugs other than cannabis among 15-16 year olds, 2015 Girls Total Boys % 14 12 10 8 6 4 2 66432 118777776666665555555544332 0 The EU average is not weighted by country population size. The data for Belgium refers to the Flanders region only. Note: ESPAD, 2015. Spanish national school survey 2014-15 for Spain. Source: http://dx.doi.org/10.1787/888933835117 1 2 4.11. Lifetime use of cannabis among 4.12. Lifetime use of illicit drugs other than 15-16 year olds, average across EU countries cannabis among 15-16 year olds, average across EU countries and Norway, 1995 to 2015 and Norway, 1995 to 2015 Boys Girls Boys Girls % % 9 25 8 20 7 6 Boys Boys 15 5 4 10 3 Girls 2 Girls 5 1 0 0 2015 2011 2003 1999 1995 2007 2011 2007 2003 1999 1995 2015 The average is not weighted by country population size. Note: Note: The average is not weighted by country population size. ESPAD. ESPAD. Source: Source: http://dx.doi.org/10.1787/888933835155 1 2 1 2 http://dx.doi.org/10.1787/888933835136 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 121

124 II.4. RISK FACTORS ILLICIT DRUG CONSUMPTION AMONG ADULTS The use of illicit drugs remains an important from 0.2% in Italy and Norway to 1.7% in Romania. public health issue in Europe. Over a quarter of adults While consumption levels of new psychoactive in the European Union aged 15-64, or over 92 million substances are low overall in Europe, over two-thirds of people, have used illicit drugs at some point in their countries report their use by high-risk drug users. In lives. In most cases, they have used cannabis, but some particular, the use of synthetic cathinones by opioid have also used cocaine, amphetamines, ecstasy and and stimulant injectors has been linked to serious other drugs (EMCDDA, 2018). The use of illicit drugs, health and social problems (EMCDDA, 2017). particularly among people who use them regularly, is The consumption of opioids (i.e. heroin and other associated with higher risks of cardiovascular diseases, drugs) is responsible for the majority of drug overdose mental health problems, accidents, as well as deaths (reported in about 80% of fatal overdoses). The infectious diseases such as HIV. Illicit drug use is a main opioid used in Europe is heroin, but there are major cause of mortality among young adults in concerns in several countries about the increasing use Europe, both directly through overdose and indirectly of other synthetic opioids (such as buprenorphine, through drug-related diseases, accidents, violence and methadone, fentanyl and tramadol). The prevalence of suicide. high-risk opioid use among adults aged 15-64 is Cannabis is the illicit drug most used among estimated at 0.4% of the EU population; this was young adults in Europe. Over 14% of people aged 15 equivalent to 1.3 million high-risk opioid users in 2016. to 34 in EU countries report having used cannabis in the last year (Figure 4.13). This proportion is the highest in France and Italy (20% or more). Cannabis use has Definition and comparability increased over the past decade in some Nordic Data on drug use prevalence come from national countries which initially had low levels (Denmark and population surveys, as gathered by the European Finland), and are now converging towards the Monitoring Centre for Drugs and Drug European average. Among those countries with above- Addiction (EMCDDA). The data presented in this average use of cannabis, the decreasing trends section focus on the percentage of young adults previously observed in Spain have now stabilised, aged 15 to 34 years old reporting to have used while France reported a marked increase in recent different types of illicit drugs in the last year. Such years. estimates of recent drug use produce lower figures Cocaine is the most commonly used illicit than “lifetime experience”, but better reflect the stimulant in Europe: around 2% of young adults current situation. The information is based on the reported having used cocaine in the last year latest survey available for each country. The study (Figure 4.14). This percentage is highest in Denmark, years range from 2008 to 2017. To obtain estimates the Netherlands, Spain and the United Kingdom of the overall number of users in Europe, the EU (3% or more). After years of reported decreases in cocaine average is applied to those countries with missing use, there are now signs of stabilisation and possible data. increase in some countries. www.emcdda. For more information, please see: The use of amphetamines and ecstasy (or MDMA) europa.eu/data/stats2018_en . is slightly lower than the use of cocaine, with about 1% of young adults in EU countries reporting to have used amphetamines and 1.8% ecstasy (or MDMA) in the last year. The use of amphetamines tends to be higher in References some Nordic and Baltic countries (Estonia and Finland) and in Croatia, Germany and the Netherlands. The use European Monitoring Centre for Drugs and Drug Addiction of ecstasy is highest in Bulgaria, the Czech Republic, (2017), High-risk drug use and new psychoactive substances, EMCDDA Rapid Communication, Ireland and the Netherlands (EMCDDA, 2018). Over the Publications Office of the European Union, Luxembourg, last decade, the use of amphetamines has remained www.emcdda.europa.eu/system/files/publications/4540/ relatively stable in most European countries. In many TD0217575ENN.pdf . countries, the use of ecstasy declined after reaching a European Monitoring Centre for Drugs and Drug Addiction peak in the early and mid-2000s, but recent surveys (2018), European Drug Report 2018: Trends and point to increased use in some countries. Developments, Publications Office of the European The prevalence of use of new psychoactive www.emcdda.europa.eu/publications/ Union, Luxembourg, substances among young people in the last year ranges . edr/trends-developments/2018_en HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 122

125 II.4. RISK FACTORS 4.13. Cannabis use over the last 12 months among people aged 15 to 34, 2017 (or nearest year) n.a. Percent < 5.1 5.1 - 10 10.1 - 15 > 15 n.a. n.a. n.a. n.a. n.a. n.a. n.a. (Malta) EMCDDA, 2018. Source: http://dx.doi.org/10.1787/888933835174 1 2 4.14. Cocaine use over the last 12 months among people aged 15 to 34, 2017 (or nearest year) n.a. Percent < 0.6 0.6 - 1 1.1 - 2.5 > 2.5 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. (Malta) Source: EMCDDA, 2018. 1 2 http://dx.doi.org/10.1787/888933835193 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 123

126 II.4. RISK FACTORS OBESITY AMONG CHILDREN public procurement based on nutritional food quality Children who are overweight or obese are at a standards (European Commission, 2017). greater risk of poor health in adolescence as well as in adulthood. Obesity among children is also often related In the area of food marketing, the revised to psychosocial problems such as poor self-esteem, Audiovisual Media Services Directive allows the bullying at school, underachievement at school, eating Commission and the Member States to continue disorders, and depression, leading to health and working together with stakeholders to develop economic problems in adulthood. voluntary codes of conduct to reduce the exposure of children to aggressive marketing of foods high in fat, Nearly one in eight children aged 7-8 is obese on sugar or salt (European Commission, 2018). In 2018, a average in EU countries (Figure 4.15) (WHO Europe, project started to measure children’s exposure to food 2018). Cyprus, Italy, Greece, Malta and Spain show the marketing especially in the digital sphere. highest obesity rates in 7-8 year olds. The lowest child obesity rates are in the Czech Republic, Denmark, France, Ireland and Latvia. The obesity rate among children aged 7-8 has in fact shown signs of decrease in Definition and comparability several EU countries between 2007-08 and 2015-17. Estimates of obesity are based on body mass This decrease has been particularly strong in Greece, index (BMI) calculations using measured height Italy, Portugal and Slovenia, although child obesity and weight. Obese children are defined as those rates in Greece and Italy still remain relatively high. with a BMI above the WHO age- and sex-specific On average across 23 EU countries, 14% of boys and cut-off points (de Onis et al, 2017). 10% of girls aged 7-8 year olds are obese, according to Measured data on height and weight are theCOSIstudy(Figure4.16).Boystendtocarryexcess collected by the WHO Childhood Obesity weight more often than girls, with the largest gender Surveillance Initiative (COSI), which has monitored differences observed in Austria, Italy, Greece and trends in overweight and obesity among primary- Romania (about 6-7 percentage points). In particular, school-aged children for over 10 years (WHO, 2018). more than one in five boys is obese in Cyprus, Greece, Data refer to children aged 7 years old in most and Italy. countries, except in Albania, Austria, Croatia, The WHO European Food and Nutrition Action France, Italy, Norway, Poland, Romania and Plan 2016-2020 was adopted by the WHO Regional Sweden, where children are aged 8. Committee for Europe in 2014. Specific policy options in this action plan include stronger restrictions on the marketing of foods high in saturated fat, sugars and salt to children, the promotion of better labelling on the References front of food packages, and strict standards for the foods available in schools. Using a life-course approach, De Onis et al. (2007), “Development of a WHO growth the actions range from the protection and promotion of reference for school-aged children and adolescents”, exclusive breastfeeding, to the improvement of the Bulletin of the World Health Organization. baby food market landscape, to the increase of intake of JANPA (2017), Joint Action on Nutrition and Physical Activity fruit and vegetable (WHO Europe, 2017). website, . www.janpa.eu/outcomes/outcomes.asp The EU Action Plan on Childhood Obesity 2014-20 European Commission (2014), EU Action Plan on Childhood aims to halt the rise in overweight and obesity in children Obesity 2014-2020, Brussels, https://ec.europa.eu/health/ and young people by 2020. It is based on several key areas sites/health/files/nutrition_physical_activity/docs/ for action, including the support of a healthy start in life . childhoodobesity_actionplan_2014_2020_en.pdf and promoting healthier environments, especially in European Commission (2017), “Childhood Obesity: Halting schools and pre-schools (e.g. limiting exposure to less the Rise”, Updated on 22 February 2017, https:// healthy food options and ensuring access to free drinking ec.europa.eu/jrc/en/news/childhood-obesity-halting-rise . water) (European Commission, 2014). A mid-term evaluation report on its implementation will be delivered European Commission (2018), Digital Single Market: in the second half of 2018. Updated audiovisual rules, 7 June 2018, http://europa.eu/ rapid/press-release_MEMO-18-4093_en.htm . The Joint Action on Nutrition and Physical Activity (JANPA), run from 2015 to 2017, was a direct contributor to WHO Europe (2018), Children Obesity Surveillance this action plan, notably by using the economic Initiative, Highlights 2015-17, Preliminary data, evaluation of the cost of obesity to encourage public www.euro.who.int/__data/assets/pdf_file/0006/372426/ actions, and by identifying multilevel, multi-sectorial and wh14-cosi-factsheets-eng.pdf?ua=1 . life-course approaches for preventing obesity, sedentary WHO Europe (2017), European Food and Nutrition Action lifestyle and unhealthy nutrition (JANPA, 2017). Plan 2015-2020, WHO Regional Office for Europe, Another focus of action is through improving the www.euro.who.int/__data/assets/pdf_file/ Copenhagen, availability of healthy food in schools through better . 0008/253727/64wd14e_FoodNutAP_140426.pdf HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 124

127 II.4. RISK FACTORS 4.15. Changes in obesity rates among children aged 7-8 years old, 2007-08 (or nearest year) and 2015-17 2015-17 2007-08 % of 7-8 years old children 25 20 20 18 18 14 15 17 17 15 14 10 13 13 13 10 12 12 12 11 11 11 10 10 999 8 888 5 7 6 5 0 Note: The EU average is not weighted by country population size. WHO-Europe (Children Obesity Surveillance Initiative). Source: http://dx.doi.org/10.1787/888933835212 1 2 4.16. Obesity among children aged 7-8 years old, by gender, 2015-17 Girls Boys % of 7-8 years old children 25 20 15 10 5 0 The EU average is not weighted by country population size. Note: Source: WHO-Europe (Children Obesity Surveillance Initiative). 1 2 http://dx.doi.org/10.1787/888933835231 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 125

128 II.4. RISK FACTORS OBESITY AMONG ADULTS encourage cycling and worksite wellness programmes, Obesity is a known risk factor for numerous health have also become increasingly popular (OECD, 2017). problems, including hypertension, high cholesterol, diabetes, cardiovascular diseases, and some forms of The taxation of foods high in fat, sugar or salt cancer. As obesity is associated with higher risks of and/orsugarydrinksisalsousedbyanincreasing chronic illnesses, it is linked to significant additional number of countries to tackle overweight and obesity. healthcarecostsaswellassubstantialindirectcosts At least nine countries in Europe (Belgium, Estonia, due to lower employment and loss of work productivity Finland, France, Hungary, Ireland, Norway, Portugal (OECD/EU, 2016). and the United Kingdom) have taxes in place on sugar-sweetened beverages in 2018. On average across EU countries, 16% of adults were obese in 2014, according to data self-reported by people. At EU level, the 2007 Strategy for Europe on Obesity rates among adults vary greatly across EU Nutrition, Overweight and Obesity-related Health Issues countries, from 9% in Romania to 26% in Malta promotes a balanced diet and active lifestyle. It also (Figure 4.17). Obesity has increased in almost all encourages action by Member States and civil society European countries since 2000. It has notably increased in (notably through the EU Platform for action on Diet, Finland, France, Ireland, the Netherlands, and Sweden Physical Activity and Health) on food reformulation, where obesity rates used to be much lower. On the other marketing and advertising, physical activity, consumer hand, obesity rates among adults seem to have remained information, and advocacy and information exchange relatively stable between 2008 and 2014 in Belgium, the (European Commission, 2016). A project on food Czech Republic, Greece, Italy, Latvia and Poland. reformulation will start at the end of 2018 to provide a baseline to help Member States monitor the removal of Obesity rates based on the actual measurement of excess sugars, salt and fat from products that are height and weight are much higher than those based on bought every day in European supermarkets. self-reported data (as many people either overestimate their height or underestimate their weight), but these more reliable data are only available in a limited number of countries. These data show that obesity rates have Definition and comparability increased over the past decade in Finland, Hungary, Obesity is defined as excessive weight Luxembourg and the United Kingdom, while they have presenting health risks because of the high igure 4.18). plateaued in France and Ireland (F proportion of body fat. The most frequently used The prevalence of obesity is generally greater measure is based on the body mass index (BMI), among people with primary education (20% based on which is a single number that evaluates an self-reported data) than those with tertiary education individual’s weight in relation to height (weight/ (12%) on average (Figure 4.19). The gap in obesity by 2 height , with weight in kilograms and height in education level is particularly large in Luxembourg, metres). Based on the WHO classification, adults Portugal, Slovenia and Spain, while it is smaller in over age 18 with a BMI greater than or equal to 30 Latvia and Romania. are defined as obese. A number of behavioural and environmental factors Obesity rates can be assessed through self- have contributed to the long-term rise in obesity rates reported estimates of height and weight derived across EU countries, including the widespread availability from population-based health interview of energy-dense foods and an increasingly sedentary surveys, or measured estimates derived from lifestyle. These factors have created obesogenic health examinations. Estimates from health environments, putting people, and especially those in examinations are generally higher and more socially disadvantaged groups, more at risk. reliable than from health interviews. A growing number of countries have adopted policies to prevent and reverse obesity from spreading further. One approach has been to improve the information available to citizens to make more healthy choices (e.g. through food and menu labelling, public References awareness campaigns, mobile apps, restrictions or bans European Commission (2016), EU Platform on Diet, Physical on food product advertising targeting children). For Activity and Health, 2016 Annual Report, Brussels. instance, easy-to-understand interpretative labels put OECD/EU (2016), “The labour market impact of ill-health”, on the front of prepacked food have been used on a Health at a Glance: Europe 2016 in , OECD Publishing, voluntary basis in England (traffic-light system), France Paris, . https://doi.org/10.1787/health_glance_eur-2016-4-en (Nutri-Score), Denmark, Norway, Sweden and Lithuania (Keyhole logo) (OECD, 2017). Policies and programmes to www.oecd.org/ OECD (2017), “Obesity Update”, available at promote regular physical activity, such as subsidies to . els/health-systems/Obesity-Update-2017.pdf HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 126

129 II.4. RISK FACTORS 4.17. Changes in self-reported obesity rates among adults, 2000 to 2014 (or nearest year) 2014 2008 2000 % 30 25 25 21 20 20 20 19 19 19 19 18 18 18 20 17 17 17 17 17 16 16 16 16 15 15 15 14 14 14 13 15 12 12 10 10 9 10 5 0 Source: https:// Eurostat (EHIS 2008 and 2014) complemented with OECD Health Statistics 2018 for 2000 data and data for non-EU countries, . doi.org/10.1787/health-data-en http://dx.doi.org/10.1787/888933835250 1 2 4.18. Changes in measured obesity rates among adults, 2006 to 2016 (or nearest year) 2006 2016 % 35 30 29 30 26 24 23 23 23 25 19 18 17 20 17 15 10 https://doi.org/10.1787/health-data-en OECD Health statistics 2018, . Source: 1 2 http://dx.doi.org/10.1787/888933835269 4.19. Self-reported obesity rates by education level, 2014 Lowest level of education Highest level of education % 35 30 25 20 15 10 5 0 The lowest level of education refers to people with less than a high-school diploma, while the highest level refers to people with a Note: university or other tertiary diploma. Eurostat, EHIS 2014. Source: 1 2 http://dx.doi.org/10.1787/888933835288 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 127

130 II.4. RISK FACTORS MORTALITY DUE TO AIR POLLUTION AND EXTREME WEATHER CONDITIONS Environmental degradations, in particular air pollution and extreme weather conditions due at least Definition and comparability partly to climate change, expose people to health risk The first indicator presented here refers to and excess mortality. mortality due to air pollution (specifically PM 2.5 Air pollution increases the risk of various health and ozone) and is based on estimates from the problems (including respiratory diseases, lung cancer Global Burden of Disease study (IHME, 2016). and cardiovascular diseases), with children and older Fine particulate matter (PM) is a mixture of very people being particularly vulnerable. Some projections small particles and liquid droplets released into have estimated that outdoor air pollution may cause the air. PM refers to suspended particulates 2.5 6 to 9 million premature deaths a year worldwide by less than 2.5 microns in diameter that are 2060 and cost 1% of global GDP as a result of sick capable of penetrating deep into the respiratory days, medical bills and reduced agricultural output tract and causing significant health damage. (OECD, 2016). They are potentially more toxic than PM as they 10 In Europe, exposure to some serious air pollutants may include heavy metals and toxic organic such as fine particulate matter 2.5 (PM ) and ozone is 2.5 substances. Most fine particulate matters come estimated to have caused the death of 238 400 people from fuel combustion, including from vehicles, in 2016. Mortality rates due to air pollution are highest power plants, factories and households. in Central and Eastern Europe (e.g. Bulgaria and Ozone is a secondary pollutant (meaning that it Hungary) while they are lowest in Nordic countries is not emitted directly by any emission source), (Figure 4.20). formed in the lower part of the atmosphere from Climate change-related events – such as extreme complex chemical reactions following emissions temperatures, floods, and drought – also have serious of precursor gases such as nitrogen dioxides consequences on health and premature death. Heat (which are emitted during fuel combustion). waves can cause health problems such as fatigue, Ozone exposure is generally highest in emission- dehydration and stress, and can lead to respiratory dense countries with warm and sunny summers. and cardiovascular diseases, and aggravated allergies Data on fatalities due to extreme temperature (European Environment Agency, 2017; OECD, 2017). events come from the Emergency Events Database Some population groups, such as the elderly and (EM-DAT). EM-DAT includes all disasters people with chronic diseases, are more vulnerable to worldwide since 1900, conforming to at least one heat waves, but also to cold waves in some countries, of the following criteria: a) 10 or more people dead; particularly the Northern and Eastern part of Europe. b) 100 or more people affected; c) the declaration of Figure 4.21 shows the death rate related to a state of emergency; d) a call for international extreme temperature events in Europe, cumulated over assistance. Empty fields in the EM-DAT database the 2000-2016 period. Heat waves had a much bigger usually indicate missing values or non-reported impact than cold waves, particularly in Southern and information. Missing information in EM-DAT was Western Europe. Several Southern European countries complemented with data from national registry on were mostly impacted by the heat wave in 2003 when deaths by cause collected in the WHO Mortality more than 55 000 persons died in France, Italy, and Database. Deaths due to exposure to excessive Spain, and more recently in 2015 when 3 700 people natural heat (ICD code X30) and exposure to died in France and Belgium. Cold waves have had an excessive natural cold (X31) were selected. impact on mortality mainly in Eastern Europe and Nordic countries, with the latest largest event causing 350 deaths in Poland and Romania in 2012. References Cross-sectoral policy actions to limit greenhouse gas emissions and control the rise in temperature are European Environment Agency (2017), Climate change essential to limit the detrimental impacts on human adaptation and disaster risk reduction in Europe, health and the environment. While there have been Copenhagen. improvements in reducing the emission of a number OECD (2016), The Economic Consequences of Outdoor Air of air pollutants in the past decade, further efforts are https://doi.org/ Pollution, OECD Publishing, Paris, needed to reduce air pollution, notably by reducing . 10.1787/9789264257474-en emissions from motor vehicles, but also from power OECD (2017), Healthy People, Healthy Planet – The role of stations, which produce more pollution than any health systems in promoting healthier lifestyles and other industry. Health care systems also have a role to www.oecd.org/health/healthy-people- agreenerfuture, play in reducing environmental risk factors, for healthy-planet.htm . instance by supporting the implementation of nutritional guidelines for healthier food consumption IHME (2016), “Global Burden of Disease Study 2016 Results”, that can put less stress on the environmental resources http://ghdx.healthdata.org/gbd- Seattle, United States, used in food production (OECD, 2017). results-tool . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 128

131 II.4. RISK FACTORS 4.20. Deaths due to exposure to outdoor PM and ozone, 2016 2.5 180 206 219 150 300 852 340 867 Mortality rate 236 per 1 000 000 pop. 389 704 396 0 - 350 475 460 350 - 600 633 600 + 321 675 396 958 261 269 457 876 838 n.a. 700 1310 460 675 n.a. 486 306 370 573 362 308 373 (Malta) Source: IHME (Global Burden of Disease, 2016). 1 2 http://dx.doi.org/10.1787/888933835307 4.21. Deaths due to extreme weather conditions (heat waves and cold waves), cumulative from 2000 to 2016 57 209 62 76 8 44 32 21 Mortality rate per 1 000 000 pop. 11 0 - 20 16 48 115 20 - 100 114 223 100 - 250 45 291 250 + 24 41 67 125 362 140 22 201 6 n.a. 333 10 6 16 5 326 310 4 n.a. 57 (Malta) 11 Note: In France, Italy and Spain, most of the deaths are related to the heat wave in 2003. Source: Emergency Events Database (EM-DAT), complemented with WHO Mortality Database for Denmark, Finland, Ireland, Malta, Sweden, Iceland and Norway. 1 2 http://dx.doi.org/10.1787/888933835326 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 129

132

133 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART II Chapter 5 Health expenditure and financing This chapter looks at recent trends in health spending, both at an overall level but also disaggregated according to the type of health care service or medical good, and by health care provider. A particular focus is given to analysing spending in the hospital sector and on pharmaceuticals. The chapter ends by analysing how health care is financed in Europe, both in terms of the type of financing arrangements in place and the revenues that ultimately fund health care spending. Data presented in this chapter are jointly collected by OECD, Eurostat and WHO, and comply with internationally standardised definitions of health spending provided under the System of Health Accounts (SHA 2011) framework. In 2017, spending on health care in the European Union stood at 9.6% of gross domestic product, ranging from over 11% in France and Germany to less than 6% in Romania. This share remained largely unchanged from the previous two years as health spending grew in line with the economy in Europe. In most countries, payments for curative and rehabilitative care services made up the bulk of health spending, while spending on pharmaceuticals also accounted for a large share of health expenditure in some countries. Regarding the financing of health care, compulsory schemes, either government financed or through compulsory public or private health insurance, were the dominant method of financing accounting for more than three- quarters of overall health spending. However, out-of-pocket expenditure also played an important role in health financing for several Southern as well as Central and Eastern European economies. 131

134 II.5. HEALTH EXPENDITURE AND FINANCING HEALTH EXPENDITURE PER CAPITA Bulgaria and Romania saw health spending continue to The amount a country spends on health and the grow strongly. While nearly all EU countries have seen rate at which it can grow over time is influenced by a positive growth between 2013 and 2017, per capita wide array of social and economic determinants, as health spending in countries such as Greece and well as the financing arrangements and organisational Portugal continued to be at a lower level in 2017 than in structure of the health system itself. In particular, 2009. Outside of the EU, Iceland also experienced there is a strong relationship between the overall negative growth between 2009 and 2013 while Turkey income level of a country and how much the also saw a significant slowdown. Switzerland on the population of that country spends on health care. other hand appeared to be little affected with constant Given these factors, there are large variations to be annual growth of 2-2.5% throughout. observed in the level and growth of health spending across Europe. High-income countries such as Luxembourg, Norway and Switzerland are the European countries that spent the most on health in 2017 Definition and comparability (Figure 5.1). With spending at EUR 4 713 per person – Expenditure on health measures the final adjusted for differences in countries’ purchasing powers – consumption of health goods and services, as Luxembourg was the biggest spender in the European defined in the System of Health Accounts (OECD, Union. Among EU member states, Germany (EUR 4 160), Eurostat and WHO, 2017). This refers to current Sweden (EUR 4 019) and Austria (EUR 3 945) were also big spending on medical services and goods, public spenders. At the other end of the scale, Romania (EUR 983) health and prevention programmes, and and Bulgaria (EUR 1 234) were the lowest spending EU administration irrespective of the type of financing countries. Taking the European Union as a whole, per arrangement. capita health spending reached EUR 2 773 in 2017. Among Under Commission Regulation 2015/359, all EU some of the other European states, Switzerland countries are now obliged to produce health (EUR 5 799) and Norway (EUR 4 653) rank among the high expenditure data according to the boundaries and spenders overall while health spending per capita in definitions of the System of Health Accounts 2011 Turkey, Montenegro, the Former Yugoslav Republic of (SHA, 2011). Data on health expenditure for 2017 Macedonia and Albania were all below that of Romania. are considered preliminary, either estimated by After a number of years of slow or even negative national authorities or projected by the OECD health spending growth across Europe following the Secretariat, and are therefore subject to revision. economic crisis in 2008, growth rates picked up again in Countries’ health expenditures are converted nearly all countries in recent years. Across the European to a common currency (Euro) and are adjusted to Union as a whole, health spending per capita increased take account of the different purchasing power of by around 1.9% each year in real terms (adjusted for the national currencies, in order to compare inflation) between 2013 and 2017, compared with an spending levels. Economy-wide gross domestic annual growth rate of only 0.6% between 2009 and 2013. product (GDP) PPPs are used to compare relative During the crisis, ten EU countries saw expenditure on expenditure on health in relation to the rest of health retract in real terms with only Bulgaria and the economy. Romania among the member countries continuing to see growth above 5% per year. During the subsequent For the calculation of growth rates in real terms, four-year period, there has been a large-scale economy-wide GDP deflators are used. Although turnaround with all but two EU countries seeing some some countries (e.g. France and Norway) produce growth in health spending, albeit growth has remained their own health-specific deflators, based on slow in some countries (Figure 5.2). national methodologies, these are not currently used due to the limited availability and On an individual country basis, Greece comparability for all countries. experienced one of the biggest falls in health spending growth following the crisis. During the period 2009 to 2013, per capita health spending in Greece averaged an 8.7% annual drop. It is notable, however, that during the period 2003-2009, Greece experienced a much Reference steeper increase in real per capita health spending than the average for EU countries. Portugal, Croatia, A System of Health Accounts OECD/Eurostat/WHO (2017), Cyprus and Spain also experienced negative growth 2011: Revised edition , OECD Publishing, Paris, http:// between 2009 and 2013. On the other hand, Malta, dx.doi.org/10.1787/9789264270985-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 132

135 II.5. HEALTH EXPENDITURE AND FINANCING 5.1. Health expenditure per capita, 2017 (or nearest year) EUR PPP 7000 5 799 6000 4 713 5000 4 653 4 160 4 019 3 945 3 930 3 885 3 831 4000 3 572 3 493 3 309 3 045 3 013 2 773 3000 2 568 2 551 2 446 2 066 2 023 1 873 1 722 2000 1 678 1 625 1 551 1 473 1 463 1 409 1 367 1 252 1 234 987 983 824 728 638 1000 583 0 https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: ; Eurostat Database; WHO Global Health Expenditure Database. 1 2 http://dx.doi.org/10.1787/888933835345 5.2. Annual average growth rate (real terms) in per capita health spending, 2009 to 2017 (or nearest year) 2009-2013 2013-2017 10 7.6 7.3 8 6.1 5.8 5.5 5.5 5.4 6 4.9 4.6 4.5 3.5 3.4 3.3 3.2 4 2.7 2.5 2.4 2.3 2.3 2.2 2.1 2.1 1.9 1.9 1.8 1.7 1.6 1.6 1.6 1.5 1.4 1.4 1.3 1.1 1.1 1.1 2 1.1 1.1 1.0 0.8 0.8 0.8 0.8 0.8 0.7 0.7 0.6 0.6 0.4 0.4 0.1 0.0 0.0 0 -0.1 -0.4 -0.4 -2 -0.6 -0.7 -0.8 -0.8 -0.9 -1.4 -1.9 -1.9 -4 -3.1 -6 -8 -8.7 -10 1. Mainland Norway GDP price index used as deflator. Source: ; Eurostat Database. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933835364 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 133

136 II.5. HEALTH EXPENDITURE AND FINANCING HEALTH EXPENDITURE IN RELATION TO GDP the growth rate in GDP per capita. With the exception How much a country spends on health care in of a handful of countries – Greece, Hungary, Ireland, relation to all other goods and services in the Luxembourg and Romania – the share of GDP economy and how that changes over time depends allocated to health has increased in all EU countries. not only on the level of health spending but on the size of the economy as a whole. Looking at some of the larger EU economies, both France and Germany saw their health spending to In 2017, the EU as a whole devoted 9.6% of its GDP GDP ratio jump in 2009, stabilise, and then show a to health care (Figure 5.3). This figure stayed largely gradually increasing trend in subsequent years unchanged from the levels in both 2015 and 2016 as (Figure 5.5). While Italy and Spain also experienced a growth in health spending remained broadly in line similar increase in 2009, growth in health spending with overall economic growth. Among the EU member has been much more closely aligned with economic states, seven had spending on health at 10% or more of growth since then, resulting in the health-to-GDP GDP, with France (11.5%) and Germany (11.3%) having ratio remaining stable over the last five years or so. the highest shares of GDP spent on health. Nevertheless, these shares remain well below that of the United States, where health expenditure accounted for 17.2% of GDP in 2017. At the other end of the scale, the Definition and comparability share of health spending in GDP was lowest in Romania (5.2%), Luxembourg (6.1%), Latvia and Lithuania (both at Gross domestic product (GDP) is the sum of 6.3%). Across all of Europe, Switzerland allocated the final consumption, gross capital formation and biggest share, spending 12.3% of its GDP on health, net exports. Final consumption includes all the while Turkey at 4.2% of GDP had the lowest share. goods and services used by households or the community to satisfy their individual needs. It For a better understanding of the different includes final consumption expenditure of dynamics, the health spending to GDP ratio should be households, general government and non-profit considered together with health spending per capita. institutions serving households. While higher income countries tend to devote more of their income to health care, some countries with Data on health expenditure for 2017 are relatively high health expenditure per capita can have considered preliminary, either estimated by relatively low health spending to GDP ratios, and vice national authorities or projected by the OECD versa. For example, while Slovenia and Bulgaria spent Secretariat, and are therefore subject to revision. roughly the same share of their GDP on health in 2017, The GDP figures used to calculate the per capita health spending (adjusted to EUR PPP) was indicator health expenditure to GDP are based 64% higher in Slovenia (see Figure 5.1). Luxembourg, on official GDP data available as of mid-June despite having the highest per capita spending in the EU, 2017. Any subsequent revisions to GDP data are has one of the lowest shares of health spending relative not reflected in the indicator. to GDP. This reflects its high level of economic wealth. In countries, such as Ireland and Luxembourg, Over time, changes in health spending often where a significant proportion of GDP refers to reflect changes in GDP, though there is often a lag profits exported and not available for national before changes in economic conditions are reflected in consumption, gross national income (GNI) may be a health spending. When overall economic conditions more meaningful measure than GDP, but for rapidly deteriorated in many European countries from international comparability, GDP is used throughout. 2008 onwards, overall health spending was initially Both GDP and GNI increased significantly in maintained or even continued to grow (see Figure 5.4). Ireland in 2015 primarily due to the relocation of As a result, the health spending to GDP ratio in the a limited number of big economic operators to European Union jumped sharply to reach 9.6% – up Ireland leading to a substantial drop in the health from 8.8% in 2008. After a slight decline – as countries expenditure to GDP/GNI indicators in that year. introduced a range of measures in attempts to rein in government health spending and reduce burgeoning budgetary deficits (Morgan and Astolfi, 2014) – the pattern of health expenditure growth per capita has become more aligned to economic growth in many References European countries. Consequently, the ratio of health spending to GDP has been relatively stable. Morgan, D. and R. Astolfi (2014), “Health Spending Continues to Stagnate in Many OECD Countries”, OECD Changes in the ratio of health spending to GDP Health Working Papers ,No.68,OECDPublishing,Paris, are the result of the combined effect of growth in both http://dx.doi.org/10.1787/5jz5sq5qnwf5-en . GDP and health expenditure. Even taking into account the economic crisis, the long-term growth in health OECD/Eurostat/WHO (2017), A System of Health Accounts expenditure per capita (in real terms) in the European http:// 2011: Revised edition , OECD Publishing, Paris, Union between 2005 and 2017 has been greater than dx.doi.org/10.1787/9789264270985-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 134

