BETTER EVIDENCE FOR BETTER MENTAL HEALTH POLICY Francisca Vargas Lopes Mind the inequalities gap

Better Evidence for Better Mental Health Policy: Mind the inequalities gap Francisca Vargas Lopes

COLOFON Doctoral thesis, Erasmus University Rotterdam Francisca Vargas Lopes Copyright © 2024 January Francisca Vargas Lopes All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior permission of the author or the copyright-owning journals for previously published chapters. ISBN 978-94-6473-404-1 Printed by Ipskamp Printing | Layout and design by Indah Hijmans, Cover illustration by Diana Reis, This thesis was printed with financial support of the Department of Public Health, Erasmus Medical Center and of the Erasmus University Rotterdam.

CONTENTS Chapter 1 General Introduction 7 Chapter 2 Income inequalities beyond access to mental health care: a Dutch nationwide record-linkage cohort study of baseline disease severity, treatment intensity, and mental health outcomes 23 Chapter 3 Patient cost-sharing, mental health care and inequalities: A population-based natural experiment at the transition to adulthood 67 Chapter 4 The effects of supported housing for individuals with mental disorders 101 Chapter 5 Antidepressant therapy prescription, do psychologists help? Evidence from Portugal 153 Chapter 6 Natural experiments: A Nobel Prize awarded research design for strengthening causal inference on global health challenges 189 Chapter 7 General Discussion 197 Appendices Summary 226 Samenvatting 229 About the author 233 List of Publications 234 Portfolio 235 Acknowledgments 239

CHAPTER 1 General Introduction

8 Chapter 1 THE HIGH BURDEN OF MENTAL DISORDERS AND ITS DISTRIBUTION Mental disorders pose an extremely high burden to society, both in terms of their direct impact on health but also in terms of societal welfare losses [1]. According to the most recent estimates of the Global Burden of Disease (GBD) study, mental disorders rank as the second cause of years lived with disability (YLDs) worldwide and the seventh cause of disability-adjusted life years (DALYs; the sum of YLDs and years of life lost (YLLs)) [2]. The main contributors to this burden are depression (37% of DALYs) and anxiety (23%) due to their high prevalence, followed by schizophrenia (12%) due to extremely disabling states of acute psychosis. Importantly, recent studies point towards a systematic underestimation of mental health burden mainly through unaccounted YLLs. According to these estimates, the actual burden should be at least two to three times higher than the conventional results, making mental disorders responsible for 13% to 16% of all DALYs worldwide [3, 4]. This population health burden translates into substantial economic consequences. Mental disorders are costly not only due to their direct medical costs, such as outpatient care and hospitalisations but mainly due to indirect costs, such as losses in labour productivity and income resulting from absenteeism and presentism [3]. Disability resulting from mental health disorders reduces countries´ economic output by undermining labour and capital supply and contributes to the persistence of the vicious poverty and illness cycle. A recent estimate of the economic value associated with mental disorders suggests global losses of 4.7 trillion US dollars (USD) in 2019, ranging from 4% of the gross domestic product (GDP) in Eastern Sub-Saharan Africa to 8% in North America [3]. Mental health is a global public good and a vital determinant of every country’s sustainable development, regardless of its level of development. In fact, in what concerns addressing population mental health, all countries qualify as developing countries [1]. Within countries, the distribution of mental disorders burden varies substantially by age, gender, and socioeconomic status. The number of DALYs increases consistently during childhood and adolescence, peaks for the age group of 25 to 34 years, and decreases steadily from there onwards [2]. Mental disorders are the leading cause of disability among adolescents and young adults in high-income countries [5, 6]. Approximately 75% of adult mental disorders have their onset during adolescence, increasing the risk of recurrence and disabling physical1 conditions in adulthood [6, 7]. Gender patterns are also marked, at least for some conditions. Females have the highest prevalence and burden of depressive, anxiety and eating disorders. Males have the highest prevalence 1 For the lack of a better term I use physical conditions / diseases / disorders to refer to all the other disorders that are not mental or neurological. The reader should, however, be aware of the limitations of using such terminology. Referring to physical or somatic disorders as the opposite of mental disorders perpetuates the idea that the later are not equally attributable to our phenotype; supporting the segregation of mental disorders that has made these neglected for decades.

9 General Introduction of autism spectrum disorders and attention deficit hyperactivity disorder (ADHD). Bipolar disorders and schizophrenia vary the least with gender [2]. Last, and most important for this thesis, mental disorders are known to be strongly concentrated among those of low socioeconomic status, measured either at the individual level (income, education, socioeconomic position, employment [8-12]) or through area-based measures (neighbourhood socioeconomic conditions, social capital, built environment. [13, 14]). The quality of the evidence varies considerably in the literature, often reporting associations and less frequently causation. Inequalities driven by socioeconomic status have been mostly documented for common mental disorders, namely depression [13, 15, 16]. There is also evidence of disparities in suicide and suicidal behaviour [13, 14, 17-19]. The evidence of socioeconomic disparities for the severe conditions such as schizophrenia and bipolar disorders is the least consistent, particularly because some early findings were not robust to methodological progress on measurement and analysis [20-23]. Shocks negatively impacting socioeconomic status such as substantial reductions in household income or becoming unemployed have also