137 II.5. HEALTH EXPENDITURE AND FINANCING 5.3. Health expenditure as a share of GDP, 2017 (or nearest year) % GDP 14 12.3 11.5 12 11.3 10.9 10.4 10.3 10.2 10.1 10.0 9.6 9.6 9.4 10 9.2 9.0 8.9 8.9 8.8 8.5 8.4 8.4 8.0 7.5 8 7.2 7.1 7.1 7.1 6.8 6.8 6.7 6.7 6.3 6.3 6.1 6.1 5.9 6 5.2 4.2 4 2 0 ; Eurostat Database; WHO Global Health Expenditure Database. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: http://dx.doi.org/10.1787/888933835383 1 2 5.4. Annual average growth (real terms) 5.5. Health expenditure as a share of GDP, in per capita health expenditure EU28 and selected countries, 2005 to 2017 and GDP, EU28, 2005 to 2017 GDP Health expenditure France Italy Germany Spain EU28 % % GDP 5 12 4 11 3 2 10 1 0 9 -1 8 -2 -3 7 -4 -5 6 2013 2015 2017 2005 2007 2009 2011 2011 2013 2015 2017 2005 2007 2009 OECD Health Statistics 2018, Source: OECD Health Statistics 2018, https://doi.org/10.1787/health- Source: https://doi.org/10.1787/health- ; Eurostat Database. data-en ; Eurostat Database. data-en http://dx.doi.org/10.1787/888933835421 1 2 http://dx.doi.org/10.1787/888933835402 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 135

138 II.5. HEALTH EXPENDITURE AND FINANCING HEALTH EXPENDITURE BY TYPE OF GOOD AND SERVICE average rate of 0.7% between 2008 and 2012. Spending A variety of factors, from disease burden and then recovered between 2012 and 2016, rising by an system priorities to organisational aspects and costs, average of 0.8% per year. The same trend was seen for help determine the share of spending on the various preventive care spending, which increased between types of health care goods and services. In 2016, EU 2004 and 2008 across the EU, but then contracted by member states spent three-fifths of their health 1.4% on average through the crisis years, despite expenditure on curative and rehabilitative care services, countries’ intentions to protect public health budgets. 20% went on medical goods (mainly pharmaceuticals), while 13% was on health-related long-term care. The While the growth in spending on inpatient and remaining 7% was spent on collective services, such as outpatient care was reduced during the years of the prevention and public health as well as the governance economic crisis, it remained positive, at 1.2% and 1.0% and administration of health care systems. respectively. During the crisis, some governments decided to protect expenditure for primary care and In 2016, the share of current health expenditure front-line services while looking for cuts elsewhere in going to curative and rehabilitative care ranged from the health system. Long-term care was the only major around half of total health spending in Bulgaria to three- health care service to experience an increase in quarters in Portugal. Greece had the highest proportion of spending growth over this period compared to the spending on inpatient care (including day care in pre-crisis years (2004-08), rising from 3.0% to 4.3%. hospitals), accounting for 42% of health spending. Inpatient care also accounted for more than one-third of all expenditure in Romania, Poland, Bulgaria and Austria. However, for most EU countries, spending on outpatient Definition and comparability care (including home-based curative and rehabilitative care and ancillary services) exceeded that on inpatient The System of Health Accounts (OECD, Eurostat care, notably in Portugal where outpatient care accounted and WHO, 2017) defines the boundaries of the for just under half of all health spending (49%). health care system. The functional dimension defines the type of health care consumed. Current The other major category of health spending is health expenditure comprises personal health care medical goods consumed in outpatient settings. A range (curative and rehabilitative care, long-term care, of factors can influence spending, including differences ancillary services and medical goods) and in distribution channels, the prevalence of generic drugs, collective services (prevention and public health as well as relative prices in different countries.The share services as well as health administration). of medical goods spending tended to be highest in Curative, rehabilitative and long-term care can also Southern and Central European countries and be classified by mode of provision (inpatient, day represented the largest component of health spending in care, outpatient and home care). Concerning Bulgaria and the Slovak Republic. In contrast, the shares long-term care, only care that relates to the in Western European and Scandinavian countries management of the deterioration in a person’s tended to be smaller: medical goods accounted for less health is reported as health expenditure, although than 15% of overall health spending in Denmark, the it is difficult in certain countries to clearly separate Netherlands, the United Kingdom, Luxembourg, Ireland out the health and social aspects of long-term care. and Sweden in 2016. Some countries can have difficulties separating There are also differences in countries’ spending spending on pharmaceuticals used as an integral on health-related long-term care. Countries such as part of hospital care from those intended for use Sweden and the Netherlands, with established formal outside of the hospital, potentially leading to an arrangements for the elderly and the dependent underestimate of pharmaceutical spending and an population, allocated more than a quarter of all health overestimate of inpatient and/or outpatient care. spending to long-term care in 2016. In many Southern and Central European countries, with more informal The variation between countries in price levels arrangements, the expenditure on formal long-term of medical goods (tradable) is generally smaller care services accounts for a much smaller share of total than that for health services (non-tradable). spending. Hence, spending on medical goods will tend to make up a larger share of health spending in Figure 5.7 presents the growth in key health goods low-income countries. and services for three time periods: before the financial crisis (2004-2008), during and immediately after the financial crisis (2008-12) and in the most recent period (2012-16). The financial crisis hampered growth in most parts of the health sector. Growth rates did Reference recover, but not to match pre-crisis levels. Following an average annual per capita increase of A System of Health Accounts OECD/Eurostat/WHO (2017), 1.8% over the years leading up to the financial crisis, EU 2011: Revised edition , OECD Publishing, Paris, http:// retail pharmaceutical expenditure fell by an annual dx.doi.org/10.1787/9789264270985-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 136

139 II.5. HEALTH EXPENDITURE AND FINANCING 5.6. Health expenditure by function, 2016 (or nearest year) Outpatient care** Collective services Long-term care Inpatient care* Medical goods % 0 10 3 4 4 4 4 4 4 5 4 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 7 7 7 7 8 8 8 9 11 90 10 14 12 15 11 20 19 14 12 16 17 17 22 14 21 23 21 20 23 19 22 20 20 30 31 80 32 32 30 34 32 3 3 44 25 21 28 6 19 23 70 15 26 6 22 10 19 19 9 26 10 13 21 15 13 16 18 3 8 4 5 60 6 40 22 30 50 49 28 31 42 18 31 32 27 30 37 26 27 31 33 31 31 34 36 32 25 36 29 17 28 32 33 29 40 30 34 30 42 20 37 35 34 34 33 32 32 32 31 31 31 30 30 29 28 28 28 28 28 27 26 26 26 26 26 25 25 25 25 24 22 10 0 * Refers to curative-rehabilitative care in inpatient and day care settings. ** Includes home care and ancillary services. Countries are ranked by the sum of inpatient and outpatient care as a share of current health expenditure. Note: Source: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en ; Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933835440 5.7. Growth rates of health expenditure per capita for selected functions, EU average, 2004-2016 2008-2012 2004-2008 2012-2016 % 5 4.7 4.3 4 3.1 3.0 3 2.5 2.4 2.2 2.2 1.8 2 1.3 1.2 1.2 1.0 0.9 0.8 1 0.0 0 -0.7 -1 -1.4 -2 Inpatient care Outpatient care Long-term care Pharmaceuticals Prevention Administration OECD Health Statistics 2018, ; Eurostat Database. https://doi.org/10.1787/health-data-en Source: 1 2 http://dx.doi.org/10.1787/888933835459 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 137

140 II.5. HEALTH EXPENDITURE AND FINANCING HEALTH EXPENDITURE IN HOSPITALS provided to outpatients. Outpatient care accounted for Breaking down health spending by provider offers an 45% of hospital spending in Portugal, and more than 40% organisational perspective, by identifying the setting in in Sweden, Estonia, Finland and Denmark. On the other which different health services are delivered. Care can be hand, in Poland, Bulgaria, Romania, Belgium, Germany provided in a variety of institutions, ranging from hospitals and Greece, less than 10% of hospital expenditure goes on and medical practices, to pharmacies, care homes and even outpatient care. private households caring for family members. Hospitals have traditionally been the key health care provider, in terms of their share of health spending. In 2016, health services in hospitals accounted for nearly Definition and comparability two-fifths of all EU health expenditure and represented The classification of health care providers is the largest spending category for most EU countries. In defined in the System of Health Accounts (OECD, Estonia, Cyprus and Italy almost half of health care Eurostat and WHO, 2017) and encompasses primary expenditure related to hospital services. In contrast, in providers, i.e. organisations and actors that deliver Germany a greater proportion of health services are health care goods and services as their primary provided in ambulatory settings and, by consequence, activity, as well as secondary providers, for which hospital services in Germany accounted for less than 30% health care provision is only one among a number of of health spending, the lowest share in the EU. Part of the activities. variation in the share of hospital spending can also be attributed to the provision of outpatient pharmaceuticals, The main categories of primary providers are with hospital pharmacies playing a larger role in some hospitals, residential long-term care facilities, countries than in others. ambulatory providers (e.g. offices of general and specialised physicians, dental practices, ambulatory While expenditure on hospital services varies health care centres), providers of ancillary services considerably between EU member states, it tends to be in (e.g. ambulance services), retailers (e.g. pharmacies), line with overall health care expenditure, with high- and providers of preventive care (e.g. public health income countries spending the most. Overall, EU countries institutes). spent EUR 1 059 per person on hospitals in 2016. Spending was highest in Denmark at EUR 1 653 per person, and was Secondary providers include, for example, more than EUR 1 500 per capita in Luxembourg and supermarkets that sell over-the-counter medicines, or Sweden (Figure 5.8). By comparison, spending on hospitals facilities that provide health care services to a in Romania accounted for less than EUR 350 per person restricted group of the population such as prisons or and was at a similar level for Latvia, Bulgaria and Poland. police health services. Secondary providers also include providers of health care system administration While the types of care delivered in hospital settings and financing (e.g. government agencies, health differ across EU countries (Figure 5.9), inpatient and insurance agencies) and households as providers of outpatient care remain the most common services home health care. provided. In 2016, spending on inpatient curative and rehabilitative care accounted for at least half of all Differences can exist in the administering and hospital expenditure for the majority of EU countries and dispensing of pharmaceuticals to outpatients in more than 90% in Poland, Germany and Greece. This high hospitals. Some countries have a larger range of share is due, to a large extent, to the use of alternative pharmaceuticals dispensed in hospital outpatient settings for care delivery; for example, specialised settings, which should be considered when comparing outpatient services delivered in ambulatory centres or overall hospital spending. In addition, some of these private practices in Germany (Busse and Blümel, 2014). costs may erroneously be accounted under curative The share of hospital inpatient care is lowest in Portugal, care rather than under pharmaceuticals. Estonia, the Netherlands, Croatia and Finland, where hospital outpatient services play a much greater role. The share of hospital spending on day care services has generally been increasing in most EU countries. This References often reflects an explicit policy to generate efficiency Busse, R. and M. Blümel (2014), “Germany: Health system gains and reduce hospital waiting times (OECD, 2017). review”, Health Systems in Transition, Vol. 16/2, Moreover, for some interventions, guidelines advise that www.euro.who.int/__data/assets/pdf_file/0008/255932/ day care procedures are the most appropriate treatment HiT-Germany.pdf?ua=1 . method. For all EU countries, inpatient and day care services together account for at least half of hospital , OECD (2017), Tackling Wasteful Spending on Health spending. http://dx.doi.org/10.1787/ OECD Publishing, Paris, . 9789264266414-en Hospitals can also be important providers of outpatient care, for example through accident and OECD/Eurostat/WHO (2017), A System of Health Accounts emergency departments, hospital-based specialist http:// , OECD Publishing, Paris, 2011: Revised edition outpatient units, or laboratory and imaging services dx.doi.org/10.1787/9789264270985-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 138

141 II.5. HEALTH EXPENDITURE AND FINANCING 5.8. Hospital spending in per capita terms and as a share of health spending, 2016 Percentage of current health care expenditure EUR (PPP) per capita EUR PPP per capita % CHE 100 2 500 90 2 030 80 2 000 1 829 1 653 70 1 516 1 505 1 491 1 407 60 1 500 1 349 1 331 1 277 1 195 1 168 1 166 1 149 50 1 087 1 066 1 059 1 006 40 1 000 865 859 762 722 744 684 30 537 530 513 508 459 394 371 20 500 340 10 0 0 ; Eurostat Database. OECD Health Statistics 2018, Source: https://doi.org/10.1787/health-data-en 1 2 http://dx.doi.org/10.1787/888933835478 5.9. Hospital expenditure by type of service, 2016 (or nearest year) Day care Inpatient care Other** Long-term care Outpatient care* % 0 10 3 3 3 3 4 3 3 4 5 5 3 5 6 6 77 7 7 4 3 9 4 12 13 14 18 12 9 90 9 11 21 4 18 21 15 15 15 26 32 29 24 23 25 80 40 34 35 43 15 10 38 45 7 42 43 7 9 24 9 70 21 17 17 4 4 9 9 12 60 5 5 13 5 10 5 6 50 94 93 15 92 88 86 85 83 82 40 73 72 71 71 70 69 67 66 63 61 61 61 30 60 59 58 57 56 55 52 49 48 48 47 20 39 10 0 * Refers to curative-rehabilitative care in outpatient and home-based settings and ancillary services. ** Includes medical goods and collective health services. Countries are ranked by inpatient care as a share of hospital expenditure. Note: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en ; Eurostat Database. Source: 1 2 http://dx.doi.org/10.1787/888933835497 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 139

142 II.5. HEALTH EXPENDITURE AND FINANCING PHARMACEUTICAL EXPENDITURE crisis – such as cutting manufacturer prices and margins Pharmaceuticals play a vital role in the health for pharmacists and wholesalers, introducing system. After inpatient and outpatient care, compulsory rebates, de-listing some pharmaceuticals pharmaceuticals (excluding those used in hospitals) and incentivising the use of generics (Belloni, Morgan represent the third largest item of health care and Paris, 2016). Patent expiries for a number of spending, accounting for a sixth of health expenditure blockbuster drugs also contributed to the fall in spending in the EU in 2016. The challenge for policymakers, over this period. However, new high cost treatments acknowledging that health care budgets are limited, is such as for Hepatitis C and some oncological drugs help to balance access for new medicines while providing explain a return to positive growth rates in more recent the right incentives to industry. years for some countries. The total retail pharmaceutical bill across the The retail pharmaceutical sector only tells part of European Union was more than EUR 210 billion the story, since spending on pharmaceuticals used (adjusted for purchasing power parities) in 2016 and an during hospital care can typically add another 20% to increase of around 5% (in nominal terms) since 2010. a country’s pharmaceutical bill (Belloni et al., 2016). The variations in per capita pharmaceutical spending Available data in a number of European countries across countries can reflect differences in suggest that pharmaceutical spending growth in the pharmaceutical prices, consumption and dispensing hospital setting has outpaced that of retail practices, as well as the market penetration of generics igure 5.12). Average annual growth pharmaceuticals (F (Figure 5.10 ). Among EU Member States, Germany spent of pharmaceuticals consumed in hospitals was the most on pharmaceuticals on a per capita basis significantly higher in Iceland and Denmark than (EUR 572), around 40% above the EU average. Ireland retail pharmaceutical spending between 2009 and (EUR 498) and Belgium (EUR 491) spent nearly 20% more 2016. Although on a smaller scale, the same is true for on medicines per capita than the EU average. At the Germany, Finland, Estonia and Spain. other end of the scale, Denmark (EUR 203), Romania (EUR 255), Estonia (EUR 262) and Poland (EUR 267) had relatively low spending levels. Outside the EU, Switzerland (EUR 742) spent significantly more on Definition and comparability medicines per capita than any other country in Europe. Pharmaceutical expenditure covers spending Around four out of every five euros spent on retail on prescription medicines and self-medication, pharmaceuticals goes on prescription medicines, with often referred to as over-the-counter products. the rest on over-the-counter medicines (OTC). OTC Final expenditure on pharmaceuticals includes medicines are pharmaceuticals that are generally wholesale and retail margins and value-added bought without prescription with their costs, in most tax. Total pharmaceutical spending refers in cases, fully borne by patients. However, it should be most countries to “net” spending, i.e. adjusted for noted that pharmaceuticals classed as prescriptions in possible rebates payable by manufacturers, one country might be classed as an OTC medicine in wholesalers or pharmacies. another. The share of OTC medicines is particularly high in Poland, accounting for half of pharmaceutical Pharmaceuticals consumed in hospitals and spending, and stands at 30% or more in Spain (36%), other health care settings as part of an inpatient or Latvia (31%) and Cyprus (30%). day case treatment are excluded (data available suggest that their inclusion would add another 30% The cost of pharmaceuticals is predominantly to pharmaceutical spending on average). In some covered by government or compulsory insurance countries, expenditure associated with the schemes in Europe (F igure 5.11). In the EU, these schemes administering and dispensing of pharmaceuticals cover around 64% of all retail pharmaceutical spending, for outpatients in hospitals may be incorrectly with out-of-pocket payments (34%) and voluntary private accounted for under curative care, affecting the insurance (1%) financing the remaining part. Coverage is comparability in retail pharmaceutical expenditure. most generous in Germany and Luxembourg where government and compulsory insurance schemes pay for Pharmaceutical expenditure per capita is 80% or more of all pharmaceutical costs. By contrast, in adjusted to take account of differences in around a quarter of EU Member States, public or purchasing power. mandatory schemes cover less than half the amount spent on medicines and coverage is particularly low in Bulgaria (19%) and Cyprus (18%). During the financial crisis, average annual spending Reference growth on retail pharmaceuticals in the EU was much lower compared to other health services (see indicator Belloni, A., D. Morgan and V. Paris (2016), “Pharmaceutical “Health expenditure by type of good and service”) and Expenditure and Policies: Past Trends and Future was negative in some years. Several countries took Challenges”, OECD Health Working Papers ,No.87,OECD measures to reduce pharmaceutical spending during the http://dx.doi.org/10.1787/5jm0q1f4cdq7-en Publishing, Paris, . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 140

143 II.5. HEALTH EXPENDITURE AND FINANCING 5.10. Expenditure on retail pharmaceuticals per capita, 2016 Total (no breakdown) Non-prescribed (over-the-counter) Prescribed EUR PPP 742 800 700 572 600 498 484 491 464 456 447 440 500 438 417 416 398 398 391 357 364 358 368 350 400 326 324 316 308 299 299 290 267 262 275 300 255 203 200 100 0 1. Includes medical non-durables (resulting in an overestimation of around 5-10%). ; Eurostat Database. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: http://dx.doi.org/10.1787/888933835516 1 2 5.11. Expenditure on retail pharmaceuticals by type of financing, 2016 Other Out-of-pocket payments Voluntary health insurance Government/compulsory schemes 100 5 6 15 13 17 23 23 29 30 30 31 31 32 34 6 37 80 41 7 42 43 44 44 45 45 45 46 47 48 51 57 11 65 66 66 59 7 26 1 60 81 73 3 1 4 6 84 3 40 80 77 76 71 70 70 68 68 64 63 61 59 57 56 56 55 55 54 54 53 52 51 50 44 20 40 3 36 35 34 33 19 18 0 1. Includes expenditure on medical non-durables. “Other” includes non-profit-schemes, enterprises and rest of world. Note: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en ; Eurostat Database. Source: http://dx.doi.org/10.1787/888933835535 1 2 5.12. Annual average growth in retail and hospital pharmaceutical expenditure, in real terms, 2009 to 2016 (or nearest year) Retail pharmaceuticals Hospital pharmaceuticals % 12 9.7 10 8 5.5 6 3.2 2.8 2.8 4 2.3 1.4 1.1 0.9 0.7 2 0.0 0 -2 -1.8 -4 -2.5 -6 -5.4 -8 Spain Portugal Iceland Germany Finland Estonia Denmark OECD estimates for Portugal exclude expenditure on other medical products from retail spending. Note: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en . Source: 1 2 http://dx.doi.org/10.1787/888933835554 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 141

144 II.5. HEALTH EXPENDITURE AND FINANCING FINANCING OF HEALTH EXPENDITURE Health care can be paid for through a variety of overall budgets. Hence, health is competing for public financing arrangements. In countries where individuals fundswithmanyothersectorssuchaseducation, are entitled to health care services based, for example, on defence and housing. In 2016, around 17% of total their residency, government schemes are the government expenditure was allocated to health in the predominant arrangement. In others, some form of EU (Figure 5.15). In Germany, the United Kingdom and compulsory health insurance (either social health Sweden the share of public spending dedicated to insurance or cover organised through private insurers) health care was closer to 20%, while in Hungary and usually covers the bulk of health expenditure. In addition, Poland it was just over 10%. payments by households (either standalone payments or as part of co-payment arrangements) as well as various forms of voluntary health insurance intended to replace, Definition and comparability complement or supplement automatic or compulsory coverage make up the rest of health spending. The financing of health care can be analysed from the point of view of financing schemes In 2016, around 77% of EU health spending was (financing arrangements through which health financed through government and compulsory services are paid for and obtained by people, insurance (Figure 5.13). In Denmark, Sweden and the e.g. social health insurance) and types of revenues United Kingdom, central, regional or local government of financing schemes (e.g. social insurance financed around 80% or more of all health spending. In contributions) (OECD, Eurostat and WHO, 2011). Germany, France, the Netherlands, the Slovak Republic and Croatia, compulsory health insurance financed Financing schemes include government more than three-quarters of all health expenditure. schemes, compulsory health insurance as well as Cyprus was the only EU country where less than half of voluntary health insurance and private funds all health spending was financed through government such as households’ out-of-pocket payments, or compulsory insurance schemes. NGOs and private corporations. Out-of-pocket payments are expenditures borne directly by In five EU countries – Malta, Greece, Latvia, Bulgaria patients, which can take the form of cost-sharing and Cyprus – households’ out-of-pocket payments of services included in the publicly defined accounted for more than one-third of health spending, benefit package and also direct purchases of while only in Slovenia and Ireland did voluntary health goods and services. insurance cover more than 10% of health spending. Health financing schemes have to raise Financing schemes can be funded by different types revenues in order to pay for health care goods and of revenue streams. Public revenues include service for the population they are covering. governmental transfers (mainly coming from taxation) Financing schemes can receive transfers from the and social insurance contributions paid by employees, government, social insurance contributions, employers and others. Private revenues include the voluntary or compulsory prepayments premiums paid to both voluntary and compulsory private (e.g. insurance premiums), other domestic insurance as well as any other resources coming directly revenues and revenues from abroad as part of from households and corporations. In 2016, among a development aid. group of 12 EU countries with comparable data, public sources funded 76% of all health spending, (F igure 5.14). Total government expenditure is used as defined in the System of National Accounts and The types of revenues are closely related to the includes as major components: intermediate system of health care financing. In Denmark, for consumption, compensation of employees, example, where health care is predominantly interest, social benefits, social transfers in kind, purchased through local government schemes, this is subsidies, other current expenditure and capital almost entirely funded via government transfers. Other expenditure payable by central, regional and local types of financing may rely on a mix of different revenue governments as well as social security funds. sources. For example, if a social health insurance scheme exists, like in the case of Belgium and Germany, social insurance contributions will typically be a major revenue source. However, social health insurance schemes can also receive governmental transfers to a References varying extent. Analysing the structure of financing schemes with the types of revenues that these schemes Mueller, M. and D. Morgan (2017), “New insights into health financing: First results of the international receive can give important insights into the overall data collection under the System of Health Accounts financing of health: in many countries, the government’s Health Policy , Vol. 121/7, pp. 764-769, 2011 framework”, role in funding health care is typically more than being http://dx.doi.org/10.1016/j.healthpol.2017.04.008 . just a simple purchaser of health services (Mueller and Morgan, 2017). OECD/Eurostat/WHO (2017), A System of Health Accounts Governments (including social security schemes) http:// 2011: Revised edition , OECD Publishing, Paris, finance many different public services out of their dx.doi.org/10.1787/9789264270985-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 142

145 II.5. HEALTH EXPENDITURE AND FINANCING 5.13. Health expenditure by type of financing, 2016 (or nearest year) Out-of-pocket Compulsory health insurance Government schemes Voluntary health insurance Other % 10 0 4 5 5 5 4 6 5 3 8 7 7 5 6 5 5 15 14 12 15 14 12 6 90 17 18 15 21 10 15 23 11 16 16 11 18 23 20 15 19 24 32 32 36 23 30 35 28 80 11 12 34 13 41 30 45 48 57 70 5 13 45 60 41 44 56 50 61 70 30 78 65 72 78 75 65 76 84 42 84 82 40 60 75 79 57 69 58 74 74 72 58 66 66 65 63 62 55 30 41 55 34 42 20 36 31 30 23 21 10 18 13 12 11 10 10 9 9 9 8 7 6 6 5 4 4 4 2 1 0 Countries are ranked by government schemes and compulsory health insurance as a share of current health expenditure. Note: https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: ; Eurostat Database; WHO Global Health Expenditure Database. http://dx.doi.org/10.1787/888933835573 1 2 5.14. Public financing as a share of health 5.15. Government transfers and social spending, by source of funding, contributions for health care as share of total government expenditure, 2016 (or nearest year) 2016 (or nearest year) Social contributions Government transfers % % GDP 90 22 8 19.6 80 20 18.8 18.5 10 70 17.6 5 18 16.7 16.2 45 31 15.9 60 41 45 17 16 15.1 15.0 50 13.6 63 14 13.1 62 84 84 12.4 82 58 40 80 67 77 11.8 12 68 66 11.0 10.6 30 10.4 1 49 45 10 20 39 38 36 28 8 10 15 13 11 6 6 0 Chart only contains the countries reporting revenues of Chart only contains the countries reporting revenues of Note: Note: financing schemes. financing schemes. OECD Health Statistics 2018, https://doi.org/10.1787/health- OECD Health Statistics 2018, https://doi.org/10.1787/health- Source: Source: ; Eurostat Database. data-en ; Eurostat Database. data-en http://dx.doi.org/10.1787/888933835611 1 2 http://dx.doi.org/10.1787/888933835592 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 143

146

147 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART II Chapter 6 Effectiveness: Quality of care and patient experience This chapter starts with a broad indicator of avoidable mortality, providing a general assessment of the effectiveness of public health and health care systems in reducing premature deaths. In 2015, more than 1.2 million people in EU countries died prematurely from diseases and injuries that could potentially have been avoided through more effective public health policies or health care. The main causes of avoidable mortality include ischaemic heart diseases, lung cancer and accidents. Vaccine-preventable diseases have resurged in some parts of Europe in recent years, highlighting the importance of assuring effective vaccination coverage across all European countries. In some EU countries, at least 10% of children were not vaccinated against infectious diseases such as measles and hepatitis B in 2017, increasing the risk that these communicable diseases will spread. The quality of acute care in hospital for life-threatening conditions has generally improved over the past decade. Mortality rates following a hospital admission for acute myocardial infarction (AMI) has reduced by 30% on average between 2005 and 2015, and mortality rates following an admission for stroke has also come down by over 20%. Yet, wide disparities in the quality of acute care persist not only between countries but also between hospitals within each country. Health systems in Europe have also made progress in tackling cancer through the implementation of population-based screening programmes and the provision of effective and timely cancer care, as reflected by increased survival following diagnosis and reduced cancer mortality in most countries. Health care needs to be provided by putting patients at the centre. While data on patient-reported experience remain limited, the available data indicate that patients generally report positive experiences from their contacts with doctors. 145

148 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE AVOIDABLE MORTALITY (PREVENTABLE AND AMENABLE) ischemic heart diseases and cerebrovascular diseases, Indicators of avoidable mortality provide a but also by higher mortality from some types of cancer general “starting point” to assess the effectiveness of and other treatable diseases. These three countries are public health and health care systems in reducing also among those that spend the least on health across premature deaths from various diseases and injuries, the EU. Hence, additional expenditure on health could but further analysis is required to assess more contribute to reductions in amenable mortality. precisely different causes of potentially avoidable deaths and possible interventions to reduce them. Looking at trends over time, the age-standardised rate of amenable mortality has declined by In 2015, over 1 million deaths across EU countries approximately 25% between 2005 and 2015 across the were considered to be potentially preventable through EU as a whole. This reduction has been particularly effective public health and prevention interventions rapid in Denmark and Finland (over 30%), driven and more than 570 000 deaths were considered to be mainly by a rapid decline in ischaemic heart diseases amenable (or treatable) through more effective and mortality due partly to reduced mortality rates for timely health care (Figure 6.1). The overall number of people admitted to hospital for a heart attack (see potentially avoidable deaths was around 1.2 million indicator “Mortality following AMI”). deaths in 2015, taking into account that some diseases are considered to be both preventable and amenable (Eurostat, 2018). The main causes of preventable mortality are Definition and comparability ischaemic heart diseases (which are also considered to Based on the Eurostat definitions, preventable be amenable to health care when these diseases occur), mortality is defined as deaths that could be lung cancer, road accidents and other types of accidents, avoided through public health and prevention alcohol-related deaths, colorectal cancer and suicides. interventions, whereas amenable (or treatable) Combined, these causes of death account for over mortality is defined as deaths that could be two-thirds of all deaths considered to be preventable avoided through effective and timely health care through more effective public health and prevention (Eurostat, 2018). interventions in EU countries. The two lists of preventable and amenable The main causes of amenable (or treatable) mortality focus on premature deaths, defined as mortality are ischaemic heart diseases and deaths under age 75. However, a lower or higher cerebrovascular diseases, which together account for age threshold is used for some selected causes nearly half of all amenable deaths. Mortality from of death. colorectal cancer and breast cancer also account for a A number of causes of death are included in considerable number of amenable deaths (20% of the both the preventable and amenable mortality lists total) that could be reduced both through earlier as they are considered to be both potentially detections and more effective and timely treatments preventable through public health interventions (see indicators on screening, survival and mortality and treatable through effective and timely health for breast cancer and colorectal cancer). care when they occur. For example, ischemic heart The age-standardised rate of preventable mortality diseases, colorectal cancer and breast cancer are is lowest in Italy, Cyprus and Spain, with rates at least considered to be both 100% preventable and 100% 25% lower than the EU average. By contrast, preventable amenable to health care.This “double counting” of mortality rates are about two times greater than the EU several causes of death means that the sum of the average in Lithuania, Hungary and Latvia (Figure 6.2). preventable and amenable mortality lists is much The high rates of preventable mortality in these three greater than the overall number of avoidable countries are due mainly to much higher death rates deaths. from ischaemic heart diseases, accidents, alcohol- The two current lists of preventable and related deaths, suicides (particularly in Lithuania) and amenable mortality were adopted by a Eurostat lung cancer (particularly in Hungary). Task Force in 2013. These lists may be subject to The age-standardised rate of amenable mortality future revisions. is lowest in France, Spain and the Netherlands, due to these countries having among the lowest death rate from ischaemic heart diseases and cerebrovascular diseases. Lithuania, Latvia and Romania have the Reference highest rates of amenable mortality, more than two- and-a-half times higher than the EU average , Eurostat (2018), Amenable and preventable deaths statistics (Figure 6.3), driven mainly by higher death rates from Statistics Explained, June 2018. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 146

149 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.1. Leading causes of preventable and amenable mortality in the European Union, 2015 Amenable mortality Preventable mortality (570 791 deaths in 2015) (1 003 027 deaths in 2015) Colorectal Colorectal cancer, 7% cancer, 12% Others, 22% Lung cancer, Others, 29% Breast cancer, 17% 9% Hypertension, Influenza and 5% Hypertension, pneumonia, 5% 5% Alcohol, 7% Cerebrovascular Sui c ide, 7% es diseas , 16% Ischaemic heart Ischaemic heart di es seas , 32% es diseas , 18% Accidents, 16% Eurostat Database. Source: http://dx.doi.org/10.1787/888933835630 1 2 6.3. Amenable mortality rates, 2015 6.2. Preventable mortality rates, 2015 France 78 Italy 151 Spain 88 Cyprus ¹ 155 Netherlands 91 Spain 159 Italy 93 Malta ¹ 171 Belgium 94 Sweden 174 Cyprus ¹ 94 Greece 182 Sweden 97 France 184 Luxembourg ¹ 97 Portugal 186 Denmark 98 Ireland 188 Austria 109 Netherlands 189 Ireland 111 Luxembourg ¹ 203 Portugal 111 Denmark 206 Finland 111 United Kingdom 211 Germany 116 Finland 213 United Kingdom 117 Germany 215 Malta ¹ 119 EU28 216 Greece 127 Belgium 216 EU28 127 Austria 221 Slovenia 128 Slovenia 265 Poland 169 Bulgaria 271 Czech Rep. 179 Poland 276 Croatia 216 Czech Rep. 285 Estonia 224 Estonia 307 Slovak Rep. 250 Croatia 327 Hungary 268 Slovak Rep. 362 Bulgaria 282 Romania 363 Romania 318 Latvia 415 Latvia 326 Hungary 418 Lithuania 326 Lithuania 446 Switzerland 75 Switzerland 163 Norway 87 Iceland ¹ 175 Iceland ¹ 91 Norway 182 Turkey 189 Turkey 231 Serbia 236 Serbia 276 300 400 500 0 100 200 0 200 300 400 500 100 Age-standardised rates per 100 000 population Age-standardised rates per 100 000 population 1. Three-year average (2013-15). 1. Three-year average (2013-15). Eurostat Database. Source: Eurostat Database. Source: http://dx.doi.org/10.1787/888933835668 1 2 1 2 http://dx.doi.org/10.1787/888933835649 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 147

150 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE CHILDHOOD VACCINATIONS Viruses do not respect national borders. About Diseases such as measles, diphtheria, pertussis 10% of all measles cases are due to infections acquired and influenza are highly infectious and spread through by people while travelling outside of their home human contact, while the hepatitis B virus is country or imported by visitors from other countries transmitted by contact with blood or body fluids of an (ECDC, 2018). This highlights the importance of infected person, by sex or from mother to child. maintaining high vaccination coverage across Effective vaccination is available to prevent all these countries in an EU and global context. infectious diseases. All EU countries have established childhood vaccination programmes, contributing to On average, 93% of children at age 1 receive reducing many deaths related to these diseases, hepatitis B vaccination across EU countries where this although the number and type of vaccines vary to vaccination is part of the national immunisation some extent across countries. programme. The vaccination rates are particularly high in Latvia and Portugal, but less than 90% in Germany, Malta, Vaccine-preventable diseases have resurged in Slovenia and Sweden (Figure 6.5). In Sweden, hepatitis B some parts of Europe recently due to a combination of was included in the national childhood vaccination declining vaccine coverage, increasing supply programme only in 2016, which partly explains why the shortages and growing vaccine hesitancy. The coverage is still relatively low. In Denmark and Finland, European Commission has called for stronger efforts where data are not available, hepatitis B vaccination is not and cooperation to tackle hesitancy against vaccines, yet part of the general infant vaccination programme, but improve vaccination coverage and develop sustainable is provided to high-risk groups. Hungary and Slovenia vaccination policies in the EU (European Commission, have also not yet included hepatitis B vaccine in their 2018). infant vaccination programme. Vaccination against measles is included in all Between 2007 and 2017, vaccination rates for national childhood vaccination programmes, whereas hepatitis B among children have increased by vaccination against hepatitis B has been included in a 8 percentage points on average across EU countries that growing number of countries, but is available only for have this vaccination included in their national certain risk groups in a few Nordic countries (Denmark, immunisation programme. The increase was particularly Finland and Iceland). The WHO recommends at least large in France, the Netherlands and Sweden. 95% coverage with two doses of measles-containing vaccine by 2020 (WHO, 2012). As for hepatitis B, a proportion of infections become chronic, and this risk is high particularly among infants and children. Definition and comparability Infected people are at high risk of death from cancer or cirrhosis of the liver. The hepatitis B vaccine is Vaccination rates reflect the percentage of considered to be 95% effective in preventing infection children under one year old who have received and its chronic consequences. WHO recommends that the respective vaccination (at least one dose all infants should receive their first dose of hepatitis B of measles-containing vaccine and three doses of vaccine as soon as possible after birth (WHO, 2017). hepatitis B vaccine) in a given year. The age of complete immunisation differs across countries On average across EU countries, 94% of children due to different immunisation schedules. For those received at least one dose of measles vaccination countries recommending the first dose of measles before turning age 1 (Figure 6.4). However, the vaccine after age one, the indicator is calculated as vaccination rate in 2017 did not reach more than 90% of the proportion of children less than two years of age children in Romania, Croatia, Cyprus and France. who have received that vaccine. Thus, these data Measles continues to spread in some parts of reflect the actual policy in a given country and are Europe. Between May 2017 and May 2018, 13 475 cases not always strictly comparable across countries. of measles were reported, up from 8 523 cases for the preceding 12-month period (see Chapter 3). Almost 85% of these cases were reported in Italy (4 032), Greece (2 752), France (2 436) and Romania (2 127). Most measles cases were reported among people who were References not vaccinated, particularly children below age 1 who were too young to have received the first dose of the ECDC (2018), Surveillance Report: Monthly measles and vaccine and older individuals who had missed , ECDC, Stockholm. rubella monitoring report – June 2018 vaccination (ECDC, 2018). European Commission (2018), Proposal for a Council Small decreases in vaccination rates can result in Recommendation on Strengthening Cooperation against Vaccine large increases in measles cases. In Romania, , European Commission, Brussels. Preventable Diseases vaccination coverage has decreased by 10 percentage Global Measles and Rubella: Strategic Plan 2012-20 , WHO (2012), points over the past decade. In Italy, the rate decreased WHO, Geneva. from 91% in 2010 to 85% in 2016, but it went back up to 92% in 2017. , WHO, Geneva. Global Hepatitis Report 2017 WHO (2017), HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 148

151 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.4. Vaccination against measles, children aged 1, 2017 (or nearest year) 92 94 96 97 93 96 97 % of children 94 vaccinated (measles) < 90 92 92 96 90 - 93.9 93 97 96 94 - 96.9 97 99 > 97 96 96 99 95 90 93 87 89 86 92 94 47 83 97 96 98 97 98 90 91 (Malta) WHO/UNICEF. Source: http://dx.doi.org/10.1787/888933835687 1 2 6.5. Vaccination against hepatitis B, children aged 1, 2017 (or nearest year) n.a. n.a. n.a. 76 92 98 n.a. % of children 94 vaccinated (hepatitis B) < 90 95 n.a. 95 90 - 93.9 92 87 97 94 - 96.9 94 94 > 97 96 90 n.a. 17 90 89 92 94 93 n.a. 94 92 75 91 93 99 98 96 98 97 88 (Malta) Note: Data for Denmark, Finland, Hungary, Iceland and Norway are not available because national infant vaccination programmes do not cover Hepatitis B. Data is not available for the United Kingdom. Source: WHO/UNICEF. 1 2 http://dx.doi.org/10.1787/888933835706 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 149

152 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE PATIENT EXPERIENCE WITH AMBULATORY CARE Across EU countries, delivering health care that is the perceived quality of family doctor/GP or health patient-centred is becoming a priority in health care centre services as reported in the European Quality of policy. Given the importance of utilising people’s voice Life Survey. The perceived quality of care is high in for developing health systems and improving quality of Austria and Luxembourg, while it is low in Poland and care, national efforts to develop and monitor patient- Greece (Figure 6.8). reported measures have been intensified in recent years. In recent years, reported patient experiences In many countries, responsible organisations have been have not changed significantly in most countries. established or existing institutions have been identified However, Estonia and Sweden have improved some for measuring and reporting patient experiences. These aspects of patient experiences recently. organisations develop survey instruments for regular collection of patient experience data and standardise procedures for analysis and reporting. An increasing Definition and comparability number of countries collect not only Patient-Reported Experience Measures (PREMs) but also Patient-Reported In order to monitor general patient experience Outcome Measures (PROMs) which collect patients’ with ambulatory care, the OECD recommends perception on their specific medical conditions and collecting data on patient experience with any general health, including mobility, pain/discomfort and doctor in ambulatory care settings. An increasing anxiety/depression, before and after a specific medical number of countries have been collecting patient intervention (OECD, 2018). experience data based on this recommendation through nationally representative population Countries use patient-reported data differently to surveys, while Portugal collects them through a drive quality improvements in health systems. In order nationally-representative service user survey. to promote quality of health care through increased provider accountability and transparency, many In 11 countries, the Commonwealth Fund’s countries report patient experience data in periodic International Health Policy Surveys of 2010 and national health system reports and/or on public 2016 were used as a data source, even though websites, showing differences across providers, regions there are limitations relating to the sample size and over time. In addition, Belgium and Norway use and response rates. Data from this source refer patient experience measures in payment mechanisms to patient experience with a GP, instead of with to promote quality improvement and patient-centred any doctor including both GP and specialist. In care. The Czech Republic, Denmark, France and the 2016, the Netherlands developed a national United Kingdom use patient experience data to inform population survey and this resulted in improved health care regulators for inspection, regulation and/or response rates and data quality. Poland collects accreditation. Patient-reported measures are also used data through national survey and the data refer in Belgium, Denmark and the Netherlands to provide to patient experience with a regular doctor. specific feedback for providers’ quality improvement. Rates for Figure 6.6 and Figure 6.7 are age-sex Several countries including Belgium and Denmark also standardised to the 2010 OECD population, to use patient-reported outcome measures systematically remove the effect of different population structures for quality improvement (Fujisawa and Klazinga, 2017; across countries. Desomer et al., 2018). Patients generally report positive experiences in relation to communication and autonomy in the ambulatory health care system. For example, the References majority of patients reported that they spent enough Desomer, A. et al. (2018), “Use of patient-reported outcome time with a doctor during consultation (Figure 6.6) and and experience measures in patient care and policy”, a doctor involved them in care and treatment decisions , No. 303, Health Care Knowledge Centre (KCE) Reports (Figure 6.7). For these and other aspects of patient Belgian KCE Health Services Research, Brussels. experience, Belgium and Luxembourg have high rates Fujisawa, R. and N. Klazinga (2017), “Measuring Patient with above 95% of patients reporting positive Experiences (PREMs): Progress Made by the OECD and experiences, while Poland has lower rates. For its Member Countries 2006-2015”, OECD Health Working , No. 102, OECD Publishing, Paris, https://doi.org/ Papers example, only one in two patients report having been 10.1787/893a07d2-en . involved in their care and treatment decisions during consultation in Poland. Across European countries, OECD (2018), “Patient-Reported Indicators Survey (PaRIS)”, www.oecd.org/health/paris.htm . these patient experiences are generally consistent with HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 150

153 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.6. Doctor spending enough time 6.7. Doctor involving patient with patient in consultation, in decisions about care and treatment, 2010 and 2016 (or nearest year) 2010 and 2016 (or nearest year) 2010 Confidence Interval 2016 Confidence Interval 2016 2016 2016 2010 Age-sex standardised rate per 100 patients Age-sex standardised rate per 100 patients 100 100 94 90 98 97 96 87 96 95 91 79 79 79 89 87 86 88 87 85 84 87 82 75 75 79 73 60 50 50 48 25 25 France² Estonia¹ Estonia¹ Norway² Norway² Sweden² Sweden² Belgium¹ Belgium¹ Poland¹ ² Poland¹ ² Portugal¹ Portugal¹ Germany² Germany² Czech Rep.¹ Czech Rep.¹ Switzerland² Switzerland² Netherlands² Netherlands² Luxembourg¹ Luxembourg¹ United Kingdom² United Kingdom² 1. National sources. 1. National sources. 2. Data refer to patient experiences with GP. 2. Data refer to patient experiences with GP. 95% confidence intervals have been calculated for all Note: Note: 95% confidence intervals have been calculated for all countries, represented by grey areas. countries, represented by grey areas. Commonwealth Fund International Health Policy Survey Commonwealth Fund International Health Policy Survey Source: Source: 2016 and other national sources. 2016 and other national sources. 1 2 http://dx.doi.org/10.1787/888933835744 1 2 http://dx.doi.org/10.1787/888933835725 6.8. Perceived quality of GP (family doctor) or health centre services, 2016 Mean score 10 8.4 8.4 9 8.1 8.0 7.9 7.9 7.8 7.7 7.7 7.5 7.5 7.5 7.4 8 7.3 7.2 7.1 7.1 7.1 7.0 7.0 7.0 7.0 7.0 6.9 6.9 6.9 6.9 6.8 6.8 6.8 6.6 6.6 6.5 7 6.0 6 5 4 3 2 1 0 The mean score is based on a scale from 1 to 10, where 1 means very poor quality and 10 means very high quality. The EU average Note: is unweighted. European Quality of Life Survey 2016. Source: 1 2 http://dx.doi.org/10.1787/888933835763 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 151

154 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE MORTALITY FOLLOWING ACUTE MYOCARDIAL INFARCTION (AMI) were initially admitted. The differences between upper Mortality due to coronary heart diseases has and lower quartile rates are largest in Latvia (over declined substantially over the past few decades (see 7 deaths per 100 admissions between different indicator “Mortality from circulatory diseases” in hospitals) and the smallest in Sweden (about 2 deaths Chapter 3). Important advances in both public health per 100 admissions). policies, including reductions in smoking and improved treatment for heart diseases, have Multiple factors contribute to variations in contributed to these declines (OECD, 2015). Clinical outcomes of care across hospitals, including hospital practice guidelines such as those developed by the structure, processes of care and organisational culture. European Society of Cardiology have helped optimise In Sweden, a system of evaluating and reporting quality treatment. Despite these advances, acute myocardial and outcomes of care is likely to have contributed to the infarction (AMI or heart attack) remains the leading small variation in mortality of patients after an AMI cause of cardiovascular deaths across European (Chung et al., 2015). countries, making further improvements a priority. A good indicator of acute care quality is the 30-day AMI mortality rate after hospital admission. The Definition and comparability measure reflects the processes of care, such as timely transport of patients and effective medical The thirty-day mortality rate measures the interventions. However, the indicator is influenced not percentage of people aged 45 and over who died only by the quality of care provided in hospitals but within 30 days following admission to hospital for also differences in hospital transfers, average length of an AMI (heart attack). Rates based on unlinked stay and AMI severity. data refer to a situation where the death occurred in the same hospital as the initial admission. Rates Figure 6.9 shows mortality rates within 30 days of based on linked data refer to a situation where the admission to hospital for AMI using unlinked data to death occurred in the same hospital, a different measure where the death occurs in the same hospital. hospital, or out of hospital. Rates are age-sex Across EU countries, the lowest rates (below 4.5%) are standardised to the 2010 OECD population aged found in Denmark and Sweden. The rate is also low in 45+ admitted to hospital for AMI (ICD-10 I21, I22). Poland but this is because the data refer mainly to patients admitted to cardiology wards while about 65% The specific methodology used to calculate the of patients with AMI are admitted to other wards. The hospital mortality rates presented in Figure 6.11 highest rates are in Latvia and Estonia. differs from that used for Figure 6.9 and Figure 6.10 and is likely to vary from the methods Using linked data, Figure 6.10 shows 30-day used by country for national monitoring and mortality rates where fatalities are recorded regardless reporting purposes. Different analytical methods of where they occur (in the hospital where the patient can result in quite different rates for and rankings was initially admitted, after transfer to another hospital of organisations and countries, limiting the or after discharge). This is a more robust indicator comparability of results. For more details on the because it records deaths more widely than the same- methodology used to calculate data presented in hospital indicator, but it requires a unique patient Figure 6.11, see Brownwood et al. (forthcoming). identifier and linked data which are not available in all countries. Using linked data, the AMI mortality rates range from less than 8% in Italy, Denmark and Sweden to over 14% in Latvia and Estonia. Thirty-day mortality rates for AMI have decreased References substantially between 2005 and 2015. Across the 20 EU countries for which data are available, they fell by 30% Brownwood, I. et al. (forthcoming), “OECD Hospital Performance Project: Methodological Development of (from 9.7% to 6.8%) when considering deaths occurring International Measurement of Acute Myocardial only in the hospital where patients were initially Infarction 30-Day Mortality Rates at the Hospital Level”, admitted and by over 25% (from 12.8% to 9.5%) in the , OECD Publishing, Paris. OECD Health Working Papers smaller group of countries providing data on deaths occurring in and out of hospital. Better access to high- Chung, S.C. et al. (2015), “Comparison of Hospital Variation quality acute care for heart attack, including timely in Acute Myocardial Infarction Care and Outcome transportation of patients, evidence-based medical between Sweden and United Kingdom: Population interventions and specialised health facilities such as Based Cohort Study Using Nationwide Clinical percutaneous catheter intervention-capable centres have Registries”, British Medical Journal ,Vol.351, http:// helped to reduce 30-day mortality rates (OECD, 2015). doi.org/10.1136/bmj.h3913 . Figure 6.11 presents the differences in dispersion Cardiovascular Disease and Diabetes: Policies for OECD (2015), of AMI 30-day mortality rates across hospitals within Better Health and Quality of Care , OECD Health Policy countries based on data which include deaths Studies, OECD Publishing, Paris, http://dx.doi.org/ occurring outside of these hospitals where patients 10.1787/9789264233010-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 152

155 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.9. Thirty-day mortality after admission to hospital for AMI based on unlinked data, 2005 and 2015 (or nearest years) Confidence Interval 2005 2015 Age-sex standardised rate per 100 admissions of adults aged 45 years and over 15 13.4 10 10.6 8.8 8.6 7. 9 7. 9 7.7 7. 4 7. 3 7.1 7. 0 6.9 6.8 6.4 6.4 5.9 5 6.1 5.6 5.6 5.4 5.4 5.1 4.4 4.2 4.0 3.7 0 Italy EU20 Spain Latvia Malta¹ Turkey France Poland Ireland Austria Finland Estonia Norway Sweden Iceland¹ Belgium Slovenia Portugal Denmark Germany Czech Rep. Slovak Rep. Switzerland Netherlands Luxembourg¹ United Kingdom 1. Three-year average. Note: 95% confidence intervals for the latest year are represented by grey areas. The EU average is unweighted and only includes countries with data covering the whole time period. https://doi.org/10.1787/health-data-en . Source: OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933835782 6.10. Thirty-day mortality after admission to hospital for AMI based on linked data, 2005 and 2015 2015 2005 Confidence Interval Age-sex standardised rate per 100 patients aged 45 years and over 20 18.0 14.1 15 13.7 10 10.9 10.7 9.9 9.7 9.5 9.0 8.9 8.8 8.3 8.2 8.2 8.1 7. 8 7.7 7.7 7. 6 5 0 Italy EU12 Spain Latvia Malta² Turkey Poland Finland Estonia Norway Sweden Slovenia Portugal Denmark Czech Rep. Switzerland Netherlands Luxembourg¹ United Kingdom 1. Three-year average. 2. Two-year average. 95% confidence intervals for the latest year are represented by grey areas. The EU average is unweighted and only includes Note: countries with data covering the whole time period. https://doi.org/10.1787/health-data-en Source: . OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933835801 6.11. Thirty-day mortality after admission to hospital for AMI based on linked data, 2013-15 (or nearest years) Age, sex, co−morbidity standardised mortality rates 40 30 20 10 0 Slovenia Denmark Italy Sweden Finland United Kingdom Norway Latvia Note: The width of each line in the figure represents the number of hospitals (frequency) with the corresponding rate. The data for the United Kingdom relate to England only and are presented at trust-level (i.e. multiple hospitals). The countries are ranked by interquartile range of mortality rate. Source: OECD Hospital Performance Data Collection 2017. 1 2 http://dx.doi.org/10.1787/888933835820 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 153

156 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE MORTALITY FOLLOWING STROKE Across EU countries, some 610 000 stroke events that occur not just in the same hospital but also in occurred in 2015 and the number is expected to rise by other hospitals and out of hospital. one-third by 2035 due to population ageing and Between 2005 and 2015, 30-day mortality rates for increases in some risk factors (King’s College London, ischaemic stroke have decreased in nearly all countries 2017). Stroke is the second leading cause of death after (and by over 25% on average), with the exception of heart disease (see the indicator “Mortality from Latvia where the rates have increased when considering circulatory diseases” in Chapter 3), and is also the fatalities in and out of hospital, although this may second leading cause of disability after depression. reflect improved data accuracy (OECD, 2016). The A stroke occurs when the blood supply to a part of reduction in 30-day mortality rates was substantial in the brain is interrupted. Of the two types of stroke that Denmark and the United Kingdom. Across European exist, about 85% are ischaemic (caused by clotting) and countries, better access to high-quality stroke care, 15% are haemorrhagic (caused by bleeding). including timely transportation of patients, evidence- Pneumococcal infections and influenza infections, both based medical interventions and high-quality vaccine-preventable, have a marked effect on triggering specialised facilities such as stroke units have helped to strokes. Treatment for ischaemic stroke has advanced reduce 30-day mortality rates (OECD, 2015). dramatically over the last decades with systems and Despite the progress so far, there is still room to processes now in place in many European countries, improve implementation of best practice acute care for which include specialised stroke units that are devoted stroke and other cardiovascular diseases across countries. to care for stroke patients by a multidisciplinary team, Targeted strategies can be highly effective to shorten and medical progress such as thrombolysis and acute care treatment time. Advances in technology are thrombectomy. now leading to models of care to deliver reperfusion Figure 6.12 shows the mortality rates within therapy in an even more rapid and efficient manner, 30 days of admission for ischaemic stroke using whether through pre-hospital triage via telephone, unlinked data to measure deaths occurring in the same administration via telemedicine, or administering the hospital. Using linked data, Figure 6.13 shows the therapy in the ambulance (Chang and Prabhakaran, 2017). mortality rate where deaths are recorded regardless of where they occurred (in the hospital admitted initially, after transfer to another hospital or after discharge). Definition and comparability This indicator is more robust because it takes account Thirty-day mortality rates are defined in the of hospital transfers and captures fatalities more indicator “Mortality following acute myocardial comprehensively. Although more countries report the infarction” in Chapter 6. Rates are age-sex same-hospital measure using unlinked data, an standardised to the 2010 OECD population aged increasing number of countries are investing in their 45+ admitted to hospital for ischaemic stroke data infrastructure and using linked data to provide (ICD-10 I63-I64). more comprehensive measures. Across EU countries, 8.6% of patients admitted for ischaemic stroke in 2015 died within 30 days in the same hospital in which the initial admission for References ischaemic stroke occurred (Figure 6.12). Thirty-day Chang, P. and S. Prabhakaran (2017), “Recent Advances in mortality rates were highest in Latvia (18.3%), Malta the Management of Acute Ischaemic Stroke”, (15.9%) and Lithuania (15.3%). Rates were less than 5% in , F1000Research http://doi.org/ , F1000 Faculty Rev-484, 6 Denmark and Finland. Generally, countries that have 10.12688/f1000research.9191.1 . 30-day mortality for ischaemic stroke lower than the EU King’s College London (2017), The burden of stroke in Europe: average also tend to have lower 30-day mortality rates , Stroke Alliance for The challenge for policy makers for acute myocardial infarction (AMI) (see indicator www.strokeeurope.eu/downloads/The_ Europe, London, “Mortality following acute myocardial infarction”). This Burden_of_Stroke_in_Europe_-_Challenges_for_policy_ suggests that certain aspects of acute care may be makers.pdf . influencing outcomes for both stroke and AMI patients. Cardiovascular Disease and Diabetes: Policies for OECD (2015), Across those EU countries that reported in- and Better Health and Quality of Care , OECD Publishing, out-of-hospital mortality rates, 11.7% of patients died Paris, http://dx.doi.org/10.1787/9789264233010-en . within 30-days of being admitted to hospital for stroke OECD (2016), OECD Reviews of Health Systems: Latvia 2016 , (Figure 6.13). This figure is higher than the same- http://dx.doi.org/10.1787/ OECD Publishing, Paris, hospital based indicator because it captures deaths . 9789264262782-en HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 154

157 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.12. Thirty-day mortality after admission to hospital for ischaemic stroke based on unlinked data, 2005 and 2015 (or nearest years) 2015 Confidence Interval 2015 2005 Age-sex standardised rate per 100 admissions of adults aged 45 years and over 20 18.3 15 15.9 15.3 10.1 12.1 10 10.1 9.9 9.7 9.6 9.7 9.2 8.6 8.4 8.6 8.3 8.1 7. 3 7.1 6.8 6.3 5 6.2 6.1 5.4 5.0 4.8 4.6 0 Italy EU18 Spain Latvia Malta¹ Turkey France Ireland Austria Finland Estonia Norway Sweden Iceland¹ Belgium Slovenia Portugal Denmark Germany Lithuania Czech Rep. Slovak Rep. Switzerland Netherlands Luxembourg¹ United Kingdom 1. Three-year average. 95% confidence intervals for the latest year are represented by grey areas. The EU average is unweighted and only includes Note: countries with data covering the whole time period. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, . Source: 1 2 http://dx.doi.org/10.1787/888933835839 6.13. Thirty-day mortality after admission to hospital for ischaemic stroke based on linked data, 2005 and 2015 (or nearest years) 2005 2015 Confidence Interval 2015 Age-sex standardised rate per 100 patients aged 45 years and over 30 25.7 25 19.9 20 17.3 15 15.0 14.1 13.1 10 11.7 10.8 10.9 10.6 10.2 10.1 9.6 9.4 9.3 8.4 8.2 7.7 5 0 Italy EU10 Spain Latvia Malta² Turkey Finland Estonia Norway Sweden Slovenia Portugal Denmark Czech Rep. Switzerland Netherlands Luxembourg¹ United Kingdom 1. Three-year average. 2. Two-year average. Note: 95% confidence intervals for the latest year are represented by grey areas. The EU average is unweighted and only includes countries with data covering the whole time period. OECD Health Statistics 2018, Source: https://doi.org/10.1787/health-data-en . 1 2 http://dx.doi.org/10.1787/888933835858 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 155

158 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE WAITING TIMES FOR HIP FRACTURE SURGERY The main risk factors for hip fractures are In Portugal, the proportion of patients operated associated with ageing, including an increased risk of within two days after a hip fracture has decreased from falling and loss of skeletal strength from osteoporosis. 57% in 2008 to 47% in 2015, despite greater efforts to With increasing life expectancy, it is anticipated that monitor this performance target at the hospital level hip fractures will become an even greater public health and the provision of financial incentives to achieve issue in the coming years. more timely hip fracture repairs (OECD, 2015). In nearly all instances following a hip fracture, The waiting time for surgery after a hip fracture surgical intervention is required to repair or replace the is influenced by many factors, including hospitals’ hip joint. There is general agreement that early surgical surgical theatre capacity and flow, and targeted policy intervention maximises patient outcomes and interventions, including public reporting and minimises the risk of complications, and that monitoring of performance (Siciliani et al., 2013). surgery should occur within two days (48 hours) of hospitalisation. The guidelines in some countries stipulate even more rapid intervention. For example, in Definition and comparability the United Kingdom, the National Institute for Health This indicator is defined as the proportion of and Care Excellence (NICE) clinical guidelines patients aged 65 years and over admitted to recommend hip fracture surgery to be performed on the hospital for an upper femur fracture, who had day of hospital admission or the next day (NICE, 2017). surgery initiated within two calendar days of On average across EU countries, more than three their admission. Data are also provided for the quarters (77%) of patients aged 65 and over admitted proportion of patients who had surgery within forahipfracturewereoperatedwithintwodaysin one day of their admission to hospital, and for 2015, with most of them being treated in fact either on patients who had surgery on the same day as the same day of their admission or the next day their admission. (Figure 6.14). In Denmark and the Netherlands, the The capacity to capture time of admission and proportion of patients operated within two days was surgery in hospital administrative data varies greater than 95%. By contrast, only about half of across countries, resulting in the inability to patients aged 65 and over were operated within two precisely record surgery within 48 hours. Recent days following their admission for a hip fracture in research and development data indicate that the Latvia, Portugal, Spain and Italy in 2015. impact of measuring days rather than hours However, substantial progress has been achieved may result in marginally higher rates. over the past 10 years in Italy and Spain in meeting the recommended clinical guideline of operating patients within two days following a hip fracture (Figure 6.15). In Italy, this proportion nearly doubled from 28% in 2005 to 53% in 2015, whereas it increased from 36% to 48% in References Spain. Remarkable improvement also occurred in NICE (2017), “Hip Fracture: The Management of Hip Fracture Switzerland, where the proportion doubled from 46% NICE Clinical Guideline No. 124 , issued June in Adults”, to 91%. 2011, last updated May 2017. In Italy, the progress in providing more rapid OECD (2014), OECD Reviews of Health Care Quality: Italy surgical treatment for patients admitted with a hip http:// , OECD Publishing, Paris, 2014: Raising Standards fracture was mainly achieved by reducing the waiting . dx.doi.org/10.1787/9789264225428-en time in those regions and hospitals that were lagging OECD Reviews of Health Care Quality: Portugal OECD (2015), behind a decade ago. Italian authorities implemented a http:// , OECD Publishing, Paris, 2015: Raising Standards policy of public reporting of hospital performance . dx.doi.org/10.1787/9789264225985-en indicators that included the waiting time for surgery for Siciliani, L., M. Borowitz and V. Moran (eds.) (2013), patients admitted with a hip fracture, which helped to Waiting Time Policies in the Health Sector: What Works?, identify those regions and hospitals that were falling http://dx.doi.org/10.1787/ OECD Publishing, Paris, short of meeting the recommended target (OECD, 2014). . 9789264179080-en HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 156

159 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.14. Hip fracture surgery initiation after admission to the hospital, 2015 (or nearest year) Same day Next day Day Two % of patients aged 65 years and over 100 96 96 95 91 92 92 90 88 87 87 90 83 82 81 77 80 75 69 67 70 60 53 48 47 46 50 40 30 20 10 0 1. Sweden provided data within 12, 24 and 48 hours. 2. Three-year average. Note: The EU average is unweighted. Source: . https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, http://dx.doi.org/10.1787/888933835877 1 2 6.15. Hip fracture surgery initiation after admission to hospital, 2005 and 2015 (or nearest year) 2015 2005 % of patients aged 65 years and over being operated within 2 days ες ερ 100 εϮ εϭ εϮ εϬ ΘΘ Θϳ Θϳ ες 90 Θϯ ΘϮ Θϭ ΘϬ ϳρ 80 ςε ςϳ 70 60 ρϯ κΘ κς 50 40 κϳ 30 20 10 0 1. Three-year average. Note: The EU average is unweighted and only includes countries with data covering the whole time period. https://doi.org/10.1787/health-data-en Source: OECD Health Statistics 2018, . 1 2 http://dx.doi.org/10.1787/888933835896 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 157

160 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE SCREENING, SURVIVAL AND MORTALITY FOR CERVICAL CANCER More than 100 000 women in EU countries are diagnosed each year with cervical cancer (see Definition and comparability indicator on “Cancer incidence” in Chapter 3). Cervical Screening rates are based on programme or survey cancer is highly preventable if precancerous cells are data. Programme data are collected to monitor detected and treated before progression occurs. The national screening programmes, but differences in human papilloma virus (HPV) is found in over 90% of target population and screening frequency may lead cervical cancers (European Commission, 2018), and to variations in the data reported across countries. vaccination against the main types of HPV Survey data may be affected by recall bias. responsible for cervical cancer is expected to reduce Five-year net survival is the cumulative probability incidence. that cancer patients survive their cancer for at least European countries follow various approaches to 5 years, after controlling for the risks of death from the prevention and early diagnosis of cervical cancer. other causes. Net survival is expressed as a percentage Over half of the countries have implemented in the range of 0-100%. Five-year net survival for population-based cervical cancer screening programmes patients diagnosed during 2000-04 is based on a (IARC, 2017). WHO recommends HPV vaccination for cohort approach, since all patients have been followed girls aged 9-13 years (WHO, 2018). Most European up for at least 5 years. For patients diagnosed during countries now have national HPV vaccination 2010-14, a period approach was used, allowing estimation of 5-year survival when complete 5-year programmes, but the target populations vary, based on follow-up data were not yet available for all patients. epidemiological and other evidence such as cost- Survival estimates are age-standardised with the effectiveness that is specific to each country (ECDC, International Cancer Survival Standard weights. 2014). Vaccination for boys is also considered effective when coverage for girls is low. Data collection, quality control and analysis were performed as part of the CONCORD programme, the On average, the proportion of women in EU global programme for the surveillance of cancer countries aged 20-69 years who have been screened survival (Allemani et al., 2018). In some countries, for cervical cancer within the past three years has not all regional registries participated. Survival increased from 56% to 61% over the past decade. estimates for cervical cancer are based on the However, the proportion has fallen in several International Classification of Diseases for Oncology countries. The proportion of screened women across (ICD-O-3 C53.0-C53.1 and C53.8-C53.9). EU countries still varies widely, from about 25% only Chapter 3 See indicator “Mortality from cancer” in in Latvia and Romania to over 80% in Austria and for the definition of cancer mortality rates. Mortality Sweden (F igure 6.16). from cervical cancer is based on ICD-10 C53. Cancer survival is one of the key measures of the effectiveness of health care systems in managing cancer, reflecting both early detection and the effectiveness of treatment. Among women diagnosed References with cervical cancer between 2010 and 2014, age- Allemani, C. et al. (2018), “Global surveillance of trends in standardised five-year net survival ranged from 70% in cancer survival 2000-14 (CONCORD-3): Analysis of Denmark to 54% in Latvia (Figure 6.17). The average individual records for 37 513 025 patients diagnosed with among EU countries has increased from 61% to 63% one of 18 cancers from 322 population-based registries in over the past decade. The variation across countries , Vol. 391 (10125), pp. 1023-1075, 71 countries”, The Lancet has decreased, because some of the countries that had . http://dx.doi.org/10.1016/S0140-6736(17)33326-3 among the lowest survival have converged to some European Commission (2018), Human papilloma virus (HPV) , extent towards the best performers. https://ec.europa.eu/ European Commission, Brussels, Trends in cervical cancer mortality rates reflect . health/vaccination/hpv_en the underlying trends in incidence and survival. The HPV vaccination in EU countries: Review of new ECDC (2014), mortality rates for cervical cancer have declined https://ecdc.europa.eu/en/news-events/hpv- , evidence across EU countries from 6.0 per 100 000 women in vaccination-eu-countries-review-new-evidence . 2000 to 5.1 in 2015 (F igure 6.18). IARC (2017), Cancer Screening in the European Union – Report on However, in many Central and Eastern European the implementation of the Council Recommendation on cancer countries, cervical cancer screening rates are low, https:// , European Commission, Brussels, screening incidence has not yet declined, five-year net survival ec.europa.eu/health/sites/health/files/major_ chronic_diseases/ remains low and mortality is still high or even rising. . docs/2017_cancerscreening_2ndreport implementation_en.pdf These trends suggest the need for greater policy attention to prevention, early diagnosis and effective WHO (2018), Human papillomavirus (HPV) and cervical cancer – treatment for cervical cancer. www.who.int/en/news-room/fact- ,WHO,Geneva, Key facts . sheets/detail/human-papillomavirus-(hpv)-and-cervical-cancer HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 158

161 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.16. Cervical cancer screening in women aged 20-69 within the past 3 years, around 2006 and around 2016 2016 2006 % of women screened 100 87 82 80 80 77 90 76 76 75 73 72 72 71 70 69 80 61 70 54 77 50 74 60 46 48 68 64 41 65 50 34 32 60 54 40 26 25 46 30 35 11 20 5 10 0 1. Programme. 2. Survey. 3. Three-year average. The EU average is unweighted and only includes countries with data covering the whole time period. Note: . https://doi.org/10.1787/health-data-en Source: OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933835915 6.17. Cervical cancer five-year net survival, 2000-04 and 2010-14 2010-14 2000-04 Confidence Interval 2010-14 Age-standardised net survival, % 90 80 73 71 80 70 67 67 68 68 66 65 70 65 65 64 65 65 63 64 63 61 67 59 57 66 64 60 54 55 55 61 61 50 40 Italy EU23 Spain Turkey France Latvia¹ Malta¹ ³ Poland¹ Ireland¹ Croatia¹ Austria¹ Finland¹ Estonia¹ Norway¹ Sweden¹ Belgium¹ Romania Bulgaria¹ Germany Slovenia¹ Portugal¹ Iceland¹ ² Denmark¹ Lithuania¹ Czech Rep.¹ Slovak Rep. Switzerland Netherlands¹ United Kingdom¹ 1. Data with 100% coverage of the national population. 2. Data not age-standardised. 3. Data for 2000-04 not age-standardised. 95% confidence intervals have been calculated for all countries, represented by grey areas. The EU average is unweighted and only Note: includes countries with data covering the whole time period. CONCORD programme, London School of Hygiene and Tropical Medicine. Source: 1 2 http://dx.doi.org/10.1787/888933835934 6.18. Cervical cancer mortality in women, 2000 and 2015 2000 2015 Age-standardised rates per 100 000 women 20 18 16 16 14 12 10 9 10 12 8 11 9 5 8 8 6 8 4 7 6 4 1 5 4 4 2 3 3 3 3 3 3 3 3 3 3 2 2 3 2 2 2 2 2 0 1. Three-year average. EU average for 2000 has been calculated by the OECD. Note: Eurostat Database. Source: http://dx.doi.org/10.1787/888933835953 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 159

162 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE SCREENING, SURVIVAL AND MORTALITY FOR BREAST CANCER Breast cancer is the most frequent cancer among following the introduction of a breast cancer screening women across EU countries, with more than 400 000 programme and a National Cancer Control Programme cases diagnosed each year across EU countries (see in the early 2000s (OECD, 2014). Survival also increased indicator “Cancer incidence” in Chapter 3). The main strongly in Denmark, Malta, Portugal and the United risk factors for breast cancer are age, genetic Kingdom. predisposition, oestrogen replacement therapy, and Mortality from breast cancer has fallen in most EU lifestyle factors including obesity, physical inactivity, countries since 2000. On average across EU countries, nutrition habits and alcohol consumption. the age-standardised rates of mortality from breast Most European countries have adopted breast cancer fell from 39 to 33 per 100 000 women per year cancer screening programmes as an effective way for between 2000 and 2015 (Figure 6.21). Particularly strong detecting the disease early (OECD, 2013; IARC, 2017). reductions occurred in Denmark and Malta, although However, due to recent progress in treatment these countries still have higher age-standardised outcomes and concerns about false-positive results, mortality rates. Croatia is one of the few EU countries over-diagnosis and overtreatment, breast cancer where breast cancer mortality rate has increased since screening recommendations have been re-evaluated in 2000 and now has the highest mortality rates of all EU recent years. WHO now recommends organised countries. population-based mammography screening for women aged between 50 and 69 in EU countries, if specific criteria are met such as whether women are Definition and comparability able to make an informed decision based on the Screening coverage and survival are defined in benefits and risks of mammography screening. Other the indicator “Screening, survival and mortality criteria are related to quality assurance and monitoring for cervical cancer”. Survival estimates for breast and evaluation mechanisms (WHO, 2014). cancer are based on the International The proportion of women in the EU aged 50-69 Classification of Diseases for Oncology (ICD-O-3 who have been screened for breast cancer within the C50.0-C50.6 and C50.8-C50.9). See indicator past two years is lowest in Romania, Bulgaria and “Mortality from cancer” in Chapter 3 for the Latvia, and highest in Nordic countries (Sweden, definition of cancer mortality rates. Mortality from Finland and Denmark) and Portugal (Figure 6.19). On breast cancer is based on ICD-10 C50. average across EU countries, the proportion of screened women increased from 54% to 58% between 2006 and 2016. A large increase has occurred in some countries that had a low screening rate a decade ago, References such as Lithuania, Poland and the Czech Republic. However, breast cancer screening rate has decreased IARC (2015), “IARC Handbooks of Cancer Prevention: substantially over the past decade in several countries, Benefits of mammography screening outweigh adverse Press release effects for women aged 50-69 years”, likely due partly to concerns over potential harms www.iarc.fr/en/media-centre/pr/2015/pdfs/ ,Lyon, No. 234 related to mammography screening, although a . pr234_E.pdf number of studies have found that the benefits IARC (2017), Cancer Screening in the European Union – Report overweigh the potential risks (IARC, 2015). on the implementation of the Council Recommendation on Breast cancer survival reflects early diagnosis as cancer screening , European Commission, Brussels, well as effective treatment. All Western European https://ec.europa.eu/health/sites/health/files/major_ countries have attained five-year net survival of at chronic_diseases/docs/2017_cancerscreening_2ndreport least 80%, but survival is still lower in several Central implementation_en.pdf . and Eastern European although it has increased in OECD (2013), Cancer Care: Assuring Quality to Improve recent years (Figure 6.20). http://dx.doi.org/ , OECD Publishing, Paris, Survival 10.1787/9789264181052-en . Over the last decade, the five-year net survival improved from 79% to 83% on average across EU OECD Reviews of Health Care Quality: Czech OECD (2014), , OECD Publishing, Republic 2014: Raising Standards countries. Net survival increased particularly rapidly in . Paris, http://dx.doi.org/10.1787/9789264208605-en Bulgaria, the Czech Republic, Estonia, Latvia and Lithuania, converging towards the level of other EU , WHO Position Paper on Mammography Screening WHO (2014), countries. In the Czech Republic, survival improved WHO, Geneva. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 160

163 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.19. Mammography screening in women aged 50-69 within the past 2 years, around 2006 and around 2016 2016 2006 % of women screened 100 82 90 84 80 90 80 62 60 82 60 61 60 70 59 58 78 56 77 75 75 75 51 73 46 60 47 60 50 35 33 57 31 54 27 40 52 21 42 30 12 6 20 10 0 1. Programme. 2. Survey. 3. Three-year average. Note: The EU average is unweighted and only includes countries with data covering the whole time period. . OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en Source: http://dx.doi.org/10.1787/888933835972 1 2 6.20. Breast cancer five-year net survival, 2000-04 and 2010-14 2010-14 Confidence Interval 2010-14 2000-04 Age-standardised net survival, % 100 89 87 89 89 88 87 86 90 87 86 86 86 86 85 86 85 84 83 82 81 87 79 78 77 77 76 80 75 77 82 74 70 60 50 Italy EU23 Spain Malta¹ Turkey France Latvia¹ Poland¹ Ireland¹ Croatia¹ Austria¹ Iceland¹ Finland¹ Estonia¹ Norway¹ Sweden¹ Romania Belgium¹ Bulgaria¹ Germany Slovenia¹ Portugal¹ Denmark¹ Lithuania¹ Czech Rep.¹ Slovak Rep. Switzerland Netherlands¹ United Kingdom¹ 1. Data with 100% coverage of the national population. 95% confidence intervals have been calculated for all countries, represented by grey areas. The EU average is unweighted and only Note: includes countries with data covering the whole time period. CONCORD programme, London School of Hygiene and Tropical Medicine. Source: 1 2 http://dx.doi.org/10.1787/888933835991 6.21. Breast cancer mortality in women, 2000 and 2015 2000 2015 Age-standardised rates per 100 000 women 60 50 43 41 40 40 35 33 32 32 43 39 38 37 37 37 37 37 36 30 35 34 33 33 35 33 33 32 32 31 31 30 15 29 20 27 24 27 26 23 10 16 0 1. Three-year average. EU average for 2000 has been calculated by the OECD. Note: Source: Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933836010 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 161

164 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE SURVIVAL AND MORTALITY FOR COLORECTAL CANCER Colorectal cancer is the second most common Nonetheless, differences across countries in cause of cancer deaths after lung cancer among men, survival following a diagnosis for colon and rectal and the third most common cause of cancer deaths cancer is larger than for other types of cancer, such as after breast and lung cancers among women across EU cervical and breast cancer. This indicates that there is countries (see indicator “Mortality from cancer” in still large room for improvements in early detection Chapter 3). The main risk factors for colorectal cancer and treatment in countries (mainly in Central and include age, ulcerative colitis, a personal or family Eastern Europe) that are lagging behind. history of colorectal cancer or polyps, and lifestyle Looking at overall mortality rates from colorectal factorssuchasadiethighinfatandlowinfibre, cancer, they fell by over 10% on average across EU physical inactivity, obesity, tobacco and alcohol countries between 2000 and 2015 (Figure 6.24). The consumption. The incidence of colorectal cancer is decline was particularly large in Austria, Belgium, the significantly higher among men. Generally, rectal Czech Republic and Germany with a reduction of over cancer is more difficult to treat than colon cancer due 30% in age-standardised mortality rates. However, to a higher probability of spreading to other tissue, mortality rates from colorectal cancer have increased recurrence and postoperative complications. in some countries, notably in Romania and Croatia, Following screening programmes for cervical and reflecting higher incidence. breast cancers, a growing number of countries have introduced free population-based colorectal cancer screening programmes over the past few years, Definition and comparability targeting people in their 50s and 60s (OECD, 2013). In Net survival is defined in the indicator most countries that use the faecal occult blood test, “Screening, survival and mortality for cervical screening is available every two years. The screening cancer”. See the indicator “Mortality from cancer” schedule is less frequent with colonoscopy and flexible in Chapter 3 for the definition of cancer mortality sigmoidoscopy, generally every ten years (IARC, 2017). rates. Mortality rates from colorectal cancer are These differences complicate international based on ICD-10 codes C18-C21 (colon, comparisons of screening coverage. Based on survey rectosigmoid junction, rectum, and anus) while data collected in 2014, less than half of people aged 50 survival estimates are based on C18-C19 for colon to 74 in EU countries reported having ever been cancer and C20-C21 for rectum cancer. screened for colorectal cancer through a faecal occult blood test (Eurostat, 2017). Advances in diagnosis and treatment of colorectal cancer, including improved surgical techniques such as mesorectal excision, radiation therapy and combined References chemotherapy, and wider and more timely access, Eurostat (2017), Cancer statistics – specific cancers, Statistics have contributed to increased survival over the last Explained, September 2017. decade. On average across EU countries, five-year net IARC (2017), Cancer Screening in the European Union – Report survival for colon cancer improved from 54% to 60% on the implementation of the Council Recommendation on between 2000-04 and 2010-14, and from 52% to 58% for , European Commission, Brussels, cancer screening rectal cancer over the same period (Figure 6.22 and https://ec.europa.eu/health/sites/health/files/major_ Figure 6.23). Survival for colon cancer increased chronic_diseases/docs/2017_cancerscreening_2ndreport implementation_en.pdf . particularly rapidly in Denmark, Estonia, Latvia, Lithuania and Slovenia, and the same countries along OECD (2013), , Cancer Care: Assuring Quality to Improve Survival with Ireland achieved the biggest progress in survival http://dx.doi.org/10.1787/ OECD Publishing, Paris, for rectal cancer. . 9789264181052-en HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 162

165 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.22. Colon cancer five-year net survival, 2000-04 and 2010-14 2010-14 2000-04 Confidence Interval 2010-14 Age-standardised net survival, % 80 68 68 67 65 65 65 65 64 64 64 63 63 62 62 61 61 60 60 58 58 57 56 55 60 53 52 52 51 49 40 20 Italy EU23 Spain Malta¹ Turkey France Latvia¹ Poland¹ Ireland¹ Austria¹ Croatia¹ Finland¹ Iceland¹ Estonia¹ Norway¹ Sweden¹ Belgium¹ Bulgaria¹ Germany Slovenia¹ Portugal¹ Denmark¹ Lithuania¹ Czech Rep.¹ Switzerland Slovak Rep.¹ Netherlands¹ United Kingdom¹ 1. Data with 100% coverage of the national population. Note: 95% confidence intervals have been calculated for all countries, represented by grey areas. The EU average is unweighted. CONCORD programme, London School of Hygiene and Tropical Medicine. Source: 1 2 http://dx.doi.org/10.1787/888933836029 6.23. Rectal cancer five-year net survival, 2000-04 and 2010-14 2010-14 Confidence Interval 2010-14 2000-04 survival, % ndardised net Age-sta 80 68 67 67 65 65 65 64 64 63 62 62 61 61 60 60 60 58 58 55 56 60 53 53 52 63 50 49 48 48 46 40 20 ¹ ¹ ¹ y d a¹ d¹ 23 tvia Ital EU Spain Malt Tur key rtugal France La Poland¹ Irelan Croatia¹ Austria¹ Finland¹ Iceland¹ Estonia¹ Norway¹ Sweden¹ Belgium¹ Romania Bulgaria¹ Germany Slovenia¹ Po Denmark¹ Lithuania¹ Czech Rep.¹ Switzerlan Slovak Rep. Netherlands¹ United Kingdom¹ 1. Data with 100% coverage of the national population. 95% confidence intervals have been calculated for all countries, represented by grey areas. The EU average is unweighted. Note: CONCORD programme, London School of Hygiene and Tropical Medicine. Source: 1 2 http://dx.doi.org/10.1787/888933836048 6.24. Colorectal cancer mortality, 2000 and 2015 2000 2015 Age-sex standardised rates per 100 000 population 70 60 50 50 38 54 36 35 35 34 34 34 40 49 40 38 23 30 19 37 35 36 33 33 33 32 30 20 29 29 28 28 27 27 26 26 26 25 23 19 22 6 10 0 1. Three-year average. Note: EU average for 2000 has been calculated by the OECD. Eurostat Database. Source: http://dx.doi.org/10.1787/888933836067 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 163

166 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE LATE-DIAGNOSED HIV AND TUBERCULOSIS TREATMENT OUTCOMES available. Multi-drug resistant tuberculosis requires The prevention and management of infectious longer and more intensive treatment and is associated diseases such as the Human Immunodeficiency Virus with lower success rates. (HIV) and tuberculosis remain a high priority in many European countries. The EU is committed to play an Figure 6.27 shows the percentage of newly important role in achieving target 3.3 of the United diagnosis tuberculosis cases classified as being Nations’ Sustainable Development Goals, which is to Rifampicin-resistant or multi-drug resistant. While a end the epidemics of Acquired Immunodeficiency number of countries did not report any case, the Baltic Syndrome (AIDS), tuberculosis and other communicable countries (Estonia, Lithuania and Latvia) reported the diseases by 2030 (European Commission, 2018). highest proportions of multi-drug resistant cases in 2016. Although HIV is preventable through effective public health measures, significant HIV transmission continues in Europe with nearly 30 000 newly- Definition and comparability diagnosed cases reported in EU countries in 2016. In some countries such as Latvia and Malta, rates of HIV Late diagnosis of HIV cases is defined as patients transmission have increased in recent years (see with a CD4 cell count under 200 per mm3 of blood indicator “New reported cases of HIV/AIDS and at diagnosis (ECDC and WHO Regional Office for tuberculosis” in Chapter 3). Europe, 2017). Surveillance systems for HIV are not identical across Europe and differences in data HIV weakens the human immune system, leaving collection methods and testing policies can affect those affected vulnerable to infections and other data comparability. Official reports of newly health issues including tuberculosis or hepatitis C. The diagnosed cases of HIV do not represent true most advanced stage of HIV infection is AIDS. Early incidence. Newly reported HIV diagnoses include testing for HIV allows infected individuals to be put on recently infected individuals as well as those who treatment quickly leading to earlier viral suppression, were infected several years ago but only recently thus allowing them to continue to live a normal life tested for HIV.These reports are also influenced by and to avoid infecting others. several factors such as the uptake of HIV testing, Figure 6.25 shows the percentage of late diagnosis patterns of reporting, the long incubation period among newly diagnosed HIV cases in 2016. The and a slow progression of the disease. Czech Republic, the Slovak Republic and Belgium New tuberculosis cases include patients who report the lowest proportion of late diagnosis cases have never been treated for tuberculosis or have among newly diagnosed HIV infections, with taken anti-tuberculosis drugs for less than one percentages of 18% or less. The proportion in Romania, month. All pulmonary cases of tuberculosis have Greece, Italy, Lithuania and Malta is two-times greater been bacteriologically confirmed. Successful (at 36% or more). The high rates in some countries treatment outcomes are defined as the sum of: suggest that screening and testing services need to be 1) cured: a TB patient with bacteriologically- substantially improved to identify and treat HIV cases confirmed TB at the beginning of treatment who earlier, particularly among at-risk populations. was smear or culture-negative in the last month of Tuberculosis also remains an important public treatment and on at least one previous occasion; health issue in some European countries. Although the and 2) treatment completed, but does not meet the number of new cases of tuberculosis has generally criteria to be classified as cure or treatment failure declined over the past decade, further efforts are (a TB patient whose smear or culture is positive at needed to prevent the spread of this disease in some month five or later during treatment) (ECDC, 2018). countries (see indicator “New reported cases of HIV/ AIDS and tuberculosis” in Chapter 3). Figure 6.26 shows the percentage of new tuberculosis cases and relapses with successful treatment outcome after 12 months. Sweden, the References Netherlands, the Slovak Republic, Romania and Bulgaria Commission Staff Working European Commission (2018), have rates of 85% or more of successful treatment Document on Combatting HIV/AIDS, viral hepatitis and outcomes, while Croatia has the lowest success rate. tuberculosis in the European Union and neighbouring Success rates are driven by the availability of treatment countries – State of play, policy instruments and good programmes, patient adherence, and the proportion of , EC, Brussels. practices multi-drug resistant tuberculosis infections. ECDC and WHO Regional Office for Europe (2017), HIV/ Drug-resistant tuberculosis can occur when the , ECDC, AIDS surveillance in Europe 2017 – 2016 data drugs used to treat the condition are misused or Stockholm. mismanaged, including where people do not complete a full course of treatment, providers prescribe the ECDC (2018), Tuberculosis Surveillance and Monitoring in wrong treatment or where proper treatments are not , ECDC, Stockholm. Europe 2018 – 2016 data HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 164

167 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.25. Percentage of late diagnosis among newly diagnosed HIV cases, 2016 % 60 50 50 40 37 37 36 36 35 40 34 34 33 32 30 29 29 29 29 28 27 27 27 26 26 25 30 24 23 21 21 18 17 17 20 10 0 Minimum of 30 HIV cases needed for inclusion. EU average unweighted. Note: Source: ECDC (2017). 1 2 http://dx.doi.org/10.1787/888933836086 6.26. Percentage of treatment success after 12 months of new TB cases and relapses, 2015 % 100 93 89 88 87 87 87 86 86 86 85 85 84 83 81 79 79 79 73 73 72 71 80 69 65 59 58 57 55 60 42 40 20 0 1. Three-year average. Minimum of 30 TB cases needed for inclusion. EU average unweighted. Note: ECDC (2018). Source: http://dx.doi.org/10.1787/888933836105 1 2 6.27. Estimated percentage of notified new tuberculosis cases with multi-drug resistance, 2016 % 14 14 12 12 10 8 8 6 4 4 3 3 3 33 3 3 4 2 2 22 2 2 22 2 1 1 1 1 1 11 1 2 0 00000 0 Minimum of 30 TB cases needed for inclusion. EU average unweighted. Note: Source: ECDC (2018). http://dx.doi.org/10.1787/888933836124 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 165

168 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE HEALTHCARE-ASSOCIATED INFECTIONS The European Centre for Disease Control estimates that 3.8 million people acquire a healthcare-associated Definition and comparability infection each year in acute care hospitals in EU countries The data are based on a point prevalence survey and Norway and Iceland (Suetens et al., 2018), and an (PPS) of healthcare-associated infections conducted in estimated 90 000 people in the EU die each year due to the 2016-17 in 1 275 acute care hospitals covering all EU six most common infections in health care settings countries (except Denmark and Sweden), Norway, (Cassini, 2016). At least 20% of healthcare-associated Iceland and Serbia (Suetens et al., 2018). Validation infections are considered to be avoidable through better studies of national PPS data were carried out in a infection prevention and control (Harbath, 2003). subgroup of hospitals and generally found an Figure 6.28 shows the percentage of patients reported underestimation of the true prevalence, which allowed by selected hospitals in EU countries to have acquired a to make a more robust estimation of the burden of healthcare-associated infection in 2016-17, together with healthcare-associated infections. Different the predicted percentage of patients that would be expected sensitivities and specificities of infections’ detection to have acquired such an infection based on patient may explain, in part, differences between the observed characteristics. On average across EU countries (weighted), versus expected prevalence. Estimates were used for 5.5% of patients acquired an infection during their hospital Denmark and Sweden to come up with a total burden stay in 2016-17. The observed percentage was lowest in for the EU, Norway and Iceland as a whole, using EU Lithuania, Bulgaria, Germany, Latvia, the Netherlands and averages to the hospital discharge data for these two Romania (less than 4%), and highest in Greece, Portugal, countries. Norway participated in this survey with a Italy, Finland and Cyprus (more than 8%). protocol that required the imputation of data for Figure 6.29 shows the proportion of healthcare- missing types of infections. In Bulgaria and the associated infections by type of care (specialty). Across EU Netherlands, country representativeness is limited countries, patients in medical specialty areas (including because of a low number of participating hospitals, general medicine, cardiology, oncology, neurology) resulting in potential selection bias. accounted for 40% of all infection cases in 2016-17. Patients in surgical specialty areas represented another 33% of cases, while intensive care patients accounted for 13% of infections. Geriatrics, paediatrics and other specialty areas made up the remaining 14% of healthcare- References associated infections. Cassini, A. et al. (2016), “Burden of Six Healthcare- As shown in Figure 6.30, the most common types of Associated Infections on European Population Health: healthcare-associated infections were pneumonia Estimating Incidence-Based Disability-Adjusted Life (accounting for 26% of all cases), urinary tract infections Years through a Population Prevalence-Based Modelling (19%), surgical site infections (18%), bloodstream Study”, PLoS Med ., 18:13(10). infections (11%) and gastrointestinal infections (9%). ECDC (2018), “Directory of online resources for the Compounding the impact of healthcare-associated prevention and control of antimicrobial resistance infections are infections due to antimicrobial resistant (AMR) and healthcare-associated infections (HAI)”, bacteria, which can lead to complications, longer hospital https://ecdc.europa.eu/en/publications-data/directory- stays, or death. A single resistant infection has been online-resources-prevention-and-control-antimicrobial- estimated to cost about EUR 8 500 to 34 000 more than a . resistance-amr non-resistant infection, due to additional hospital days and Harbarth, S. et al. (2003), “The Preventable Proportion of additional treatment costs (OECD, 2017). Inappropriate use Nosocomial Infections: An Overview of Published of antibiotics contribute to antimicrobial-resistant bacteria JournalofHospitalInfection , Vol. 54, No. 4, Reports”, in hospitals and in the community. pp. 258-266. Healthcare-associated infections can be prevented by , OECD (2017), Tackling Wasteful Spending on Health implementing a series of measures, as set out in the http://dx.doi.org/10.1787/ OECD Publishing, Paris, Council of the European Union’s Recommendation on 9789264266414-en . Patient Safety, including the Prevention and Control of Suetens, C. et al. (2018), “Prevalence of healthcare-associated Healthcare-Associated Infections (2009/C 151/01). At the infections, estimated incidence, and composite hospital level, key components of effective infection antimicrobial resistance index in acute care hospitals prevention and control strategies include: the creation of a and long-term care facilities: Results from two European local infection control team; staff training; use of , Eurosurveillance point prevalence surveys, 2016 to 2017”, evidence-based guidelines; infection surveillance and Vol. 23, No. 46, 1800516. feedback; and rigorous maintenance of environmental hygiene (WHO, 2016). Most European countries have Guidelines on Core Components of Infection WHO (2016), established their own national guidelines for infection Prevention and Control Programmes at the National and control programmes (ECDC, 2018). Acute Health Care Facility Level , Geneva. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 166

169 II.6. EFFECTIVENESS: QUALITY OF CARE AND PATIENT EXPERIENCE 6.28. Observed and predicted percentage of hospitalised patients with at least one healthcare-associated infection, 2016-17 Predicted Observed % 12 10 8 6 4 2 0 1. Country representativeness of data is limited in Bulgaria and the Netherlands. 2. Data from Norway includes partial imputation for missing types of infections. 95% confidence intervals represented by H. Data for Denmark and Sweden are not available. The EU average includes Iceland and Note: Norway. ECDC 2016-17 Point prevalence survey. Source: 1 2 http://dx.doi.org/10.1787/888933836143 6.29. Healthcare-associated 6.30. Healthcare-associated infections by type of infection infections by type of care (specialty) across EU countries, 2016-17 across EU countries, 2016-17 Geria trics Ot her c Systemi Other 4.3% 5. 4% 7.7% 7.8% Pneumonia 4.2% Skin 25.7% Surgical 32.7% Intensive care 13.4% Gastrointestinal 8.9% Bl oodstream 10.8% Surgical site 18.3% Medical 40.2% ics 1.7% Pediat r Urinary tract 18.9% Source: Source: ECDC 2016-17 Point prevalence survey. ECDC 2016-17 Point prevalence survey. http://dx.doi.org/10.1787/888933836181 1 2 1 2 http://dx.doi.org/10.1787/888933836162 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 167

170

171 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART II Chapter 7 Accessibility: Affordability, availability and use of services Accessibility is one of the key objectives of any health system. If access to care is limited for some population groups or the population at large, the result may be poorer health outcomes and greater health inequalities. Access to care can be limited for different reasons: it may not be affordable, the distance to the closest health care facility may be too great or waiting times for treatment too long. Unmet needs for health care is an important indicator of accessibility problems as reported by the population itself. Recent survey data show that the share of the population reporting unmet care needs is generally low in most EU countries, but low-income households are much more likely to report unmet care needs than high- income households, mainly for financial reasons. The affordability of health services can be restricted when they involve high out-of-pocket payments. On average across EU countries, 18% of health spending is paid out-of-pocket by households, but with wide variations. In general, countries that have a high share of out-of-pocket spending also have a high proportion of the population facing catastrophic payments for health services, particularly among low-income groups. To promote access to care for the whole population, most EU countries have achieved universal coverage for a core set of health services, although the range of services covered and the degree of cost-sharing vary. In addition to being affordable, health services also need to be accessible when and where people need them. Although the number of doctors and nurses per population has increased over the past decade in nearly all EU countries, shortages of general practitioners persist in many countries, particularly in rural and remote areas. The use of health services varies widely across EU countries. While these variations may reflect differences in health care needs, they also suggest either some under-use of some services for population groups facing accessibility issues or a possible over-use of some services. Long waiting time for elective surgery is an important policy issue in many EU countries, as it restricts timely access to care and generates patient dissatisfaction. In countries where this is an issue, waiting times have frequently gone up in recent years as the demand for surgery increased more rapidly than the supply, following some success in bringing waiting times down before 2010. 169

172 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES UNMET HEALTH CARE NEEDS affordability and accessibility to care, such as the Accessibility to health care can be limited for a extent of health care coverage, the amount of out-of- number of reasons, including cost, distance to the pocket payments, and the actual use of health services. closest health facility and waiting times. Unmet care Strategies to improve access to care for poor people needs may result in poorer health for people forgoing and disadvantaged groups need to tackle not only care and may increase health inequalities if such unmet affordability issues, but also effective access to services needs are concentrated among poor people. As noted by by promoting an adequate supply and distribution of the Expert Panel on Effective Ways of Investing in health workers and services throughout the country. Health, there are many challenges in measuring unmet needs for particular interventions, but the data from the EU Statistics on Income and Living Conditions survey (EU-SILC) are the only timely and comparable source of Definition and comparability information available across all Member States (Expert Questions on unmet health care needs are Panel on Effective Ways of Investing in Health, 2018). included in the European Union Statistics on In all European countries, most of the population Income and Living Conditions survey (EU-SILC). in 2016 reported that they had no unmet care needs for People are asked whether there was a time in the financial reasons, geographic reasons or waiting times, previous 12 months when they felt they needed based on EU-SILC (Figure 7.1). However, in Estonia and medical care or dental care but did not receive it, Greece, at least 10% of the population reported some followed by a question as to why the need for care unmet needs for health care, with the burden falling was unmet. The data presented here focus on mostly on low-income groups, particularly in Greece. three reasons: the care was too expensive, the Nearly one in five Greek people in the lowest income distance to travel too far or waiting times too quintile reported going without some medical care long. when they needed it mainly for financial reasons. In Cultural factors may affect responses to Estonia, the main reason for people to report unmet questions about unmet care needs. There are also care needs is because of long waiting times.This can be some variations in the survey question across partly explained by the limited volume of some countries: while most countries refer to both a services (such as specialist consultations) fully medical examination and treatment, the question reimbursed by public health insurance. in some countries (e.g. Czech Republic, Slovenia In most countries, a larger proportion of the and Spain) only refer to a medical examination or population indicates some unmet needs for dental care a doctor consultation, resulting in lower rates of igure 7.2). This is mainly because than for medical care (F unmet needs. The question in Germany refers to dental care in many countries is only partially included (or unmet needs for “severe” illnesses, also resulting not included at all) in public schemes and so must either in some under-estimation compared with other be paid out-of-pocket or covered through purchasing countries. Some changes in the survey question in private health insurance (see the indicator Extent of some countries in 2015 and 2016 have also led to health care coverage). More than one in eight people (13%) substantial reductions. Caution is therefore in Portugal, Greece and Latvia reported unmet needs for required in comparing variations across countries dental care in 2016, mainly for financial reasons. On the and over time. other hand, a very small proportion of people reported Income quintile groups are computed on the unmet dental care needs in the Netherlands, Austria, basis of the total equivalised disposable income Germany, Slovenia and the Czech Republic in 2016, with attributed to each member of the household. The very little difference across income groups. first quintile group represents the 20% of the Unmet needs for medical care and dental care population with the lowest income, and the fifth have decreased since 2015 on average across EU quintile group the 20% of the population with the countries, although part of the reduction in some highest income. countriesissimplyduetoachangeinthesurvey question (Figure 7.3 and Figure 7.4). However, the gap in unmet medical and dental care needs between poor people and rich people has remained the same: people in the lowest income quintile are still five times more Reference likely to report unmet medical care needs than those in Expert Panel on Effective Ways of Investing in Health (2018), the highest quintile, and they are six times more likely Opinion on Benchmarking Access to Healthcare in the to report unmet dental care needs. EU, European Union, https://ec.europa.eu/health/expert_ Indicators of self-reported unmet care needs panel/sites/expertpanel/files/docsdir/opinion_benchmarking_ should be assessed together with other indicators of healthcareaccess_en.pdf . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 170

173 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.2. Unmet need for dental examination 7.1. Unmet need for medical examination for financial, geographic or waiting for financial, geographic or waiting times reasons, by income quintile, times reasons, by income quintile, 2016 (or nearest year) 2016 (or nearest year) High income Total population Low income High income Total population Low income Estonia Portu g al Greece Greece Latvia Latvia Poland Estonia Ital Romania y Ital y Romania Finland Finland S Lithuania p ain Bul Lithuania aria g EU28 Ireland Bel EU28 ium g Bel Denmark g ium Poland Portu g al C Slovak Re yp p ublic rus Croatia Sweden Sweden Bul aria g France Ireland Hun France y g ar Slovak Re Denmark p ublic Malta Hun g ar y United Kin g dom dom United Kin g Croatia Czech Re ublic p Malta C rus yp Luxembour S p ain g Luxembour p g ublic Czech Re Slovenia Slovenia German German y y Netherlands Austria Netherlands Austria Iceland Montene g ro Serbia Turke y Montene Serbia g ro Turke Iceland y FYR of Macedonia Norwa y Norwa y Switzerland FYR of Macedonia Switzerland 0 0 03 0 1 01 02 03 02 % % Source: Eurostat Database, based on EU-SILC. Source: Eurostat Database, based on EU-SILC. http://dx.doi.org/10.1787/888933836219 http://dx.doi.org/10.1787/888933836200 1 2 1 2 7.3. Change in unmet medical care need 7.4. Change in unmet dental care need for financial, geographic or waiting for financial, geographic or waiting times reasons, by income quintile, times reasons, by income quintile, all EU countries, 2008 to 2016 all EU countries, 2008 to 2016 High income High income Total population Low income Total population Low income % % 10 10 8 8 6 6 4 4 2 2 0 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2008 2009 2010 2011 2012 2013 2014 2015 2016 Eurostat Database, based on EU-SILC. Source: Source: Eurostat Database, based on EU-SILC. 1 2 http://dx.doi.org/10.1787/888933836257 1 2 http://dx.doi.org/10.1787/888933836238 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 171

174 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES FINANCIAL BURDEN OF OUT-OF-POCKET EXPENDITURE People experience financial hardship when direct and cap the co-payments a household has to pay over out-of-pocket payments – formal and informal – are a given time period (e.g. Austria, Czech Republic, large in relation to their ability to pay for health care. Ireland and United Kingdom). Even small out-of-pocket payments can cause financial There is clear evidence of the impact of changes in hardship for poor households and those who have to user charges policy in some countries. For example, pay for long-term treatment such as medicines for looking at catastrophic spending over time in Latvia, chronic illness. Where health systems fail to provide theintroductionofexemptionsfromallco-payments adequate financial protection, people may not have for very poor people in 2009, the extension of enough money to pay for health care or to meet other exemptions to other poor people in 2010 and the basic needs. As a result, lack of financial protection abolition of exemptions in 2012 for all except the very may reduce access to health care, undermine health poorest households coincided with an improvement status, deepen poverty and exacerbate health and and then deterioration in financial protection among socioeconomic inequalities. Because all health the poor (Thomson et al., 2018). systems involve a degree of out-of-pocket payments, financial hardship can be a problem in any country. At an aggregate level, the share of out-of-pocket Definition and comparability spending in total health spending reflects the degree of financial protection in a country along the three Out-of-pocket payments are expenditures dimensions of coverage – the share of the population borne directly by a patient where neither public covered, the range of services included in a public nor private insurance cover the full cost of the benefit basket and the proportion of costs covered by health good or service. They include cost-sharing collective third-party payer schemes for each service. and other expenditure paid directly by private Thus, the share of out-of-pocket payments are higher households and should also in principle include in those countries where significant groups of the estimations of informal payments to health care population are excluded from coverage, important providers. health services not included in the public benefit Catastrophic health spending is an indicator of package or the cost-sharing of public payers limited for financial protection used to monitor progress some services. Across the EU, around a fifth of all towards universal health coverage (UHC) at global, health spending is borne directly by private regional and national levels. It is defined as out-of- households (Figure 7.5).This figure ranges from around pocket payments that exceed a predefined 10% in France, Luxembourg or the Netherlands to over percentage or threshold of a household’s ability to 40% in Bulgaria, Latvia and Cyprus. pay for health care. Ability to pay may be defined in The indicator most widely used to measure the different ways, leading to measurement differences financial hardship associated with out-of-pocket (Cylus et al., 2018). In the data presented here, payments for households is the incidence of ability to pay is defined as household consumption catastrophic spending on health (Cylus et al., 2018). spending minus a standard amount representing The incidence of catastrophic health spending varies basic spending on food, rent and utilities (water, considerably across EU countries, ranging from fewer electricity, gas and other fuels); the threshold used than 2% of households in France, Ireland, Slovenia, to define households with catastrophic spending is Sweden and the United Kingdom, to over 8% of 40%. Microdata from national household budget households in Greece, Hungary, Latvia, Lithuania, surveys are used to calculate this indicator. Poland and Portugal (Figure 7.6). Across Europe, poor households (i.e. those in the bottom quintile in terms of consumption) are most likely to experience catastrophic health spending, despite the fact that many countries have put in place policies to safeguard References financial protection. Cylus, J., S. Thomson and T. Evetovits (2018), Catastrophic Countries with comparatively high levels of public health spending in Europe: Equity and policy implications spending on health and low levels of out-of-pocket of different calculation methods, Bulletin of WHO. payments typically have a lower incidence of catastrophic out-of-pocket payments. However, policy Financial Thomson, S., T. Evetovits and J. Cylus (2018), choices are also important, particularly choices around protection in high-income countries. A comparison of the coverage policy (WHO Regional Office for Europe, 2018). Czech Republic, Estonia and Latvia , WHO Regional Office Financial protection is demonstrably stronger in for Europe. countries that cover the whole population, although WHO Regional Office for Europe (2018), “Can people afford this in itself is not enough to guarantee protection. to pay for health care? New evidence on financial Countries with a low incidence of catastrophic protection in Europe”, WHO Regional Office for Europe, spending on health are also more likely to exempt poor www.euro.who.int/en/health-topics/Health-systems/health- people and regular users of care from co-payments; use systems-financing/universal-health-coverage-financial- low fixed co-payments instead of percentage . protection co-payments, particularly for outpatient medicines; HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 172

175 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.5. Share of total health spending financed by out-of-pocket payments, 2016 (or latest year) % 60 57 48 50 45 45 41 40 36 35 34 32 32 30 30 28 30 24 23 23 23 21 20 19 18 18 20 17 16 16 15 15 15 15 15 14 13 12 12 11 11 10 10 0 . https://doi.org/10.1787/health-data-en Source: OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933836276 7.6. Share of households with catastrophic spending on health by consumption quintile, latest year available 4th Quintile 3rd Quintile 2nd Quintile Poorest Quintile Richest Quintile Households (%) 14 12 10 8 6 4 2 0 Source: WHO Regional Office for Europe 2018. 1 2 http://dx.doi.org/10.1787/888933836295 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 173

176 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES POPULATION COVERAGE FOR HEALTH CARE any cost-sharing left after basic coverage The share of the population covered by a public or (complementary insurance), add additional services private scheme provides some indication of the (supplementary insurance) or provide faster access or financial protection against the costs associated with larger choice of providers (duplicate insurance). In health care, but this is not a complete indicator of most European countries, only a small proportion of affordability as the range of services covered and the the population has an additional private health degree of cost-sharing applied to those services also insurance. But in five countries (France, Netherlands, matter. These three dimensions – the “breadth”, Slovenia, Belgium and Croatia), half or more of the “depth” and “height” of coverage – define how population has private coverage (Figure 7.8). comprehensive health care coverage is in a country. The indicator presented here on population coverage In France, nearly all the population (96%) has a looks at the first dimension only, whereas the next complementary private health insurance to cover cost indicator on the extent of health care coverage takes a sharing in the social security system. The Netherlands broader look at these three dimensions together. has the largest supplementary market (87% of the population), whereby private insurance pays for dental Most European countries have achieved universal care that is not publicly reimbursed. Duplicate private (or near-universal) coverage for a core set of services, health insurance, providing faster private-sector which usually include consultations with doctors, tests access to medical services where there are waiting and examinations and hospital care (Figure 7.7). Yet, in times in public systems, is largest in Ireland (45%). some countries coverage of these core services may not be universal. In Ireland, for example, only around 50% The population covered by private health of the population is covered for the costs of GP visits, insurance has increased in some countries over the past although recent reform proposals suggest a gradual roll decade, particularly in Denmark, Slovenia and Belgium out of primary care coverage to the entire population (Figure 7.9). The development of private health (OECD/European Observatory on Health Systems and insurance is linked to several factors, including gaps in Policies, 2017). In Greece, a new law in 2016 (Law 4368/ access to publicly financed services, government 2016) provided universal health coverage for the whole interventions directed at private health insurance population, closing the coverage gap for the 10% of the markets and historical development. population that were previously uninsured. These previously uninsured people now have legally- recognised access to a broad range of services and Definition and comparability goods (including hospital care and prescribed pharmaceuticals), like any other Greek citizen. Population coverage for health care is defined as the share of the population covered for a Three European countries (Cyprus, Bulgaria and defined set of health care goods and services Romania) still have at least 10% of their population not under public programmes and through private covered for health services. In Bulgaria, the share of the health insurance. Public coverage refers both to population covered has decreased since 2010 when a government programmes, generally financed by tightening of the law resulted in people losing their taxation, and social health insurance, generally social health insurance coverage if they failed to pay financed by payroll taxes. The take-up of private their contribution (Dimova et al., 2012). However, it is health insurance is often voluntary, although it common for uninsured people who need medical care may be mandatory by law or compulsory for to go to emergency services in hospital, where they will employees as part of their working conditions. beencouragedtogetinsurance(withoutpayingany Premiums are generally non-income-related financial penalty for not having had an insurance prior although the purchase of private coverage can to that). In Romania, although social health insurance be subsidised by the government. is compulsory, only 89% of the population was covered in 2017. The uninsured population include mainly people working in agriculture, self-employed or unemployed people who are not registered for unemployment or social security benefits, and Roma References people who do not have identity cards (which Dimova, A. et al. (2012), “Bulgaria: Health System Review”, precludes them from enrolling into the social security Health Systems in Transition, Vol. 14, No. 3. system). The uninsured can only access a minimum benefits package, covering emergency care, treatment OECD/European Observatory on Health Systems and of communicable diseases and care during pregnancy Policies (2017), “Ireland, Country Health Profile 2017”, https://ec.europa.eu/health/ State of Health in the EU, (Vlãdescu et al., 2016). state/country_profiles_en . Basic primary health coverage generally covers a defined set of benefits, but in many cases with cost Vlãdescu, C. et al. (2016), “Romania: Health System sharing. In some countries, additional health coverage Review”, Health Systems in Transition, Vol. 18, No. 4. can be purchased through private insurance to cover HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 174

177 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.8. Private health insurance coverage, 7.7. Population coverage for a core set of services, 2016 (or nearest year) 2016 (or nearest year) Croatia (2012) 100 Complementary Czech Republic 100 Supplementary Duplicate Denmark 100 Finland 100 Germany 100 France 95.5 Greece 100 Netherlands 87.3 Ireland 100 Italy 100 Slovenia 71.6 12.7 Latvia 100 Belgium 82.7 Malta 100 Portugal Croatia 100 58.0 Slovenia 100 Ireland 45.4 Sweden 100 Austria 36.5 United Kingdom 100 Austria 99.9 Denmark 32.5 France 99.9 Portugal 26.0 Netherlands 99.9 Germany 23.2 Spain 99.9 Belgium 99.0 Cyprus 21.5 Hungary 95.0 Malta 21.2 Slovak Republic 94.5 Finland 21.0 Estonia 94.0 Lithuania 92.5 Spain 15.7 Poland 91.5 Latvia 14.6 Romania 89.0 Greece 12.0 Bulgaria (2013) 88.2 Cyprus (2013) 83.0 United Kingdom 10.5 Bulgaria 2.4 FYR of Macedonia 100 Iceland 100 Lithuania 1.5 Montenegro 100 Sweden 0.1 Norway 100 Serbia 100 Switzerland 100 Switzerland 27.9 Turkey 98.2 Turkey 7.6 100 70 80 90 0 20406080100 % of population % of population Note: This includes public coverage and primary private health Note: This excludes primary PHI. PHI can be both complementary and supplementary in Denmark and Germany. coverage. Data for Luxembourg is not available. Source: OECD Health Statistics 2018, https://doi.org/10.1787/health- https://doi.org/10.1787/health-data- OECD Health Statistics 2018, Source: ; and Voluntary health insurance in Europe: country experience, ; European Observatory Health Systems in Transition (HiT) data-en en Observatory Studies Series, 2016, for non-OECD countries. Series for non-OECD countries. 1 2 http://dx.doi.org/10.1787/888933836333 http://dx.doi.org/10.1787/888933836314 1 2 7.9. Trends in private health insurance coverage, 2005 to 2016 (or nearest year) 2016 2005 % of population 96 100 84 83 80 87 60 37 33 40 45 26 23 21 15 28 20 8 15 12 11 0 Note: These data exclude primary private health insurance. Source: . https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933836352 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 175

178 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES EXTENT OF HEALTH CARE COVERAGE In addition to the share of the population entitled in providing financial protection. This is the case in the to basic health services, the range of services included in Netherlands where more than 70% of total costs are a publicly defined benefit package and the proportion of borne by these schemes. costs covered are the other dimensions defining the Public coverage for pharmaceuticals is also typically Figure 7.10 extent of health care coverage in a country. lesscomprehensivethancoverageforinpatientand assesses these three dimensions of coverage for a outpatient care. Across the EU, around 64% of all selected number of key health care functions. For each pharmaceutical costs are covered by government and function, it displays the share of the costs that is funded compulsory schemes. Over-the-counter medications – collectively – either by government schemes or which are available without prescription and are compulsory insurance arrangements. Differences typically not covered by basic coverage schemes – play an across countries in the share of the costs covered can be important role in some countries (see indicator due to the fact that some specific goods and services are “Pharmaceutical Expenditure” in Chapter 5). Less than included in the public benefit package in one country 20% of pharmaceutical costs are covered by government but not in another (e.g. a particular drug or medical schemes in Cyprus and Bulgaria. In Germany, this procedure), that cost-sharing arrangements for the proportion reaches 84% as cost-sharing is only moderate same goods and services vary or that some services are with patients generally having to pay a co-insurance rate only covered for specific population groups in a country. of 10% for each prescribed pharmaceutical capped at In the EU, inpatient services in hospitals are more EUR 10 per item within an annual Co-payment Cap. comprehensively covered than any other type of care. Therapeutic appliances such as glasses and other Across the EU, 93% of all inpatient costs are borne by eye products, hearing aids and other medical devices are government or compulsory insurance schemes. In many typically covered to a lesser extent than other health countries, patients have access to free acute inpatient services. Government and compulsory insurance care.This is the case, for example, in Denmark, Hungary, schemes cover more than 50% of these expenses in only Poland, Spain and the United Kingdom, so government four EU countries. In case of corrective eye products, and compulsory schemes cover more than 90% of these compulsory coverage is often limited to paying partially costs in those countries. In the Netherlands, these for the cost of glasses, while private households are left services are also free of charge once the annual general to bear the full cost of the frames if they are not covered deductible has been met. Only in Cyprus, Greece and by complementary insurance. Ireland is the financial coverage for costs of inpatient care below 70%. In those countries, patients frequently choose treatment in facilities not included in the public Definition and comparability benefit package. Health care coverage is defined by the share of More than three-quarters of spending on the population entitled to services (“breadth of outpatient care in the EU are borne by government and coverage”), the range of services included in a compulsory health financing schemes (77%). With the benefit package (“depth of coverage”) and the exception of Bulgaria and Cyprus, at least 50% of all proportion of costs covered (“height of coverage”) outpatient medical care costs are financed by by government schemes and compulsory compulsory third-party payers in EU countries. In a insurance schemes. Financial coverage provided number of countries, outpatient primary and specialist by voluntary health insurance is not considered. care are generally free at the point of service but user The core functions analysed here are defined charges may still apply for specific services or if based on the definitions in the System of Health providers outside the public sector are consulted. This Accounts (SHA, 2011). Inpatient care refers to is, for example, the case in Denmark, where 92% of inpatient curative and rehabilitative care in total costs are covered but user charges exist for visits hospitals, outpatient medical care to all outpatient to psychologists and physiotherapists, and the curative and rehabilitative care excluding dental United Kingdom (84%), where care provision outside of care, pharmaceuticals to prescribed and over-the- NHS commissioned services are not covered. counter medicines including medical non- Public coverage for dental care costs is more limited durables, and therapeutic appliances mainly to in the EU with more restricted service packages eyewear, hearing aids and other medical devices. (frequently limited to children) and higher levels of cost- Comparing the shares of financial coverage for sharing. Around 30% of total dental care costs are borne different types of services is a simplification as by government schemes or compulsory insurance in reality is more complex and does not reflect Europe. More than half of this spending is only covered possible trade-offs. For example, a country with in a few EU countries: Slovenia (51%), the Slovak Republic more restricted population coverage but a very (56%), Croatia (61%) and Germany (68%). By contrast, in high depth of coverage may display a lower countries like Italy and the United Kingdom, dental care share of financial coverage than a country where costs for adults without any specific entitlement are not the entire population is entitled to services but covered. In countries where dental care is not with a more limited benefit basket. comprehensively included in the public benefit package, voluntary health insurance may play an important role HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 176

179 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.10. Health care coverage for selected goods and services, 2016 Government and compulsory insurance spending as proportion of total health spending by type of service Therapeutic Outpatient Hospital care Pharmaceuticals Dental care appliances medical care 36% EU-28 93% 77% 64% 30% 37% Aust ria 87% 78% 46% 68% 43% Be lgium 77% 79% 70% 42% N/A Bulgaria 86% 46% 19% 46% 40% Croatia 90% 83% 61% 61% 12% Cyprus 67% 37% 6% 18% 38% Czech Republic 95% 90% 47% 59% Denmark 45% 91% 92% 20% 44% Estonia 30% 85% 98% 26% 54% Finland N/A 90% 79% 30% 55% N/A France 66% 95% 77% 76% Germany 54 % 96% 90% 68 % 84% Greece N/A 65% 67% 0% 53% Hungary 55% 58% 91% 21% 51% Ireland 20% 69% 73% N/A 77% N/A Italy 59% 96% 20% 63% Latvia 51% 82% 10% 7% 35% Lithuania 75% 18% 93% 22 % 33% Luxembourg 83% 45 85% 47% % 80% Malta 58% 95% 70% 15% 40% Netherlands 84% 91% 8% 43% 68% Poland 61% 95% 29% % 43 34% Portugal 63% 85% N/A 29% 54% Romania 79% 6% 98% 42% 56% Slovak Republic 95% 89% % 56 41% 71% Slovenia 78% 87% 51% 31% 50% Spain 79% 1% 94% 5% 57% Sweden 86% 39% 99% 40% 52% United Kingdom 84% N/A 94% 16% 70% 78% Iceland 23% 99% 83% 36% Norway 85% 28 % 99% 41% 56% Switzerland 61% 6% 81% 55% 41 % Outpatient medical services mainly refer to services provided by generalists and specialists in the outpatient sector. Note: Pharmaceuticals include prescribed and over-the-counter medicines as well as medical non-durables. Therapeutic appliances refer to vision products, hearing aids, wheelchairs and other medical devices. N/A means data not available. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: . 1 2 http://dx.doi.org/10.1787/888933836371 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 177

180 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES AVAILABILITY OF DOCTORS Access to medical care requires an adequate providing financial incentives for doctors to specialise number of doctors, with a proper mix between (OECD, 2016). generalists and specialists and a proper distribution in The uneven geographic distribution of doctors and all parts of the country. the difficulties in recruiting and retaining doctors in The number of doctors per capita varies widely certain regions is another important policy issue in across EU countries (Figure 7.11). In 2016, Greece had many European countries, especially those with remote the highest number with 6.6 doctors per 1 000 and sparsely populated areas. The density of physicians population, but this number is an over-estimation as it is consistently greater in urban regions, reflecting the includes all doctors who are licensed to practice concentration of specialised services such as surgery (including retired physicians and those who might and physicians’ preferences to practice in urban have emigrated to other countries). Austria and settings. Differences in the density of doctors between Portugal also had a high number of doctors per urban regions and rural regions are highest in the population, but the number in Portugal is also over- Slovak Republic, the Czech Republic and Greece (OECD, estimated for the same reason as in Greece (without 2017). this over-estimation, the number of practising doctors Many countries provide different types of in Portugal would likely be slightly below the EU financial incentives to attract and retain doctors in average). The number of doctors per capita was lowest underserved areas, including one-time subsidies to in Poland, the United Kingdom and Romania. help them set up their practice as well as recurrent Since 2000, the number of doctors per capita has payments such as income guarantees and bonus increased in all EU countries, except in France, Poland payments. A number of countries have also introduced and the Slovak Republic where it has remained stable. measures to encourage students from underserved On average across EU countries, the number increased regions to enrol in medical schools (Ono et al., 2014). from 2.9 doctors per 1 000 population in 2000 to 3.6 in 2016. In most countries, the global economic crisis that started in 2008 did not have much impact on the Definition and comparability growth in the number of doctors. Practising physicians are defined as doctors who Projecting the future supply and demand of are providing care for patients. In some countries, doctors is challenging given the high levels of the numbers also include doctors working in uncertainty concerning retirement and migration administration, management, academic and patterns and possible changes in the demand for their research positions (“professionally active” services. Many EU countries have anticipated the physicians), adding another 5-10% of doctors. current and future retirement of a significant number Greece and Portugal report all physicians entitled of doctors by increasing their education and training to practice, resulting in an even greater efforts so that there would be enough new doctors to overestimation. In Belgium, a minimum threshold replace those who will retire (OECD, 2016). of activities (500 consultations per year) is set for general practitioners to be considered to be In many countries, the main concern is about practising, resulting in an under-estimation current and future shortages of general practitioners, compared with other countries which do not set particularly in rural and remote regions. Whereas the such a threshold. overall number of doctors per capita has increased in nearly all countries, the share of general practitioners (GPs) has come down in most countries. On average across EU countries, GPs made up less than 25% of all References eece and Poland have (Figure 7.12). Gr physicians in 2016 the lowest share of GPs, while Portugal, France, Finland OECD (2016), Health Workforce Policies in OECD Countries: and Belgium have been able to maintain a better , OECD Publishing, Right Jobs, Right Skills, Right Places balance between GPs and specialists. In response to . Paris, http://dx.doi.org/10.1787/9789264239517-en these concerns about shortages of generalists, several OECD (2017), Health at a Glance 2017: OECD Indicators , OECD countries have taken steps to increase the number of http://dx.doi.org/10.1787/health_glance- Publishing, Paris, post-graduate training places in general medicine. In 2017-en . France, the number of post-graduate training places Ono, T., M. Schoenstein and J. Buchan (2014), “Geographic filled in general medicine more than doubled between Imbalances in Doctor Supply and Policy Responses”, 2005 and 2015, rising from 1 500 to over 3 500. However, No. 69, OECD Publishing, OECD Health Working Papers, in most countries, specialists earn much more than GPs, . http://dx.doi.org/10.1787/5jz5sq5ls1wl-en Paris, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 178

181 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.11. Practising doctors per 1 000 population, 2000 and 2016 (or nearest year) 2016 2000 Per 1 000 population 7 6.6 6 5.1 4.8 4.5 4.5 5 4.3 4.3 4.2 4.1 4.0 3.9 3.8 3.8 3.8 3.7 3.7 3.6 3.5 3.5 3.5 4 3.2 3.2 3.2 3.2 3.1 3.1 3.0 3.0 3.0 2.9 2.9 2.8 2.8 2.6 2.4 3 1.8 2 1 0 1. Data refer to all doctors licensed to practice, resulting in a large over-estimation of the number of practising doctors (e.g. of around 30% in Portugal). 2. Data include not only doctors providing direct care to patients, but also those working in the health sector as managers, educators, researchers, etc. (adding another 5-10% of doctors). OECD Health Statistics 2018, ; Eurostat Database. Source: https://doi.org/10.1787/health-data-en 1 2 http://dx.doi.org/10.1787/888933836390 7.12. Share of different categories of doctors, 2016 (or nearest year) General practitioners Other doctors (not further defined) Other generalists¹ Specialists % 0 10 80 60 40 46 20 38 37 37 33 32 30 27 27 24 24 24 24 23 22 22 21 21 20 20 19 19 18 18 17 17 17 16 15 15 15 9 5 0 1. Other generalists include non-specialist doctors working in hospital and recent medical graduates who have not started yet their post-graduate specialty training. 2. In Portugal, only about 30% of doctors employed by the public sector (NHS) are working as GPs in primary care, with the other 70% working in hospital. ; Eurostat Database. https://doi.org/10.1787/health-data-en Source: OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933836409 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 179

182 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES AVAILABILITY OF NURSES (e.g. Austria, Belgium, Denmark, Germany and Malta). In Nurses greatly outnumber physicians in most EU France, the number of nurses working in hospital per countries, with a ratio of two to four nurses per doctor population also increased slightly, but the number of full- in many countries. Nurses play a critical role in time equivalents has remained relatively stable, meaning providing health care not only in hospitals and long- that the average number of working hours has decreased term care institutions, but increasingly also in primary igure 7.14). In many countries, the ratio of full- slightly (F care and in home care settings. time equivalent nurses to the absolute number ranges There are concerns in many countries about from 0.80 to 0.95, and it has been fairly stable over time. possible future shortages of nurses, given that the However, this ratio is much lower in Belgium and demand for nurses is expected to rise in a context of Germany (0.70 to 0.75), indicating that nurses generally population ageing and the retirement of the current work fewer hours in these countries. “baby-boom” generation of nurses. These concerns A growing number of nurses also work in primary have prompted actions in many countries to increase care in many countries. In response to shortages of the training of new nurses, while some countries have general practitioners, some countries have introduced addressed current shortages by recruiting nurses from or extended advanced roles for nurses to improve access other countries (OECD, 2016). to primary care. Evaluations of the experience with On average across EU countries, there were (advanced) nurse practitioners in countries like Finland 8.4 nurses per 1 000 population in 2016, a rise from 6.7 in and the United Kingdom indicate that these nurses can 2000 (Figure 7.13). The number of nurses per capita is improve access to care and reduce waiting times, while highest in Denmark and Finland, although about one- providing the same quality of care as doctors for a range third of nurses in these two countries are trained at a of patients (e.g. those with minor illnesses or requiring lower level than general nurses and perform lower tasks. routine follow-up) (Maier et al., 2017). This is also the case in Switzerland and Iceland. In other countries such as Italy and Spain, a large number of health care assistants (or nursing aids) provide assistance to nurses. Greece has the lowest number of nurses per Definition and comparability capita among EU countries, but the data only include The number of nurses includes those providing nurses working in hospital. Bulgaria, Latvia, Poland and services for patients (“practising”), but in some Cyprus also have a relatively low number of nurses. countries also those working as managers, Since 2000, the number of nurses per capita has educators or researchers (“professionally active”). increased in most European countries, except in the In countries where there are different levels of Baltic countries (Estonia, Latvia and Lithuania) where nurses, the data include both “professional” the number of nurses per capita has remained stable nurses (including general and specialist nurses) (meaning that there has been a reduction in the and “associate professional” nurses who have a absolute number of nurses given that the overall lower level of qualifications but are nonetheless population has come down in these countries) and the recognised and registered as nurses in their Slovak Republic where the number of nurses has come country. Health care assistants (or nursing aids) down both in absolute number and on a per capita who are not recognised as nurses are excluded. basis. Most of this reduction in the Slovak Republic has Austria and Greece report only nurses working occurred between 2000 and 2010, with the number in hospitals (resulting in an underestimation). stabilising at a lower level since then. Full-time equivalent employment is defined The increase in the number of nurses per capita as the number of hours worked divided by the has been particularly large in Denmark, Finland, average number of hours worked in full-time Germany, Luxembourg, France and Malta. Malta has jobs, which may vary across countries. takenaseriesofmeasurestotrainmorenurses domestically and also to attract more nurses from other countries to address current shortages. The university degree to become a nurse in Malta is free of charge for students; and once students have graduated, they are References also encouraged to take time off to pursue their training Maier, C. et al. (2017), “Nurses in Advanced Roles in while continuing to receive at least part of their salary. OECD Primary Care: Policy Levers for Implementation”, , No. 98, OECD Publishing, Paris, Health Working Papers Most nurses in EU countries continue to work in . http://dx.doi.org/10.1787/a8756593-en hospital. Relative to the overall size of the population, the number of nurses working in hospital, when measured OECD (2016), Health Workforce Policies in OECD Countries: both in absolute numbers and full-time equivalents, has , OECD Publishing, Right Jobs, Right Skills, Right Places increased over the past decade in many countries http://dx.doi.org/10.1787/9789264239517-en . Paris, HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 180

183 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.13. Practising nurses per 1 000 population, 2000 and 2016 (or nearest year) 2016 2000 Per 1 000 population 20 17.5 17.0 16.9 18 14.3 14.2 16 12.9 14 11.7 11.1 11.1 10.6 10.2 12 9.7 8.4 11.6 8.1 8.1 8.0 10 7.7 6.7 6.5 6.4 6.3 6.1 6.0 8 5.6 5.5 5.3 5.2 5.1 4.6 7.9 4.4 4.3 6 3.3 5.7 4 1.9 2 0 1. In Denmark, Finland, Iceland and Switzerland, about one-third of nurses are "associate professional" nurses with a lower level of qualifications. In Denmark and Switzerland, most of the growth in the number of nurses since 2000 has been in this category of associate professional nurses. 2. Data include not only nurses providing care for patients, but also those working as managers, educators, etc. 3. Austria and Greece report only nurses employed in hospital. Source: ; Eurostat Database. https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, http://dx.doi.org/10.1787/888933836428 1 2 7.14. Nurses working in hospital, head count vs full time equivalent, 2006 and 2016 (or nearest year) Head Count Full Time Equivalent Per 1 000 population 10 9 8 7 6 5 4 3 2 1 0 2016 2016 2006 2016 2006 2016 2006 2006 2016 2006 2016 2006 2016 2006 2016 2006 2016 2006 2016 2006 2016 2006 Belgium Malta Iceland Norway Estonia Ireland Denmark Austria France Germany Switzerland Data include professional and associate professional nurses as well as midwives working in hospital. Note: https://doi.org/10.1787/health-data-en ; Eurostat Database. OECD Health Statistics 2018, Source: 1 2 http://dx.doi.org/10.1787/888933836447 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 181

184 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES CONSULTATIONS WITH DOCTORS Consultations with doctors are, for most people, Sweden, Denmark, Austria and France, as the number the most frequent contacts with health services. These of doctors has increased more rapidly than the number consultations can take place either in doctors’ offices of consultations, whereas it has remained relatively or clinics, in hospital outpatient departments or, in stable in Germany and has increased in Poland but some cases, in patients’ own homes. mainly between 2000 and 2008 (Figure 7.17). In 2016, the number of doctor consultations per person per year was highest in the Slovak Republic, the Czech Republic, Hungary and Germany, with Definition and comparability 10 consultations or more per year. It is lowest in Consultations with doctors refer to the number Sweden, Finland and Denmark with less than of contacts with physicians, including both 5 consultations per person per year (Figure 7.15).The EU generalists and specialists. There are variations average is 7.5 consultations per person per year, with across countries in the coverage of different types most countries reporting 5 to 8 visits. Some differences of consultations, notably in outpatient in health service delivery and payment methods can departments of hospitals. The data come mainly explain some of the variations across countries. In from administrative sources, although in some Sweden and Finland, the low number of doctor countries (Ireland, Italy, the Netherlands, Spain consultations can be explained partly by the fact that and Switzerland) the data come from health nurses and other health professionals play an important interview surveys. Data from administrative role in primary care centres, lessening the need for sources tend to be higher than those from consultations with doctors (Maier et al., 2017). Some surveys because of problems with recall and non- countries which pay their doctors mainly by fee-for- response rates and also because the surveys used service (e.g. the Slovak Republic and the Czech Republic) only cover the adult population, leading to an tend to have higher consultation rates than other under-estimation. countries where doctors are mainly paid by salaries or In Hungary, the data include consultations for capitation. diagnostic exams, such as CT and MRI scans The estimated number of consultations per doctor (resulting in an over-estimation). The data for the is highest in Hungary, the Slovak Republic, Poland and Netherlands exclude contacts for maternal and the Czech Republic, with more than 3 000 consultations child care. In Germany, the data include only the per doctor per year. It is lowest in Sweden, Denmark, number of cases of physicians’ treatment according Austria and Finland, with less than 1 500 consultations to reimbursement regulations under the Social per doctor per year (Figure 7.16). This indicator should Health Insurance Scheme (a case only counts the not be taken as a measure of doctors’ productivity, since first contact over a three-month period, even if the consultations can vary in length and effectiveness, and patient consults a doctor more often, leading to an also because it excludes the services doctors deliver for under-estimation). Telephone contacts are hospital inpatients and other tasks. The duration of included in a few countries (e.g. Spain). consultations with a primary care doctor in Sweden, as reported by doctors themselves, tends to be longer than in other countries such as the Netherlands and Germany (Commonwealth Fund, 2015). However, from a patient’s perspective, a lower proportion of patients in References Sweden report that their primary care doctors spent Commonwealth Fund (2015), International Health Policy enough time with them in consultation (see indicator on Survey of Primary Care Doctors, New York. Patient experience in Chapter 6). Maier, C. et al. (2017), “Nurses in Advanced Roles in Looking at trends over time in the estimated Primary Care: Policy Levers for Implementation”, OECD number of consultations per doctor per year, the Health Working Papers , No. 98, OECD Publishing, Paris, number has decreased at least slightly since 2000 in http://dx.doi.org/10.1787/a8756593-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 182

185 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.15. Number of doctor consultations per person, 2016 (or nearest year) Annual consultations per person 14 11.5 11.1 11.1 12 10.0 9.2 8.9 8.8 10 7.7 7.6 7.5 7.5 7.0 6.9 6.8 8 6.7 6.6 6.4 6.3 6.1 6.1 5.9 5.9 5.9 5.8 5.0 6 4.4 4.3 4.2 2.9 4 2 0 Source: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en ; Eurostat Database. http://dx.doi.org/10.1787/888933836466 1 2 7.16. Estimated number of consultations per doctor, 2016 (or nearest year) Annual consultations per doctor 4000 3457 3311 3500 3104 3010 2970 2741 3000 2606 2516 2389 2245 2223 2500 2147 2060 2046 2000 1977 1976 1948 1837 1823 1760 1744 2000 1619 1474 1310 1287 1196 1500 975 679 1000 500 0 ; Eurostat Database. OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en Source: http://dx.doi.org/10.1787/888933836485 1 2 7.17. Evolution in the estimated number of consultations per doctor, selected EU countries, 2000 to 2016 Sweden Poland Germany Austria France Denmark Annual consultations per doctor 3500 3000 2500 2000 1500 1000 500 0 2006 2001 2002 2003 2004 2005 2007 2000 2008 2009 2010 2011 2012 2013 2014 2015 2016 ; Eurostat Database. Source: OECD Health Statistics 2018, https://doi.org/10.1787/health-data-en http://dx.doi.org/10.1787/888933836504 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 183

186 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES AVAILABILITY AND USE OF DIAGNOSTIC TECHNOLOGIES This section presents data on the availability and some places, which could be funded by reducing use of two diagnostic imaging technologies: computed imaging tests of lower value in other places (Public tomography (CT) scanners and magnetic resonance Health England and NHS Right Care, 2017). imaging (MRI) units. CT and MRI exams help physicians Clinical guidelines have been developed in some diagnose a wide range of conditions. Unlike European countries to promote a more rational use of conventional radiography and CT scanning, MRI exams these diagnostic technologies. Through the Choosing do not expose patients to ionising radiation. Wisely® campaign, which began in the United States There is no general guideline or benchmark in 2012 and emulated in a growing number of countries regarding the ideal number of CT scanners or MRI units since then, some medical societies have identified or exams per population. However, if there are too few different cases when an MRI or CT exam is not units, this may lead to access problems in terms of necessary. For example, the Royal College of Physicians geographic proximity or waiting times. If there are too in the United Kingdom has recommended, based on many, this may result in an overuse of these costly evidence from the National Institute for Health and diagnostic procedures, with little if any benefits for Clinical Excellence (NICE), that patients with low back patients. pain or with suspected migraine do not routinely need an imaging test (Choosing Wisely UK, 2018). The availability and use of CT scanners and MRI units have increased rapidly in most European countries over the past two decades, but there remain large differences. Hungary, Romania and the Definition and comparability United Kingdom have the lowest number of MRI units While the data in most countries cover CT and CT scanners per capita among EU countries, scanners and MRI units installed both in hospitals whereas Germany, Italy, Greece and Finland have the and the ambulatory sector, the data coverage is highest number of MRI units per capita, and Denmark, more limited in some countries. CT scanners and Greece, Latvia and Germany have the highest number MRI units outside hospitals are not included in of CT scanners per capita (Figure 7.18 and Figure 7.19). some countries (e.g. Belgium, Portugal and Sweden, In Greece, many MRI and CT scanners are over as well as Switzerland for MRI units). For the 10 years old, and considered to be outdated and no United Kingdom, the data only include scanners in longer adequate for conducting some exams, because the public sector. For Hungary, the data cover only all exams are reimbursed at the same rate regardless of equipment eligible for public reimbursement. the age of the equipment. In other countries like Denmark provides data on the number of CT France, the reimbursement for an MRI or CT exam is scanners, but not on the number of MRI units. reduced after a number of years, once the equipment is Similarly, MRI and CT exams performed outside considered to be depreciated. hospitals are not included in some countries (e.g. Data on the use of these diagnostic machines Austria, Cyprus, Portugal, Switzerland and the show that the number of MRI exams and CT exams per United Kingdom). Furthermore, MRI and CT capita is lowest in Bulgaria and Romania, while they exams for Cyprus not only cover public hospitals. are highest in Germany (for MRI exams only), France, The Netherlands only report data on publicly Belgium and Luxembourg (Figure 7.20 and Figure 7.21). financed exams. Ireland and Sweden do report These large variations in MRI and CT exams may any data on MRI and CT exams. indicate either an under-use in some countries or an over-use in others. There are wide variations in MRI and CT exams not only across countries, but also within countries, References suggesting differences in clinical practices. For example, in Belgium, recent analysis shows a 50% variation in the Choosing Wisely UK (2018), Clinicians Recommendations: use of diagnostic exams of the spine between those Royal College of Physicians, www.choosingwisely.co.uk/ i-am-a-clinician/recommendations/#1528717718592- provinces with the highest and lowest rates in 2017, and . 17c3e7e1-94f2 this variation is even larger across smaller areas (INAMI/ RIVIZ, forthcoming). In the United Kingdom, the 2nd INAMI/RIVIZ (forthcoming), Medical Practice Variations – Medical Imaging, Brussels. Atlas of Variation in NHS Diagnostic Services in England found large variations in CT and MRI exams across Public Health England and NHS RightCare (2017), The 2nd geographic regions, and concluded that there is a need Atlas ofVariation in NHS Diagnostic Services in England, for certain types of imaging tests to be increased in January 2017, London. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 184

187 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.19. CT scanners, 2016 (or nearest year) 7.18. MRI units, 2016 (or nearest year) Denmark Germany 39 34 Greece 37 Italy 28 Latvia 36 Greece 27 Germany 35 Finland 25 Bulgaria 35 Austria 22 Italy 34 Cyprus 20 Cyprus 34 EU27 17 Austria 29 Spain 16 Finland 24 Sweden¹ 16 Belgium¹ 23 Ireland 15 Lithuania 23 Latvia 14 EU28 22 Estonia 14 Sweden¹ 22 France 14 Portugal¹ 21 Netherlands 13 Malta 20 Lithuania 12 Spain 18 Luxembourg 12 Croatia 18 Belgium¹ 12 Estonia 17 Slovenia 11 Poland 17 Malta 11 Slovak Republic 17 Croatia 11 Ireland 17 Slovak Republic 9 Luxembourg 17 Czech Republic 9 France 17 Poland 8 Czech Republic 16 Bulgaria 8 Slovenia 14 Portugal¹ 8 Netherlands 13 United Kingdom 7 Romania 13 Romania 6 United Kingdom 9 Hungary² 4 Hungary² 9 Switzerland¹ 22 Switzerland 39 Iceland 21 Iceland 39 Turkey 11 Turkey 15 Serbia 3 Serbia 10 FYR of Macedonia 3 FYR of Macedonia 7 0 10203040 0 10203040 population Per million population Per million 1. Equipment outside hospital not included. 1. Equipment outside hospital not included. 2. Only equipment eligible for public reimbursement. 2. Only equipment eligible for public reimbursement. OECD Health Statistics 2018, https://doi.org/10.1787/health- OECD Health Statistics 2018, https://doi.org/10.1787/health- Source: Source: data-en ; Eurostat Database. ; Eurostat Database. data-en 1 2 http://dx.doi.org/10.1787/888933836523 1 2 http://dx.doi.org/10.1787/888933836542 7.20. MRI exams, 2016 (or nearest year) 7.21. CT exams, 2016 (or nearest year) Germany Luxembourg 136 211 France 204 France 114 Belgium 200 Belgium 89 Portugal¹ 179 Luxembourg 83 Latvia 178 Spain 83 Estonia 166 Denmark 82 Slovak Republic 162 EU26 76 Denmark 161 Italy 67 Greece 150 Greece 64 Austria¹ 148 Slovak Republic 61 Germany 143 Austria¹ 57 Cyprus 132 United Kingdom¹ 57 EU26 122 Malta 54 Hungary 116 Estonia 53 Poland 109 Slovenia 52 Spain 109 Czech Republic 50 Czech Republic 107 Netherlands² 49 Lithuania 100 Latvia 48 Malta 90 Croatia 47 Netherlands² 89 Lithuania 45 United Kingdom¹ 85 Hungary 42 Croatia 84 Finland 39 Italy 81 Portugal¹ 39 Slovenia 65 Poland 30 Bulgaria 60 Romania 11 Finland 37 Bulgaria 8 Romania 27 Cyprus 7 Iceland 205 Iceland 93 Turkey 189 Switzerland¹ 72 Switzerland¹ 104 Serbia Serbia 47 15 FYR of Macedonia 29 FYR of Macedonia 11 0 200 50 100 150 250 0 306090120150 Per 1 000 population Per 1 000 population 1. Exams outside hospital not included. 1. Exams outside hospital not included. 2. Exams privately-funded not included. 2. Exams privately-funded not included. https://doi.org/10.1787/health- OECD Health Statistics 2018, Source: https://doi.org/10.1787/health- OECD Health Statistics 2018, Source: data-en data-en ; Eurostat Database. ; Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933836561 http://dx.doi.org/10.1787/888933836580 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 185

188 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES HOSPITAL BEDS AND DISCHARGES The number of hospital beds provides an Ireland where occupancy rates for curative (acute) care indication of the resources available for delivering was approaching 100% in 2016, far above any other services to inpatients in hospitals. The influence of the countries. In countries like Belgium and Germany, bed supply of hospital beds on hospital admission rates has occupancy rates have remained relatively stable since been widely documented, confirming that a greater 2000, at around 80%. The EU average has also been supply generally leads to greater admissions (Rohmer’s stable at 77% (Figure 7.24). law that a “built bed is a filled bed”). Germany, Austria and Bulgaria have the highest number of hospital beds per capita, with more than Definition and comparability seven beds per 1 000 population in 2016, well above the Hospital beds include all beds that are regularly EUaverageofjustoverfivebeds,andmorethan maintained and staffed and are immediately two-times greater than the supply in Sweden, the available for use. They include beds in general United Kingdom and Denmark (Figure 7.22). hospitals, mental health and substance abuse Since 2000, the number of hospital beds per capita hospitals, and other specialty hospitals. Beds in has decreased to some extent at least in all EU countries. nursing and residential care facilities are On average, it fell by almost 20%. This reduction has excluded. Data for some countries do not cover all been particularly pronounced in Finland, Estonia, Latvia hospitals. In the United Kingdom, data are and Lithuania. The reduction in the supply of hospital restricted to public hospitals. In Ireland, data refer beds has been accompanied by a reduction in hospital to publicly funded acute hospitals only. admissions in some countries and a reduction in Discharge is defined as the release of a patient average length of stays in nearly all countries (see who has stayed at least one night in hospital. Chapter 8). indicator on average length of stay in Same-day separations are excluded. Healthy Hospital admissions and discharges are highest in babies born in hospitals are excluded completely the three countries that have the highest number of (or almost completely) from hospital discharge hospital beds – Bulgaria, Germany and Austria. rates in several countries (e.g. Austria, Estonia, Hospital discharge rates in these countries are about Finland, France, Greece, Ireland, Latvia, Luxembourg 50% higher than the EU average. While differences in and Spain). These comprise between 3% and 10% the clinical needs of patients may explain some of the of all discharges. Data for some countries do not variations in admission and discharge rates, these cover all hospitals. In Ireland, Latvia and the variations also likely reflect differences in the supply of United Kingdom, data are restricted to public or beds, clinical practices and payment systems. Since publicly funded hospitals only. Data for Portugal 2000, hospital discharge rates have increased in relate only to public hospitals on the mainland. Bulgaria and Germany. Data for Cyprus are not shown as they only include discharges from public hospitals, resulting in a Across EU countries, the main conditions leading large under-estimation given that most hospitals to hospitalisation in 2016 were circulatory diseases, are private. Data for Belgium, Ireland and the pregnancy and childbirth, injuries and other external Netherlands include only acute care/short-stay causes, diseases of the digestive system, respiratory hospitals, also resulting in some under-estimation. diseases and cancers. The occupancy rate for curative (acute) care Hospital discharge rates vary not only across beds is calculated as the number of hospital bed- countries but also within countries. In several days related to curative care divided by the European countries (e.g. Finland, Germany, Italy, number of available curative care beds (multiplied Portugal, Spain and the United Kingdom), hospital by 365). medical admissions (excluding admissions for surgical interventions) vary by more than two-fold across different regions in the country. This may be related not only to differences in the supply of hospital beds, but also in the availability and quality of primary care Reference services (OECD, 2014). Geographic Variations in Health Care Use: What OECD (2014), Hospital bed occupancy rates have increased over Do We Know and What Can Be Done to Improve Health time in some countries that have relatively low number System Performance? , OECD Publishing, Paris, http:// of hospital beds. This has been notably the case in dx.doi.org/10.1787/9789264216594-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 186

189 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.22. Hospital beds per 1 000 population, 2000 and 2016 (or nearest year) 2016 2000 Per 1 000 population 10 9 8 7 8.1 6 7.4 7.3 7.0 6.9 6.8 6.7 6.6 5 6.1 5.8 5.7 5.7 5.6 5.5 4 2.8 5.1 4.8 4.8 4.7 4.6 4.5 3 4.4 4.2 4.0 3.9 3.7 3.6 2 3.4 3.4 3.2 3.1 3.0 2.9 2.6 2.6 1 2.3 0 ; Eurostat Database. https://doi.org/10.1787/health-data-en Source: OECD Health Statistics 2018, http://dx.doi.org/10.1787/888933836599 1 2 7.23. Hospital discharges per 1 000 population, 2000 and 2016 (or nearest year) 2000 2016 Per 1 000 population 350 317 257 300 250 197 196 183 174 179 176 168 253 200 236 119 207 109 150 200 199 191 182 172 164 166 166 166 154 100 147 145 145 142 136 131 117 117 115 50 100 97 0 1. Data exclude discharges of healthy babies born in hospital (between 3-10% of all discharges). 2. Data include discharges for curative (acute) care only. ; Eurostat Database. Source: https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, 1 2 http://dx.doi.org/10.1787/888933836618 7.24. Occupancy rate of curative (acute) care beds, 2000 and 2016 (or nearest year) 2016 2000 % 100 94 90 83 81 78 77 75 80 74 84 71 81 67 67 80 78 65 70 76 75 74 72 71 71 70 69 69 68 60 50 OECD Health Statistics 2018, Source: ; Eurostat Database. https://doi.org/10.1787/health-data-en 1 2 http://dx.doi.org/10.1787/888933836637 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 187

190 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES WAITING TIMES FOR ELECTIVE SURGERY Looking at people who are still on the waiting lists, Long waiting times for elective (non-emergency) the percentage of patients who have been waiting for surgery are an important policy issue in many more than three months also varies widely across the European countries as they generate dissatisfaction for group of countries for which data are available.While only patients because the expected benefits of treatments about 12% of people in Sweden have been on the waiting are postponed, and the pain and disability remain lists for a cataract surgery or a hip replacement for more while waiting. than three months, this is the case for over 85% of people Waiting times are the result of a complex in Estonia and Poland (Figure 7.27 and Figure 7.28). In interaction between the demand and supply of health Ireland, the percentage of people still on the waiting lists services. The demand for elective surgery is determined after three months has increased sharply between 2010 by the health needs of the population, progress in and 2016, from about 50% to 77% for cataract surgery and medical and surgical technologies, and patient from about 50% to 63% for hip replacement. A number of preferences. However, doctors play a crucial role in the initiatives have been launched in recent years to try to decision to operate a patient or not. On the supply side, address long waiting times in Ireland, but these initiatives the availability of surgeons and other staff in surgical do not appear to have had any lasting effect. teams, as well as the supply of the required equipment Over the past decade, waiting time guarantees affect surgical activity rates. have become the most common policy tool to tackle The data presented here focus on two high-volume long waiting times in several countries. However, these procedures: cataract surgery and hip replacement. guarantees are only effective if they are enforced. There In 2016, the average waiting times for people who are two main approaches to enforcement: setting were operated on for a cataract surgery ranged from waiting time standards and holding providers just over a month in the Netherlands, to three to four accountable for achieving these standards; or allowing months in Finland, Spain and Portugal, and to well over patients to choose alternative health providers a year in Poland (Figure 7.25). The median waiting (including the private sector) if they have to wait beyond times (which are lower than the average in all countries) a maximum amount of time (Siciliani et al., 2013). ranged from about one month in Italy and Hungary, to about three months in Finland and Spain, but still to well over a year in Poland. Looking at trends over time, Definition and comparability in many countries, waiting times to get a cataract surgery declined fairly rapidly up to around 2010, but Two different measures of waiting times are have started to rise again in recent years. presented here: 1) the period from the time that a The average waiting times to get a hip replacement specialist adds a patient to the waiting list for an in 2016 ranged from about one to two months in the operation to the time that the patient receives the Netherlands and Denmark, to four to five months in operation; and 2) the waiting times for patients who Hungary, Portugal and Spain, and to well over a year in are still on the waiting lists at a given point in time. Poland (Figure 7.26). The median waiting times were Waiting times for the first measure are reported about 40 days in Denmark and 50 days in Italy, while both in terms of the average and the median they reached over 200 days in Poland and Estonia. In the number of days. The median is the value which United Kingdom, the waiting times for a hip separates a distribution in two equal parts (meaning replacement fell sharply up to 2008, but have remained that half the patients have longer waiting times and stable since then at around 80 days. In Portugal, the the other half lower waiting times). Compared with waiting times for a hip replacement followed the same the average, the median minimises the influence of pattern as for a cataract surgery: they fell substantially outliers (patients with very long waiting times). up to 2010, but have gone up since then to over 100 days, The data come from administrative databases despite a slight reduction in 2016. The waiting times for (not surveys). The management of administrative a hip replacement have also increased in Spain since data can vary across countries: in some countries, 2011 and in Estonia since 2014. patients who refuse on several occasions to Poland has the longest waiting times for both receive the procedure are removed from the list, cataract surgery and hip replacement among EU while they continue to be kept on the list in other countries reporting these data, and these waiting times countries (e.g. Estonia). have increased substantially since 2010. Surgical activities in Poland are constrained by the low number of surgeons and the lack of equipment. The uneven geographic distribution of resources and services also Reference contributes to the problem: the waiting times for some surgical specialties can be very long for people living in Siciliani, L., M. Borowitz and V. Moran (eds.) (2013), underserved regions. The Polish government has taken Waiting Time Policies in the Health Sector: What Works? , a series of measures in recent years to try to reduce OECD Health Policy Studies, OECD Publishing, Paris, these long waiting times. http://dx.doi.org/10.1787/9789264179080-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 188

191 II.7. ACCESSIBILITY: AFFORDABILITY, AVAILABILITY AND USE OF SERVICES 7.25. Waiting times of patients for cataract surgery, 2016 and trends since 2005 Denmark Estonia Average Median Netherlands Portugal United Kingdom Spain Days Days 600 250 484 500 200 420 400 150 283 300 100 200 142 120 114 109 106 97 91 89 85 74 50 69 59 57 56 52 100 48 41 36 36 28 n.a. 0 0 Note: On the right panel, data relate to median waiting times, except for the Netherlands and Spain (average waiting times). https://doi.org/10.1787/health-data-en Source: OECD Health Statistics 2018, . 1 2 http://dx.doi.org/10.1787/888933836656 7.26. Waiting times of patients for hip replacement, 2016 and trends since 2005 Denmark Estonia Average Median Portugal Netherlands United Kingdom Spain Days Days 500 250 444 200 400 326 276 300 150 211 100 200 158 134 133 130 120 110 105 104 104 90 84 82 80 75 64 100 50 52 51 45 37 n.a. 0 0 Note: On the right panel, data relate to median waiting times, except for the Netherlands and Spain (average waiting times). . https://doi.org/10.1787/health-data-en OECD Health Statistics 2018, Source: http://dx.doi.org/10.1787/888933836675 1 2 7.28. Waiting times of patients still on 7.27. Waiting times of patients still on waiting list for cataract surgery, 2005 to 2016 waiting list for hip replacement, 2005 to 2016 2010 2016 2005 2016 2010 2005 % of patients waiting more than 3 months % of patients waiting more than 3 months 100 100 88 87 87 87 86 86 84 83 83 78 77 77 74 73 73 69 75 75 66 65 63 62 59 57 55 54 51 49 45 45 42 50 50 40 33 30 30 26 22 22 19 25 25 14 12 12 0 0 OECD Health Statistics 2018, Source: OECD Health Statistics 2018, https://doi.org/10.1787/health- Source: https://doi.org/10.1787/health- . data-en data-en . 1 2 1 2 http://dx.doi.org/10.1787/888933836694 http://dx.doi.org/10.1787/888933836713 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 189

192

193 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 P ART II Chapter 8 Resilience: Innovation, efficiency and fiscal sustainability This chapter presents a series of indicators related to the resilience of health systems, defined as the capacity of health systems to adapt efficiently to changing economic, technological and demographic environments. Digital technology offers great opportunities to deliver health services more efficiently, and the European Commission supports a digital transformation of health systems to empower citizens to have access to their health data and to promote exchange of health data among health care providers across the EU. The use of eHealth and ePrescribing continues to grow in many EU countries, although some countries are lagging behind. The ageing of the population requires profound transformations in health systems, from a focus on acute care in hospitals to more integrated and people-centred care in the community. Many EU countries have begun this transformation over the past 15 years – for example by reducing the average length of stay in hospitals and promoting a greater use of day surgery combined with follow-up care, but the process still requires ongoing, long-term effort. Population ageing means not only that there will be growing needs for health care in the years ahead, but also growing needs for long-term care. The latest projections from the EC confirm that public spending on long-term care is projected to grow faster than public spending on health care over the coming decades, highlighting the importance to find more innovative ways to respond to health care and long-term care needs more efficiently. 191

194 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY ADOPTION AND USE OF ELECTRONIC MEDICAL RECORDS AND EPRESCRIBING Health care that is safe, effective, timely, efficient On the other hand, ePrescribing has not been and patient-centred relies on the right information implemented yet in several countries (such as Bulgaria, reaching the right person (or organisation) at the right Cyprus, France, Germany, Ireland, Luxembourg, Malta time. A digitalised information infrastructure that and Poland), although all these countries have stated ensures timely and reliable sharing of clinical and that they plan to start implementing ePrescribing at other information can improve health outcomes and regional or national levels over the next few years. efficiency, and also create a repository of valuable data for researchers and system managers (OECD, 2017). Enabling people to access, and interact with, their Definition and comparability electronic medical record (EMR) is an important feature An Electronic Medical Record (EMR) is a that can help people become more involved in their computerised medical record created in an health and their care. organisation that delivers care, such as a hospital The European Commission’s Digital Single Market or physician’s office, for patients of that Strategy includes three pillars to improve the health organisation. Ideally, EMRs should be shared and care sector across the EU: 1) to secure access to and between providers and settings to provide a sharing of personal health information across borders, detailed history of contact with the health care with the intention of going beyond ePrescriptions and system for individual patients from multiple patient summaries and establish full interoperability of organisations (Oderkirk, 2017). The figures member states’ EMRs and a European exchange format presented on EMR implementation come from a for electronic records; 2) to connect and share health 2016 survey of OECD countries to which 15 EU data to enable research, better diagnosis and improved countries responded.The same survey was carried health; and 3) to strengthen citizen empowerment and out in 2012, with 8 responses from EU countries. individual care through eHealth solutions and new ePrescribing is the computer-based electronic care models (European Commission, 2018). generation, transmission and filing of a medicine Many countries are implementing EMRs across prescription. It allows prescribers to write health care settings, including primary care. In 2016, the prescriptions that can be retrieved by a pharmacy proportion of primary care practices using an EMR was electronically without the need for a paper about 80% on average across 15 EU countries, although prescription. ePrescribing systems may also be While an EMR was there are wide variations (Figure 8.1). linked or integrated to the reimbursement and used in all or nearly all primary care practices in claiming system. The figures presented on Denmark, Estonia, Finland, Greece, Spain, Sweden ePrescribing are derived from a 2018 survey of the and the United Kingdom, its use was much more Pharmaceutical Group of the European Union limited in Croatia and Poland. In Denmark and the (PGEU). United Kingdom, the proportion of primary care practices using an EMR doubled between 2012 and 2016. In most of these 15 countries, patients are able to view information contained in their electronic record References (with the only exceptions being Croatia, the Czech Republic and Ireland), and in half of these countries (Denmark, European Commission (2018), Communication from the Estonia, France, Greece, Latvia, Luxembourg, Spain and Commission to the European Parliament, the Council, the European Economic and Social Committee and Sweden), patients are also able to interact with their the Committee of the Regions on enabling the digital record, for example to add or amend information transformation of health and care in the Digital Single (Oderkirk, 2017). Market; empowering citizens and building a healthier ePrescribing, which allows prescribers to write society, https://eur-lex.europa.eu/legal-content/EN/TXT/ prescriptions that can be retrieved by a pharmacy . ?uri=COM:2018:233:FIN electronically, can improve the accuracy and efficiency Oderkirk, J. (2017), “Readiness of electronic health record of pharmaceutical drug dispensing. Most countries are systems to contribute to national health information transitioning from paper-based to ePrescribing, but the and research”, OECD Health Working Papers , No. 99, OECD Publishing, Paris, https://doi.org/10.1787/9e296bf3-en . implementation of ePrescribing varies greatly across the EU (Figure 8.2). In 2018, over 90% of prescriptions were OECD (2017), New Health Technologies: Managing Access, transmitted to community pharmacies electronically in Value and Sustainability https:// , OECD Publishing, Paris, Finland, Estonia, Sweden, Denmark, Portugal and Spain. doi.org/10.1787/9789264266438-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 192

195 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY 8.1. Percentage of primary care physician offices using electronic medical records, 2012 and 2016 2016 2012 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% OECD Survey of Electronic Health Record System Development and Use, 2012 and 2016. Source: http://dx.doi.org/10.1787/888933836732 1 2 8.2. Percentage of ePrescriptions in community pharmacies, 2018 100 88 98 99 n.a. 97 n.a. ePrescribing (%) 0 0 90 0 1 - 50 0 30 51 - 100 85 0 30 n.a. 75 n.a. 0 90 n.a. 80 n.a. 80 85 0 n.a. 0 n.a. 92 96 n.a. 0 0 (Malta) Note: Greece and the Netherlands are implementing ePrescribing but the percentage was not reported. Pharmaceutical Group of the European Union (PGEU). Source: 1 2 http://dx.doi.org/10.1787/888933836751 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 193

196 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY INDIVIDUALS USING THE INTERNET TO ACCESS HEALTH SERVICES AND HEALTH INFORMATION Disparities by age and socioeconomic groups exist Access to the internet among Europeans is rising in using the internet for health-related purposes. In 2017, (85% of EU households had internet access in 2016) and only about 30% of people in EU countries aged 65-74 people are increasingly going on line to access accessed health information on line, compared to 55% of information and interact with the provider of different those aged 25-64. This “age gap” in using the internet for services. Health is no exception. The amount of health-related information was particularly large in information regarding health and illness available on Croatia, Greece and Malta. In terms of socioeconomic line is growing, as are the opportunities of interacting status, about 40% of people in EU countries living in with health care providers electronically, for example households in the lowest income quartile accessed to make medical appointments. Digital technologies health information on line, compared to over 60% in the can improve patient experience and outcomes, and the highest income quartile. This “income gap” in accessing efficiency of services, but some may generate minimal health information on the internet was particularly large benefit (at a considerable expense), and protecting in Hungary, Lithuania and Portugal. individuals’ privacy is a frequent problem and a policy priority (OECD, 2017). While online medical information can be a useful way to help people manage their health, regulation is difficult and many people are Definition and comparability not in a position to check the veracity of this type of information. The figures presented here come from an annual One in eight EU residents (13%) made an European Information and Communication appointment with a health care practitioner on line in Technologies (ICT) survey of households and 2016, up from one in twelve (8%) in 2012 (Figure 8.3). individuals. Data are collected by National Almost half (49%) of Danish residents made a medical Statistical Institutes based on Eurostat’s model appointment on line in 2016 (up from 29% in 2012). questionnaire on ICT usage. The model Finland and Spain had the second and third highest questionnaire changes every year. proportion of residents making a medical appointment Around 150 000 households and 200 000 this way in 2016, with 35% and 30% respectively. individuals aged 16-74 in the EU were surveyed Virtually no Cypriots reported making a medical appointment on line in either year. The figure was also in 2016 (Eurostat, 2016). low in Greece and Bulgaria (2% and 3% respectively in In the 2016 survey, the question related to the 2016). In all countries except Cyprus, the proportion of activities described here was: residents making appointments on line increased ● For which of the following activities did you use the between 2012 and 2016, on average by 63%. The greatest increases were observed in Denmark, Belgium, Internet in the last 3 months for private purpose? the Netherlands, Luxembourg and Hungary. (tick all that apply) Making medical appointments on line had a weak Seeking health-related information (e.g. injury, ❖ correlation with internet access (r2 = 0.34), suggesting that oving health, etc.) impr , nutrition, disease internet access is not a sufficient condition to making medical appointments on line. A moderate correlation ❖ Making an appointment with a practitioner via the (r2 = 0.51) was observed with internet banking, which was bsite (e .g . of a hospital or a health care centre) we performed by 49% of EU residents in 2016, suggesting that The 2017 survey did not include the question individuals who conduct their banking on line also tend to on making an appointment via the website. book medical appointments this way. The correlation with the percentage of individuals booking travel and Data on internet access and use for personal accommodation on line (41% across the EU) was weak banking and booking travel and accommodation (r2 = 0.32). These figures suggest that internet use for come from the same surveys. making medical appointments is lagging behind use for other personal services. Half of all EU residents sought health information on line in 2017, a figure that has almost doubled since 2008 (Figure 8.4). The highest proportions of people References seeking health information on line were in the New Health Technologies: Managing Access, OECD (2017), Netherlands and Finland (about 70%). Almost 60% of https:// Value and Sustainability , OECD Publishing, Paris, Cypriots sought health information on line in 2017, a . doi.org/10.1787/9789264266438-en large increase from only about 10% in 2008. Less than 40% of Romanian, Italian, Bulgarian and Irish residents Eurostat (2016), Methodological manual for statistics on reported that they sought health information on line https://circabc.europa.eu/faces/ the Information Society, in 2017. jsp/extension/wai/navigation/container.jsp . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 194

197 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY 8.3. Percentage of people who made an appointment with a health practitioner on line, 2012 and 2016 2016 2012 % 50 49 45 40 35 35 30 30 25 22 22 22 25 21 20 18 20 16 15 14 13 15 11 10 10 99 7777 10 66 55 4 3 3 2 5 0 0 Eurostat Database, based on the European ICT survey of individuals aged 16-74. Source: 1 2 http://dx.doi.org/10.1787/888933836770 8.4. Percentage of people who sought health-related information on line, 2008 and 2017 2017 2008 % 80 71 69 65 70 64 64 64 63 63 59 59 58 58 57 57 56 55 60 54 54 53 52 52 51 51 51 50 47 45 45 50 44 44 43 43 37 40 34 33 33 30 20 10 0 Eurostat Database, based on the European ICT survey of individuals aged 16-74. Source: 1 2 http://dx.doi.org/10.1787/888933836789 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 195

198 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY PUBLIC HEALTH LABORATORY CAPACITY TO CONTROL INFECTIOUS DISEASES THREATS Infectious diseases and resistance to antibiotics microbial genomic sequencing methods and digital are global public health threats. Resilient health connectivity of laboratory information with public systemsdependontheabilitytodetectemerging health monitoring systems at national and EU levels diseases accurately in time to stop outbreaks and avert (ECDC, 2018; Revez, 2018). major international crises such as the recent Ebola epidemic (Albiger, 2018). Public health preparedness requires adequate capacity of microbiology Definition and comparability laboratories to: 1) ensure rapid infection diagnostics to The results presented here are derived from the guide treatment, detect and control epidemics; EULabCap monitoring surveys conducted 2) characterise infectious agents for designing effective annually in EU countries, Iceland and Norway by vaccines and control measures; and 3) monitor the ECDC jointly with National Microbiology Focal impact of prevention of infections and containment of Points since 2013. The EULabCap Index is a antimicrobial resistance (AMR). composite index composed of 60 technical The ECDC is operating the EULabCap (European indicators of laboratory structure, service range Laboratory Capability Monitoring System) to assess and outputs related to 12 public health targets whether laboratory systems in EU/EEA countries have aligned with EU policies and international the critical capabilities and capacities for reliable standards and health regulations. The target communicable disease and antimicrobial resistance measures are aggregated into the EULabCap surveillance and control at Member State and EU levels Country system index, with 10 being the (ECDC, 2018). In 2016, the EULabCap Index average for EU maximum score. The methodology is described in countries was 7.5 on a maximum scale of 10 (F igure 8.5). further detail in the EULabCap report (ECDC, 2018). Country scores ranged from a low of 5.6 in Cyprus to a Data completeness is robust with 100% of high of 9.6 in France. These results indicate that the EU countries and 97% of indicators data reported. has strong public health microbiology services that However, the following limitations should be largely meet communicable disease surveillance and taken into account: 1) variable relevance for response requirements. However, only 18 EU countries applicability of some indicators according to (and Norway) showed sufficient laboratory capacity differences in national health systems or levels (defined as intermediate to high score) for at least epidemiology; 2) country self-reported data; 10 of the 12 EULabCap targets (ECDC, 2018). 3) indirect measurement of national capacity National improvements in the areas of using EU-reported surveillance data; and 4) threat vulnerability have taken place in 24 EU countries since to comparability over time caused by annual 2015. Steady increases in the EULabCap Index, and a updates of indicators following laboratory narrowing score range between countries, indicate technology innovation. convergence towards more balanced laboratory capacities across countries. Capabilities to diagnose EU notifiable diseases and antimicrobial resistance as well as laboratory References contribution to surveillance networks are well in line with EU legislation and case definitions across the EU. Albiger, B. et al. (2018), “Networking of Public Health Microbiology Laboratories Bolsters Europe’s Defenses Capacities for detection and surveillance of , against Infectious Diseases”, Frontiers in Public Health antimicrobial resistance improved steadily between Vol. 6, Issue 46. 2013 and 2016 with wider use of standardised methods (Figure 8.6). EU capabilities of national reference ECDC (2018), “EU Laboratory Capability Monitoring System (EULabCap) – Report on 2016 survey of EU/EEA country laboratories for rapid detection of (re-) emerging capabilities and capacities”, Stockholm, https:// diseases and drug resistance and participation in ecdc.europa.eu/sites/portal/files/documents/2016_ outbreak investigations also progressed over the years . EULabCap_EUreport_web_300418_final.pdf (Figure 8.7). Revez, J. et al. (2018), “Survey on the Use of Whole- Some remaining gaps and inefficiencies still need Genome Sequencing for Infectious Diseases to be addressed, including the development of clinical Surveillance: Rapid Expansion of European National guidance for and adequate utilisation of diagnostic Frontiers in Public Health , Vol. 5, Capacities, 2015-2016”, tests, upgrading surveillance programme to integrate Issue 347. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 196

199 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY 8.5. Composite index of national public health laboratory capacities, 2016 EULabCap composite index 9.6 10 9.0 8.9 8.9 8.5 8.4 8.4 8.3 8.3 8.2 9 8.1 8.1 7.9 7.7 7.5 7.5 7.4 7.2 7.2 8 7.1 7.1 7.0 6.8 6.8 6.4 6.4 6.3 6.3 7 6.0 5.9 5.6 6 5 4 3 2 1 0 The maximum score for this indicator is 10. Note: ECDC (2018). Source: http://dx.doi.org/10.1787/888933836808 1 2 8.6. Antimicrobial susceptibility testing and resistance monitoring, average across EU countries, 2013 to 2016 a. Antimicrobial susceptibility testing b. AMR monitoring EULabCap Index EULabCap Index 10 10 10.0 10.0 9.5 8 8 8.0 8.0 8.0 8.0 6 6 7.0 4 4 2 2 0 0 2013 2015 2014 2014 2015 2016 2016 2013 The shaded area shows the minimum and maximum values. Note: Source: ECDC (2018). http://dx.doi.org/10.1787/888933836827 1 2 8.7. (Re-) emerging disease preparedness and outbreak response support, average across EU countries, 2013 to 2016 b. Outbreak response support a. Preparedness response support EULabCap Index EULabCap Index 10 10 8 8.9 8 8.0 8.0 8.0 7.3 6 6 7.0 7.0 6.1 4 4 2 2 0 0 2015 2014 2013 2016 2015 2013 2016 2014 The shaded area shows the minimum and maximum values. Preparedness refers to capabilities of laboratories to detect and Note: characterise various infectious diseases. Source: ECDC (2018). 1 2 http://dx.doi.org/10.1787/888933836846 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 197

200 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY AVERAGE LENGTH OF STAY IN HOSPITAL The average length of stay in hospital is often (Figure 8.10). It was lowest in Denmark, Bulgaria and regarded as an indicator of efficiency in health service Sweden (less than five days) and highest in Germany delivery. All else being equal, a shorter stay will reduce (ten days). the cost per discharge and shift care from inpatient to Beyond differences in clinical needs, several less expensive settings. Longer stays can be a sign of factors can explain these cross-country variations in poor care coordination, resulting in some patients lengths of stay. The combination of an abundant waiting unnecessarily in hospital until rehabilitation or supply of beds together with hospital payment long-term care can be arranged (see the discussion on methods may provide incentives for hospitals to keep delayed discharges in Chapter 2). At the same time, patients longer. A growing number of countries some patients may be discharged too early, when (e.g. France, Germany, Poland) have moved to prospective staying in hospital longer could have improved their payment methods often based on diagnosis-related health outcomes or reduced chances of re-admission. groups (DRGs) to set payments based on the estimated In 2016, the average length of stay in hospital for all cost of hospital care for different patient groups in causes of hospitalisation was the lowest in the advance of service provision. These payment methods Netherlands, but the length of stay in the Netherlands is have the advantage of encouraging providers to reduce under-estimated because it only includes stays for the cost of each hospitalisation. curative (acute) care that are typically shorter.Taking into Strategic reductions in hospital bed numbers account all types of care, the average length of stay was alongside the development of community care services relatively short in Bulgaria, Denmark and Sweden can shorten the average length of stay. Lengths of stay (Figure 8.8). It was highest in France, mainly because of could often be shortened through better coordination relatively long stays for rehabilitative and psychiatric care between hospitals and post-discharge care settings. An provided in general or specialised hospitals: the length of important constraint in many countries is the shortage stay in acute care units in France is no longer than in of capacity in intermediate or long-term care facilities, most other countries. Hungary and the Czech Republic or in providing home-based care. Many countries (for also have relatively long average length of stay, partly example, the Netherlands, Sweden, Norway and parts of because many hospitals have long-term care units. the United Kingdom) have taken steps in recent years to The average length of stay in hospital has decreased increase the capacity of intermediate care facilities and since 2000 in nearly all EU countries, falling from almost home-based care to reduce lengths of stay and the risk ten days in 2000 to less than eight days in 2016 on of hospital re-admission (see Chapter 2). average. It fell particularly quickly in some countries that had relatively long stays in 2000 (e.g. Bulgaria, Croatia, Finland, Germany, Latvia, Slovak Republic and Definition and comparability United Kingdom).This reduction in average length of stay Average length of stay refers to the average has generally been accompanied by a reduction in the number of days that patients spend in hospital. number of hospital beds. For example, in Finland, the 30% It is generally measured by dividing the total reduction in average length of stay since 2000 has come number of days stayed by all inpatients during a along with an almost 50% reduction in the number of year by the number of admissions or discharges. hospital beds per capita (see indicator on hospital beds Day cases are excluded. Chapter 7). and discharges in The data cover all inpatient cases (including Focusing on average length of stay for specific not only curative/acute care cases), with the diseases or conditions can remove some of the effect of exception of the Netherlands where the data different case mix and severity of patients admitted to refer to curative/acute care only (resulting in a hospital. Figure 8.9 shows that the average length of substantial under-estimation). stay for a normal delivery in EU countries ranges from Average length of stay of healthy babies born in less than two days in the United Kingdom and the hospitals are excluded in several countries Netherlands, to almost five days in Hungary, Croatia (e.g. Austria, Cyprus, Estonia, Finland, France, and the Slovak Republic. The length of stay for a Greece, Ireland, Luxembourg), resulting in a slight normal delivery has become shorter in nearly all over-estimation of average length of stay compared countries, dropping from more than four days in 2000 with other countries. to about three days in 2016 on average in EU countries. Data for normal delivery refer to ICD-10 code The average length of stay following acute O80, and for AMI to ICD-10 codes I21-I22. myocardial infarction (AMI or heart attack) was around seven days on average in EU countries in 2016 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 198

201 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY 8.8. Average length of stay in hospital, 2000 and 2016 (or nearest year) 2016 2000 Days 14 12 9.5 10 7.8 10.2 10.1 9.3 8 9.1 9.0 8.9 8.8 8.8 6.2 8.5 8.3 8.3 8.2 7.9 7.9 7.7 7.7 7.5 7.5 7.4 7.3 6 7.1 7.1 7.0 6.9 6.8 6.0 6.0 5.8 5.8 5.4 5.3 4 5.0 4.0 2 Data refer to average length of stay for curative (acute) care (resulting in an under-estimation). Note: OECD Health Statistics 2018; Eurostat Database. Source: http://dx.doi.org/10.1787/888933836865 1 2 8.9. Average length of stay for normal 8.10. Average length of stay for acute delivery, 2016 (or nearest year) myocardial infarction (AMI), 2016 (or nearest year) Hungary 4.9 Croatia 4.8 Germany 10.0 Slovak Republic 4.8 Estonia 9.0 Romania 4.7 Lithuania 8.8 Cyprus 4.6 Hungary 8.0 Bulgaria 4.2 Austria 7.9 Italy 7.9 Czech Republic 4.1 Portugal 7.7 France 4.0 Croatia 7.5 Luxembourg 4.0 Slovenia 7.2 Poland 3.8 Spain 7.2 Slovenia 3.7 EU28 7.1 Austria 3.6 Belgium 7.0 Italy 3.4 Luxembourg 6.9 Latvia 3.3 Malta 6.7 United Kingdom 6.7 Lithuania 3.3 Ireland 6.5 EU26 3.1 Latvia 6.5 Belgium 3.1 Romania 6.4 Germany 3.0 Poland 6.2 Greece 3.0 Greece 6.0 Finland 2.9 Czech Republic 5.9 Malta 2.9 Finland 5.9 Ireland 2.4 France 5.5 Cyprus 5.2 Spain 2.4 Netherlands 5.0 Denmark 2.3 Slovak Republic 5.0 Sweden 2.3 Sweden 4.3 Netherlands 1.6 Bulgaria 4.0 United Kingdom 1.5 Denmark 3.7 Serbia 4.4 Serbia 7.8 Switzerland 3.2 Montenegro 7.2 Switzerland 6.8 Iceland 1.7 Iceland 5.0 Turkey 1.5 Turkey 4.8 Norway 3.6 0246 0481 2 Days Days OECD Health Statistics 2018; Eurostat Database. OECD Health Statistics 2018; Eurostat Database. Source: Source: http://dx.doi.org/10.1787/888933836884 1 2 http://dx.doi.org/10.1787/888933836903 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 199

202 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY DAY SURGERY between the most innovative hospitals or hospital Day surgery has expanded in EU countries over units and those lagging behind; and 3) providing the past few decades, thanks to progress in surgical proper financial incentives to ensure that health care techniques and anaesthesia, although the pace of providers (hospitals and surgical teams) do not lose diffusion has varied widely across countries. revenue by moving towards a greater use of day Cataract surgery, repair of inguinal hernia and surgery, and may even be financially better-off. These tonsillectomy provide good examples of surgical interventions are likely to be more effective if they are procedures that are now carried out mainly as day part of a comprehensive strategy to promote day surgery in many, but not all, EU countries. surgery. In Portugal, the strong growth in day surgery More than 95% of all cataract surgery are performed for cataracts and other interventions since 2000 has igure 8.11). as day surgery in about half of EU countries (F been supported by a comprehensive national plan Yet, the use of day surgery remains much more limited (Lemos, 2011). In Belgium, recent proposals for a in some Central and Eastern European countries further expansion of day surgery have also recognised (e.g. Romania, Poland, Bulgaria and Lithuania), the importance of addressing various barriers and accounting for less than half of all cataract operations. enabling factors at the same time (Leroy et al., 2017). Beyond possibly reflecting some limitations in data coverage, this low share of day surgery may also be due to higher reimbursement for inpatient stays, or legal or Definition and comparability capacity constraints imposed on the development of day surgery. In Hungary, the government recently abolished Cataract surgery consists of removing the lens of the budget cap on the number of day surgery that can be the eye because of the presence of cataracts that are performed in hospital, which has led to a substantial partially or completely clouding the lens, and growth in the number of day surgery for cataract. replacing it with an artificial lens. Repair of inguinal More than half of all inguinal hernia repair hernia is a surgery to repair a weakness in the interventions in many EU countries are now performed abdominal wall; the operation is now commonly as day surgery, whereas this proportion still remains performed laparoscopically (using minimally close to zero in other countries. On average across invasive surgery), allowing patients to return home countries, the share of day surgery for inguinal hernia more quickly. Tonsillectomy consists of removing repair rose from about 20% in 2000 to over 40% in 2016. the tonsils (glands at the back of the throat). Day surgery for inguinal hernia repair increased Day surgery is defined as the release of a patient particularly rapidly in countries like France and Portugal, who was admitted to a hospital for a planned which have moved closer to the share of over 80% in surgical procedure and was discharged the same leading countries (Denmark and the Netherlands). day. For cataract surgery and tonsillectomy, the Tonsillectomy is one of the most frequent surgical data also include outpatient cases (i.e. patients not procedures in children. Although the operation is formally admitted and discharged) where performed under general anaesthesia and generally available. However, several countries are not able involves a post-operative observation period of about 6 to report outpatient cases, leading to some under- to 8 hours, it is now carried out mainly as a day surgery estimation. In Ireland, Portugal and the in many countries, with children returning home the United Kingdom, the data only include cataract same day. As shown in Figure 8.13, more than half of all operations performed in public or publicly funded tonsillectomies are now performed as day surgery in hospitals, which may affect the share of day several EU countries, but there has not been any surgery if the volume of activities in private movement yet towards day surgery in other countries hospitals is substantial and if the practice pattern (e.g. Slovenia, Hungary, Austria, Cyprus and Bulgaria). in private hospitals differs from public hospitals. These variations in clinical practice likely reflect persisting differences in the perceived risks of postoperative complications and the maintenance of a clinical tradition in some countries of keeping children for at least one night in hospital after the operation. References As noted in Chapter 2, at least three broad policy Lemos, P. (2011), “A huge increase in ambulatory surgery levers can be used to promote the expansion of day practice in Portugal”, Ambulatory Surgery , Vol. 17, No. 1, March 2011. surgery: 1) publicly monitoring the progress in the use of day surgery at different levels (national, regional and Leroy, R. et al. (2017), Proposals for a further expansion of hospital levels); 2) supporting behavioural and clinical day surgery in Belgium, Belgian Health Care Knowledge changes, notably by promoting constructive exchanges Centre, KCE Reports 282. HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 200

203 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY 8.11. Share of cataract surgery performed as day cases, 2000 and 2016 (or nearest year) 2016 2000 % 99.6 99.2 98.9 98.8 98.8 98.2 99.2 98.0 97.8 97.2 97.1 96.7 95.5 94.9 94.0 92.2 92.1 100 84.1 84.2 82.5 81.1 76.2 80 64.8 63.4 54.9 53.3 60 44.7 36.9 36.2 35.3 32.2 40 20 0 Source: OECD Health Statistics 2018; Eurostat Database. http://dx.doi.org/10.1787/888933836922 1 2 8.12. Share of inguinal hernia repair performed as day cases, 2000 and 2016 (or nearest year) 2000 2016 % 100 84.8 80.2 76.2 72.7 68.5 68.3 80 63.2 59.6 58.2 55.5 55.0 47.1 46.9 60 41.1 40.8 31.1 40 18.1 16.5 9.3 9.0 20 4.3 3.7 3.2 2.4 0.7 0.3 0.1 0 Day cases do not include outpatient cases in countries where patients are not formally admitted to hospital. Note: OECD Health Statistics 2018; Eurostat Database. Source: 1 2 http://dx.doi.org/10.1787/888933836941 8.13. Share of tonsillectomy performed as day cases, 2000 and 2016 (or nearest year) 2000 2016 % 100 86.8 75.3 69.7 68.4 80 60.5 55.6 55.2 54.4 49.1 60 36.8 30.6 29.2 28.2 28.2 27.6 40 19.0 18.3 6.4 6.2 20 4.8 4.0 2.4 1.4 0.2 0.2 0.0 0.0 0.0 0 OECD Health Statistics 2018; Eurostat Database. Source: 1 2 http://dx.doi.org/10.1787/888933836960 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 201

204 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY CAPITAL EXPENDITURE IN THE HEALTH SECTOR While the health sector remains highly labour- Austria, Belgium and Sweden. France has seen intensive, capital investment in infrastructure and spending levels generally maintained over the period, medical equipment has been an increasingly and are typically 50% higher than in 2005. On the other important factor of production of health services in hand, a number of European countries experienced recent decades, as reflected for example by the growing severe reductions in capital spending. In Greece, importance of diagnostic and therapeutic devices or spending in 2016 was still less than half its 2005 level, the expansion of information and communications dropping to about a quarter of the level in 2012 and technology (ICT) in health care (see the indicator on the 2013. In Italy, investment has also continued to drift adoption and use of Electronic Medical Records and downwards since 2010. While capital spending ePrescribing). However, the level of resources invested increased in the United Kingdom in the immediate in buildings, machinery and technology tends to aftermath of the crisis, spending in 2015 and 2016 was fluctuate more than current spending on health still 20-30% down in real terms on 2005 levels. services, often responding to the economic climate In making capital investment decisions, policy whereby investment decisions may be postponed or makers need to carefully assess not only the short- brought forward. term costs, but also the potential benefits in the short, In 2016, it is estimated that the European Union as medium and longer term. Slowing down investment in a whole allocated around 0.6% of its total GDP on health infrastructure and equipment may also reduce capital spending in the health sector (Figure 8.14). This the capacity to treat patients and contribute to increases compares to 9.6% of GDP allocated to current spending in waiting times for different types of services. on health services and medical goods (see the indicator on health expenditure as a share of GDP in Chapter 5). As with current spending, there are differences both in Definition and comparability the current levels of investment between countries and Gross fixed capital formation in the health in recent trends. sector is measured by the total value of the fixed As a proportion of GDP, Germany was the biggest assets that health providers have acquired during spender on capital investment in the health sector in the accounting period (less the value of the 2016 with around 1.1% of its GDP allocated, followed by disposals of assets) and that are used repeatedly a group of countries – Belgium, Malta, Spain and or continuously for more than one year in the Austria – that spent between 0.7-0.85% of their GDP. At production of health services. The breakdown by the lower end, the Czech Republic, Hungary and assets includes infrastructure (e.g. hospitals, Croatia invested less than 0.15% of their GDP on capital clinics, etc.), machinery and equipment infrastructure and equipment in the health sector. (including diagnostic and surgical machinery, By its nature, capital spending fluctuates more ambulances, and ICT equipment), as well as than current spending from year to year in line with software and databases. capital projects on construction (i.e. building of Gross fixed capital formation is reported by hospitals and other health care facilities) and many countries under the System of Health investment programmes on new equipment Accounts. It is also reported under the National (e.g. medical and ICT equipment), but decisions on Accounts broken down by industrial sector capital spending also tend to be more affected by according to the International Standard Industrial economic cycles with spending on health system Classification (ISIC) Rev. 4 using Section Q: Human infrastructure and equipment often being a prime target health and social work activities or Division 86: for reduction or postponement during periods of Human health activities. The former is normally economic uncertainty. While capital spending grew broader than the SHA boundary while the latter is strongly prior to the crisis – overall capital spending in narrower. the EU rose by more than 30% between 2005 and 2008 in real terms – it fell to a level almost 10% below this over the next five years. From 2013 onwards, overall investment has increased again by about 15% and was higher than its pre-crisis levels overall by 2016 Reference (Figure 8.15). A System of Health OECD, Eurostat and WHO (2011), Despite the economic crisis, capital spending Accounts 2011: Revised edition , OECD Publishing, Paris, continued to increase fairly steadily in countries like http://dx.doi.org/10.1787/9789264270985-en . HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 202

205 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY 8.14. Gross fixed capital formation in the health sector as a share of GDP, 2016 (or nearest year) % GDP 1.2 1.1 1.0 0.8 0.8 0.8 0.7 0.7 0.6 0.6 0.6 0.6 0.6 0.5 0.5 0.5 0.5 0.5 0.4 0.4 0.4 0.4 0.4 0.3 0.4 0.3 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0.0 1. Refers to gross fixed capital formation in ISIC 86: Human health activities (ISIC Rev. 4). 2. Refers to gross fixed capital formation in ISIC Q: Human health and social work activities (ISIC Rev. 4). Source: OECD Health Statistics 2018; OECD National Accounts; Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933836979 8.15. Changes in gross fixed capital formation, selected countries, 2005 to 2016 Austria Czech Republic EU25 EU25 France Italy Belgium Greece United Kingdom Sweden Index (2005=100) Index (2005=100) 200 200 180 180 160 160 140 140 120 120 100 100 80 80 60 60 40 40 20 20 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Source: OECD Health Statistics 2018; OECD National Accounts; Eurostat Database. 1 2 http://dx.doi.org/10.1787/888933836998 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 203

206 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY PROJECTIONS OF PUBLIC EXPENDITURE ON HEALTH AND LONG-TERM CARE provision, etc.), can all have a substantial impact on the Despite a dramatic slowdown in spending on health growth in public spending on health care (de la and long-term care in many EU member states following Maisonneuve et al., 2016). the 2008 economic and financial crisis, more recent estimates show that spending is back on an upward path. Since, on average, around three-quarters of health spending is financed out of public sources, this represents Definition and comparability a sizeable share of government spending, meaning that growth in health and long-term care spending can have a Public expenditure on health is defined as the considerable impact on a country’s budgetary position. In “core” health care categories (SHA 1.0 categories addition, ageing populations will continue to exert HC.1 to HC.9), excluding long-term nursing care pressures on health and long-term care spending while at category (HC.3), but including capital investment the same time reducing the size of the working-age in health (HC.R.1). It excludes private expenditure population able to finance such expenditures, thereby in the form of direct out-of-pocket payments by raising concerns around the fiscal sustainability of health households and private health insurance. and long-term care systems (OECD, 2015). Long-term care is defined as a range of services Projections of public expenditure on both health and required by persons with reduced degree of long-term care are regularly carried out by the Ageing functional capacity (physical or cognitive) and Working Group of the Economic Policy Committee (AWG), who are consequently dependent on help with using the European Commission services’ models (EC basic and/or instrumental activities of daily living and EPC, 2017). In both health and long-term care for an extended period of time. Basic Activities of projection models, a range of scenarios tests the potential Daily Living (ADL) or personal care services are impact of different determinants of public spending frequently provided in combination with basic (including both demographic and non-demographic medical services such as nursing care, prevention, factors) to project how each may contribute to the rehabilitation or services of palliative care. evolution of public spending over the next 50 years. Instrumental Activities of Daily Living (IADL) or assistance care services are mostly linked to home The results presented here are based on the help. reference (or baseline) scenario. Among the main assumptions are that half of the future gains in life The data, methodology and assumptions used expectancy are assumed to be spent in good health and for the health and long-term care expenditure that there is a convergence of income elasticity of health projections are explained in detail in the 2017 care spending from 1.1 in 2016 to unity by 2070. The report prepared by the European Commission (DG main outcome of the 2018 projection exercise, based on ECFIN) and the Economic Policy Committee these set of assumptions, is an increase of public (Ageing Working Group). The “AWG reference spending on health of almost one percentage point (0.9) scenario” is used as the baseline scenario when of GDP in total for the 28 EU countries by 2070 calculating the overall budgetary impact of ageing. the lower end of the projections, public (Figure 8.16). At The EU averages are weighted according to GDP. expenditure on health is forecast to rise by only 0.3 of GDP in Bulgaria and Estonia, while it is projected to increase by more than 2 percentage points of GDP in Portugal and Malta (EC and EPC, 2018). References Long-term care expenditure represents an de la Maisonneuve, C. et al. (2016), “The drivers of public increasing share of GDP in many EU countries and as health spending: Integrating policies and institutions”, such is important in the long-term sustainability of OECD Economics Department Working Papers , No. 1283, public finances. Under the same AWG reference OECD Publishing, Paris. scenario of healthy life expectancy gains and converging European Commission (DG ECFIN) and Economic Policy income elasticity, the main result from the baseline Committee (Ageing Working Group) (2017), The 2018 scenario is a projected increase in public spending on Ageing Report: Underlying Assumptions and Projection long-term care across the 28 EU countries by slightly Methodologies , European Economy, No 65/2017, Brussels. more than one percentage point, from 1.6% of GDP in 2016 to 2.7% of GDP in 2070 (Figure 8.17). The results European Commission (DG ECFIN) and Economic Policy vary widely across countries, from only 0.1 percentage Committee (Ageing Working Group) (2018), The 2018 point of GDP in Greece and Bulgaria up to more than Ageing Report: Economic and Budgetary Projections for the ,EuropeanEconomy, 28 EU Member States (2016-2070) 2 percentage points of GDP in Luxembourg, the No 79/2018, Brussels. Netherlands and Denmark (EC and EPC, 2018). Additional OECD studies have shown that different OECD (2015), Fiscal Sustainability of Health Systems: Bridging policy and institutional factors (such as financing Health and Finance Perspectives , OECD Publishing, Paris, https://doi.org/10.1787/9789264233386-en . mechanisms, decentralisation, organisation of health HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 204

207 II.8. RESILIENCE: INNOVATION, EFFICIENCY AND FISCAL SUSTAINABILITY 8.16. Public spending on health care as a percentage of GDP, 2016 to 2070 – Ageing Working Group reference scenario 2016 Change 2016-70 % GDP 10 9 8 7 6 5 4 3 2 1 0 The EU28 total is weighted by GDP. Note: Source: EC and EPC (2018). 1 2 http://dx.doi.org/10.1787/888933837017 8.17. Public spending on long-term care as a percentage of GDP, 2016 to 2070 – Ageing Working Group reference scenario Change 2016-70 2016 % GDP 8 7 6 5 4 3 2 1 0 Note: The EU28 total is weighted by GDP. Source: EC and EPC (2018). 1 2 http://dx.doi.org/10.1787/888933837036 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 205

208

209 Health at a Glance: Europe 2018 State of Health in the EU Cycle © OECD/European Union 2018 Statistical annex Table A.1. Total population, mid-year, thousands, 2000 to 2017 2010 2000 2013 2014 2015 2016 2017 2011 2005 2012 8 737 8 363 8 480 8 546 8 643 8 430 8 773 8 228 8 392 Austria 8 012 10 896 11 038 11 107 11 159 11 209 Belgium 11 331 11 352 10 251 10 479 11 274 7 659 7 348 7 306 7 265 7 224 7 396 7 128 7 102 8 170 Bulgaria 7 178 4 312 4 296 4 283 4 269 4 254 4 236 Croatia 4 172 4 154 4 468 4 208 694 829 851 864 862 853 848 852 855 Cyprus 739 10 546 10 211 10 511 10 514 10 525 10 496 10 566 10 579 10 255 10 474 Czech Republic 5 419 5 548 5 571 5 592 5 615 Denmark 5 683 5 728 5 749 5 340 5 643 1 397 1 331 1 327 1 323 1 318 1 355 1 315 1 316 1 316 Estonia 1 315 5 176 5 246 5 363 5 388 Finland 5 439 5 462 5 480 5 495 5 503 5 414 France 63 001 64 819 65 128 65 439 65 771 66 084 66 593 66 860 66 989 60 762 82 212 82 469 80 275 80 426 80 646 80 983 81 687 82 349 82 522 Germany 81 777 10 987 10 892 11 105 11 045 10 965 10 806 10 821 10 776 10 768 Greece 11 121 9 866 10 211 9 972 9 920 9 893 10 000 9 843 9 814 9 798 Hungary 10 087 3 805 4 160 4 560 4 580 4 600 4 624 4 658 4 702 4 755 4 784 Ireland Italy 56 942 59 277 59 379 59 540 60 234 60 789 60 731 60 627 60 589 57 969 1 994 2 368 2 060 2 034 2 013 2 098 1 978 1 960 1 950 Latvia 2 239 3 500 3 323 3 097 3 028 2 988 Lithuania 2 932 2 905 2 868 2 848 2 958 Luxembourg 465 507 518 531 436 556 570 583 596 543 Malta 460 390 404 415 416 420 426 435 445 455 Netherlands 15 926 16 615 16 693 16 755 16 804 16 865 16 940 17 030 17 082 16 320 38 259 38 165 38 063 38 063 38 040 38 012 37 986 37 970 37 973 Poland 38 043 10 503 10 401 10 558 10 515 10 457 10 290 10 358 10 325 10 310 Portugal 10 573 19 909 22 443 20 148 20 058 19 984 20 247 19 815 19 702 19 644 Romania 21 320 5 389 5 373 5 391 5 398 5 408 5 413 5 419 5 424 5 431 5 435 Slovak Republic Slovenia 1 989 2 049 2 053 2 057 2 060 2 062 2 064 2 065 2 066 2 000 46 481 40 568 46 743 46 773 46 620 46 577 46 445 46 484 46 528 Spain 43 653 8 872 9 030 9 378 9 449 9 519 Sweden 9 696 9 799 9 923 9 995 9 600 United Kingdom 60 401 62 766 63 259 63 700 58 893 64 613 65 129 65 596 65 809 64 128 EU28 (total) 487 822 495 517 503 808 503 519 504 605 506 087 507 660 509 409 510 900 511 528 Albania 3 061 2 913 2 905 2 900 2 895 2 894 2 889 2 881 2 877 3 011 2 067 2 026 2 059 2 061 2 064 2 055 2 070 2 072 2 074 FYR of Macedonia 2 037 281 297 318 319 321 Iceland 327 331 335 338 324 Montenegro 613 619 620 621 605 622 622 622 622 621 Norway 4 491 4 623 4 889 4 953 5 019 5 080 5 137 5 190 5 235 5 258 7 164 Serbia 7 291 7 234 7 199 7 441 7 131 7 095 7 058 7 040 7 516 Switzerland 7 184 7 437 7 825 7 912 7 997 8 089 8 189 8 282 8 373 8 420 77 182 Turkey 68 435 73 142 74 224 75 176 76 148 65 809 78 218 79 278 79 815 Note: Data for 2017 are provisional and subject to revisions. Source: Eurostat Database (data extracted in June 2018). http://dx.doi.org/10.1787/888933837055 1 2 207

210 STATISTICAL ANNEX Table A.2. Share of the population aged 65 and over, January 1st, 1960 to 2017 1970 1980 2000 2010 2011 2012 2013 2014 2015 2016 2017 1960 1990 18.3 14.0 15.4 17.6 17.6 17.8 18.1 14.9 18.5 18.4 18.5 12.1 15.5 Austria 13.3 14.3 14.8 16.8 17.2 17.1 17.4 Belgium 17.8 18.1 18.2 18.5 12.0 17.6 7.4 11.8 13.0 16.2 18.2 18.5 9.4 19.2 19.6 20.0 20.4 20.7 Bulgaria 18.8 .. .. .. .. 16.1 17.8 17.7 Croatia 18.1 18.4 18.8 19.2 19.6 17.9 Cyprus 9.5 10.8 10.8 11.2 12.5 12.7 12.8 13.2 13.9 14.6 15.1 15.6 6.4 16.8 9.5 12.5 13.8 15.3 15.6 16.2 13.6 17.4 17.8 18.3 18.8 Czech Republic 11.9 10.5 12.2 14.3 15.6 14.8 16.3 16.8 Denmark 17.8 18.2 18.6 18.8 19.1 17.3 Estonia 11.7 12.5 11.6 14.9 17.4 10.5 17.7 18.0 18.4 18.8 19.0 19.3 17.4 Finland 7.2 9.0 11.9 13.3 14.8 17.0 17.5 18.1 18.8 19.4 19.9 20.5 20.9 France 11.6 14.0 13.9 16.0 16.8 16.9 17.3 17.7 18.0 18.4 18.8 19.2 12.8 1 10.8 15.6 15.3 16.2 20.7 20.7 13.0 20.8 20.9 21.0 21.1 21.2 Germany 20.7 9.4 11.1 13.1 13.7 17.3 19.0 Greece 19.7 20.1 20.5 20.9 21.3 21.5 19.3 Hungary 11.5 13.5 13.2 15.0 16.6 8.9 16.9 17.2 17.5 17.9 18.3 18.7 16.7 Ireland 11.1 11.1 10.7 11.4 11.2 11.2 11.5 11.9 12.2 12.6 12.9 13.2 13.5 Italy 9.3 13.1 14.7 18.1 20.4 20.5 20.8 21.2 21.4 21.7 22.0 22.3 10.8 18.6 Latvia 13.0 11.8 14.8 18.1 18.4 11.9 18.8 19.1 19.4 19.6 19.9 .. Lithuania .. 10.0 11.3 10.8 13.7 17.3 17.9 18.1 18.2 18.4 18.7 19.0 19.3 13.9 Luxembourg 12.5 13.7 13.4 14.3 14.0 10.8 14.0 14.0 14.1 14.2 14.2 14.2 Malta 18.8 18.5 .. .. 8.4 10.4 11.8 14.9 15.7 16.4 17.1 17.7 18.2 Netherlands 8.9 11.5 12.8 13.6 15.3 15.6 16.2 16.8 17.3 17.8 18.2 18.5 10.1 14.4 5.8 10.0 12.1 13.6 13.6 14.0 10.2 14.9 15.4 16.0 16.5 Poland 8.2 7.8 9.2 11.2 13.2 16.0 18.3 18.7 Portugal 19.4 19.9 20.3 20.7 21.1 19.0 Romania 8.5 10.3 10.3 13.2 16.1 .. 16.1 16.3 16.5 17.0 17.4 17.8 16.1 Slovak Republic 6.8 9.1 10.6 10.3 11.4 12.4 12.6 12.8 13.1 13.5 14.0 14.4 15.0 Slovenia .. 10.9 10.6 13.9 16.5 16.5 16.8 17.1 17.5 17.9 18.4 18.9 .. 17.4 Spain 11.1 13.4 16.5 16.8 17.1 9.5 17.7 18.1 18.5 18.7 19.0 8.2 Sweden 11.7 13.6 16.2 17.8 17.3 18.1 18.5 18.8 19.1 19.4 19.6 19.8 19.8 16.4 United Kingdom 12.9 14.9 15.7 15.8 16.3 11.7 16.8 17.2 17.5 17.7 17.9 18.1 9.8 EU28 (total) 11.3 13.1 13.7 15.8 17.5 17.6 17.9 18.2 18.5 18.9 19.2 19.4 Albania .. .. .. 8.1 .. .. .. .. 12.0 12.4 12.9 13.1 .. 12.0 .. 8.4 9.8 11.6 11.7 11.8 .. 12.4 12.7 13.0 13.3 FYR of Macedonia .. 8.0 8.8 9.8 10.6 11.6 12.0 12.3 Iceland 12.9 13.2 13.5 13.9 14.0 12.6 Montenegro .. .. .. 12.2 12.9 .. 12.9 13.1 13.3 13.7 14.1 14.4 12.8 Norway 10.9 12.8 14.7 16.3 15.3 14.9 15.1 15.4 15.7 15.9 16.1 16.4 16.6 17.3 Serbia .. .. 16.0 17.0 17.2 .. 17.6 18.0 18.5 19.0 19.4 .. Switzerland 10.2 11.2 13.8 14.6 15.3 16.8 16.9 17.2 17.4 17.6 17.8 18.0 18.1 7.5 Turkey 4.4 4.7 4.3 5.4 7.0 7.2 7.3 3.5 7.7 8.0 8.2 8.3 | Break in series. 1. Population figures for Germany prior to 1991 refer to West Germany. Source: Eurostat Database (data extracted in June 2018). 1 2 http://dx.doi.org/10.1787/888933837074 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 208

211 STATISTICAL ANNEX Table A.3. Crude birth rate, per 1 000 population, 1960 to 2016 1970 1980 2000 2010 2011 2012 2013 2014 2015 2016 1960 1990 9.4 15.0 9.8 9.4 9.3 9.4 11.8 9.6 9.8 10.0 17.9 12.0 Austria 14.7 12.6 12.4 11.4 11.9 11.7 Belgium 11.3 11.2 10.8 10.8 16.8 11.5 17.8 14.5 12.1 9.0 10.2 9.6 16.3 9.2 9.4 9.2 9.1 Bulgaria 9.5 18.4 13.8 14.8 11.6 9.8 10.1 9.6 Croatia 9.4 9.3 8.9 9.0 9.8 Cyprus 19.2 20.4 18.3 12.2 11.8 11.3 11.8 10.8 10.9 10.8 11.1 26.2 10.3 13.4 12.6 8.9 11.2 10.4 14.9 10.2 10.4 10.5 10.7 Czech Republic 15.0 16.6 14.4 11.2 12.3 12.6 11.4 Denmark 10.4 10.0 10.1 10.2 10.8 10.6 Estonia 15.8 15.0 14.2 9.4 11.9 16.7 10.6 10.3 10.3 10.6 10.7 11.1 Finland 18.5 14.0 13.2 13.1 11.0 11.4 11.1 11.0 10.7 10.5 10.1 9.6 France 17.9 14.9 13.4 13.1 12.8 12.5 12.4 12.4 12.4 12.0 11.7 16.7 1 17.4 10.1 11.5 9.3 8.3 8.3 13.3 8.5 8.8 9.0 9.6 Germany 8.4 18.9 16.5 15.4 10.0 9.6 10.3 Greece 9.1 8.6 8.5 8.5 8.6 9.6 Hungary 14.7 13.9 12.1 9.6 9.0 14.7 9.1 9.0 9.5 9.4 9.7 8.8 Ireland 21.5 21.8 21.7 15.1 14.4 16.5 16.2 15.6 14.9 14.4 13.9 13.4 Italy 18.1 11.3 10.0 9.5 9.5 9.2 9.0 8.5 8.3 8.0 7.8 16.7 9.1 Latvia 14.1 14.2 8.6 9.4 14.6 9.8 10.2 10.9 11.1 11.2 16.7 Lithuania 22.5 17.7 15.2 15.4 9.8 9.9 10.0 10.2 10.1 10.4 10.8 10.7 11.6 Luxembourg 13.0 11.4 12.9 13.1 16.0 10.9 11.3 11.3 10.9 10.7 10.4 9.8 26.2 Malta 9.7 17.6 17.7 15.2 11.3 9.4 10.0 9.8 9.5 9.6 Netherlands 20.8 12.8 13.2 13.0 11.1 10.8 10.5 10.2 10.4 10.1 10.1 18.3 10.1 22.6 14.4 9.9 10.9 10.2 19.6 9.7 9.9 9.7 10.1 Poland 16.8 24.1 20.8 16.2 11.7 11.7 9.6 Portugal 8.5 7.9 7.9 8.3 8.4 9.2 Romania 21.1 17.9 13.6 10.4 10.5 19.1 10.0 9.4 10.0 10.0 10.2 9.7 Slovak Republic 21.7 17.8 19.1 15.1 10.2 11.2 11.3 10.3 10.1 10.2 10.3 10.6 Slovenia 17.6 15.7 11.2 9.1 10.9 10.7 10.7 10.2 10.3 10.0 9.9 15.9 10.1 Spain 15.2 10.3 9.8 10.4 19.5 9.7 9.1 9.2 9.0 8.8 21.7 Sweden 13.7 13.7 11.7 14.5 10.2 12.3 11.8 11.9 11.8 11.9 11.7 11.8 12.8 United Kingdom 16.2 13.4 13.9 11.5 12.9 17.5 12.8 12.1 12.0 11.9 11.8 16.3 EU28 (total) 18.5 14.0 12.4 10.6 10.7 10.5 10.4 10.0 10.1 10.0 10.1 Albania 43.3 26.5 25.1 16.7 .. 11.2 .. 12.3 12.4 11.3 11.0 32.5 11.4 31.7 18.8 14.5 11.8 11.1 21.0 11.2 11.4 11.1 11.1 FYR of Macedonia 23.2 28.0 19.7 19.8 18.7 15.3 15.4 Iceland 14.1 13.4 13.4 12.5 12.0 14.1 Montenegro .. .. .. 15.2 12.0 .. 12.0 12.0 12.1 11.9 12.2 11.6 Norway 17.3 16.7 12.5 14.4 13.2 12.6 12.2 12.0 11.6 11.5 11.3 11.3 9.1 Serbia .. .. 9.8 9.4 .. 9.3 9.2 9.3 9.3 9.2 .. Switzerland 17.7 16.1 11.7 12.5 10.9 10.3 10.2 10.3 10.2 10.4 10.5 10.5 16.8 Turkey .. .. .. 21.1 16.9 16.7 17.0 .. 17.3 16.9 16.5 | Break in series. 1. Population figures for Germany prior to 1991 refer to West Germany. Note: Crude birth rate is defined as the number of live births per 1 000 population. Source: Eurostat Database (data extracted in June 2018). 1 2 http://dx.doi.org/10.1787/888933837093 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 209

212 STATISTICAL ANNEX Table A.4. Total fertility rate, number of children per women aged 15-49, 1960 to 2016 1960 1990 2000 2010 2011 2012 2013 2014 2015 2016 1970 1980 1.44 2.29 1.36 1.44 1.43 1.44 1.46 1.46 1.49 1.53 2.69 1.65 Austria 2.25 1.68 1.62 1.67 1.86 1.81 Belgium 1.76 1.74 1.70 1.68 2.54 1.80 2.31 2.05 1.82 1.26 1.57 1.51 2.17 1.48 1.53 1.53 1.54 Bulgaria 1.50 2.20 1.80 1.90 1.70 1.46 1.55 Croatia 1.51 1.46 1.46 1.40 1.42 1.48 Cyprus .. 2.48 2.41 1.64 1.44 1.35 1.39 1.30 1.31 1.32 1.37 .. 1.45 2.09 1.90 1.15 1.51 1.43 2.08 1.46 1.53 1.57 1.63 Czech Republic 1.92 2.57 1.95 1.55 1.67 1.77 1.87 Denmark 1.73 1.67 1.69 1.71 1.79 1.75 Estonia 2.17 2.02 2.05 1.36 1.72 1.98 1.56 1.52 1.54 1.58 1.60 1.61 Finland 2.72 1.83 1.63 1.78 1.73 1.87 1.83 1.80 1.75 1.71 1.65 1.57 France 2.73 1.95 1.78 1.87 2.02 2.00 1.99 1.99 2.01 1.96 1.92 2.47 1 2.37 1.56 1.45 1.38 1.39 1.39 2.03 1.42 1.47 1.50 1.60 Germany 1.41 2.23 2.40 2.23 1.39 1.25 1.48 Greece 1.34 1.29 1.30 1.33 1.38 1.40 Hungary 1.98 1.91 1.87 1.32 1.25 2.02 1.34 1.35 1.44 1.45 1.53 1.23 Ireland 3.78 3.85 3.21 2.11 1.89 2.05 2.03 1.98 1.93 1.89 1.85 1.81 Italy 2.37 1.64 1.33 1.26 1.46 1.44 1.43 1.39 1.37 1.35 1.34 2.38 1.33 Latvia 1.90 2.01 1.25 1.36 2.02 1.44 1.52 1.65 1.70 1.74 1.94 Lithuania .. 2.40 1.99 2.03 1.39 1.50 1.55 1.60 1.59 1.63 1.70 1.69 1.63 Luxembourg 1.97 1.50 1.60 1.76 2.29 1.52 1.57 1.55 1.50 1.47 1.41 Malta 1.37 1.37 .. .. 1.99 2.04 1.68 1.36 1.45 1.42 1.36 1.38 Netherlands 3.12 1.60 1.62 1.72 1.79 1.76 1.72 1.68 1.71 1.66 1.66 2.57 1.33 2.98 2.06 1.37 1.41 1.33 2.28 1.29 1.32 1.32 1.39 Poland 2.20 3.16 3.01 2.25 1.56 1.55 1.39 Portugal 1.28 1.21 1.23 1.31 1.36 1.35 Romania .. 2.43 1.83 1.31 1.59 .. 1.52 1.46 1.56 1.58 1.64 1.47 Slovak Republic 3.04 2.41 2.32 2.09 1.30 1.43 1.45 1.34 1.34 1.37 1.40 1.48 Slovenia 2.18 1.93 1.46 1.26 1.57 1.56 1.58 1.55 1.58 1.57 1.58 2.21 1.34 Spain 2.22 1.36 1.22 1.37 2.90 1.32 1.27 1.32 1.33 1.34 2.86 Sweden 2.20 1.92 1.68 2.13 1.54 1.98 1.90 1.91 1.89 1.88 1.85 1.85 1.91 United Kingdom 2.43 1.90 1.83 1.64 1.92 2.72 1.92 1.83 1.81 1.80 1.79 2.61 EU28 (total) 2.35 1.92 1.66 1.46 1.62 1.59 1.59 1.55 1.58 1.57 1.60 Albania .. .. .. 1.79 .. .. .. 1.73 1.73 .. 1.54 .. 1.51 .. 2.23 1.88 1.56 1.46 .. 1.49 1.52 1.50 1.50 FYR of Macedonia .. 4.26 2.81 2.48 2.30 2.08 2.20 Iceland 2.04 1.93 1.93 1.80 1.74 2.02 Montenegro .. .. .. .. 1.70 .. 1.72 1.73 1.75 1.74 .. 1.65 Norway 2.94 2.50 1.72 1.93 1.85 1.95 1.88 1.85 1.78 1.75 1.72 1.71 1.40 Serbia .. .. 1.48 1.40 .. 1.45 1.43 1.46 1.46 1.46 .. Switzerland 2.44 2.10 1.55 1.58 1.50 1.52 1.52 1.52 1.52 1.54 1.54 1.54 2.09 Turkey 5.00 4.63 3.07 2.27 2.04 2.03 6.40 2.08 2.17 2.14 2.11 | Break in series. 1. Population figures for Germany prior to 1991 refer to West Germany. Source: Eurostat Database (data extracted in June 2018). http://dx.doi.org/10.1787/888933837112 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 210

213 STATISTICAL ANNEX Table A.5. GDP per capita in 2017 and average annual growth rates, 2009 to 2017 GDP per capita Annual growth rate per capita in real terms in EUR PPP 2010/11 2009/10 2013/14 2014/15 2015/16 2016/17 2011/12 2017 2012/13 1.6 0.2 -0.6 0.0 0.0 0.4 2.5 38 222 Austria 2.6 Belgium 0.5 -0.4 -0.3 0.8 0.8 0.9 1.5 34 860 1.8 14 733 2.6 0.6 1.4 2.0 4.3 4.7 3.9 Bulgaria 1.9 18 250 -1.2 0.0 -1.9 Croatia 0.3 3.0 4.0 3.2 -0.3 Cyprus -1.3 -2.2 -4.5 -5.7 -0.3 2.6 2.9 3.5 25 192 26 454 1.6 -0.9 -0.5 2.6 5.1 2.4 4.2 Czech Republic 2.0 37 496 1.4 0.9 -0.1 0.5 Denmark 0.9 1.2 1.9 1.1 Estonia 2.5 7.9 4.7 23 085 3.2 1.6 2.0 4.9 2.3 Finland 32 681 2.5 2.1 -1.9 -1.2 -1.0 -0.2 1.8 2.5 France 31 172 1.6 -0.3 0.1 0.5 0.3 0.8 1.6 1.5 36 910 3.7 0.3 0.2 1.5 0.9 1.1 2.0 Germany 4.3 20 062 -5.6 -9.0 -6.8 -2.5 Greece 0.4 0.2 1.4 1.4 Hungary 0.9 2.0 -1.1 20 410 4.5 3.6 2.5 4.2 2.4 Ireland 55 360 1.2 2.5 -0.4 1.1 7.5 24.4 4.0 7.1 Italy 28 660 0.4 -3.1 -2.9 -0.8 1.0 1.0 1.6 1.4 Latvia -1.9 8.3 5.3 3.5 2.8 3.8 3.1 5.1 19 978 Lithuania 23 178 3.8 8.5 5.2 4.6 4.4 3.0 3.7 4.6 Luxembourg 84 402 3.2 0.7 -2.5 1.7 4.0 1.4 1.4 0.5 Malta 5.4 3.1 29 023 3.0 0.9 1.8 3.2 6.0 7.3 Netherlands 38 304 1.2 -1.4 -0.5 1.1 1.8 1.7 2.9 0.9 21 112 5.0 1.6 1.5 3.4 3.9 3.0 4.6 Poland 3.9 22 999 1.9 -1.7 -3.6 -0.6 Portugal 2.2 1.9 2.8 1.4 Romania -2.2 2.5 1.7 18 740 3.5 4.5 5.4 7.3 3.9 Slovak Republic 23 016 4.9 2.7 1.5 1.4 2.7 3.8 3.2 3.3 Slovenia 25 372 0.4 -2.9 -1.3 2.9 2.2 3.1 5.0 0.8 Spain -0.4 -1.3 -3.0 -1.4 1.7 3.5 3.2 3.0 27 664 Sweden 36 821 5.1 1.9 -1.0 0.4 1.6 3.4 1.9 1.7 1.4 United Kingdom 0.9 0.7 0.8 31 574 2.3 1.5 1.2 1.5 29 964 EU28 (total) 1.8 1.5 -0.6 0.0 1.5 2.0 1.7 2.5 Albania 8 283 2.8 1.6 1.2 1.8 2.4 3.6 .. 4.2 11 045 2.2 -0.6 2.8 3.5 3.7 2.8 0.0 FYR of Macedonia 3.1 38 997 -3.5 1.7 0.8 3.3 Iceland 3.2 6.0 2.8 1.1 Montenegro 4.7 3.1 -2.8 12 743 1.7 3.3 2.9 .. 3.4 Norway 44 874 1.6 4.2 2.2 -0.7 -1.3 -3.6 -3.5 4.1 Serbia 1.0 2.2 -0.5 3.1 -1.4 1.3 3.3 2.1 10 948 Switzerland 47 305 1.9 0.6 -0.1 0.7 1.2 0.1 0.3 0.5 4.7 Turkey 6.9 9.5 3.5 7.1 3.8 19 503 1.8 6.7 Note: EU28 displays a weighted average and is calculated based on total GDP divided by the total population of the 28 EU member states. Source: Eurostat Database; OECD National Accounts Database. http://dx.doi.org/10.1787/888933837131 1 2 HEALTH AT A GLANCE: EUROPE 2018 © OECD/EUROPEAN UNION 2018 211

214

215 OECD PUBLISHING, 2, rue André-Pascal, 75775 PARIS CEDEX 16 (81 2018 70 1 P) ISBN 978-92-64-30334-8 – 2018

216 Health at a Glance: Europe 2018 STATE OF HEALTH IN THE EU CYCLE Health at a Glance: Europe 2018 presents comparative analyses of the health status of EU citizens and the performance of the health systems of the 28 EU Member States, 5 candidate countries and 3 EFTA countries. Health at a Glance: cycle of knowledge brokering. This publication has two parts. State of Health in the EU It is the fi rst step in the Part I comprises two thematic chapters, the fi rst focusing on the need for concerted efforts to promote better mental health, the second outlining possible strategies for reducing wasteful spending in health. In Part II, the Europe 2018 most recent trends in key indicators of health status, risk factors and health spending are presented, together with a discussion of progress in improving the effectiveness, accessibility and resilience of European health STATE OF HEALTH IN THE EU CYCLE systems. Health at a Glance: Europe 2018 STATE OF HEALTH IN THE EU CYCLE Consult this publication on line at . https://doi.org/10.1787/health_glance_eur-2018-en This work is published on the OECD iLibrary, which gathers all OECD books, periodicals and statistical databases. Visit www.oecd-ilibrary.org for more information. ISBN 978-92-64-30334-8 81 2018 70 1 P &RIXQGHGE\ WKH+HDOWK3URJUDPPH 9HSTCQE*daddei+ RIWKH(XURSHDQ8QLRQ

Related documents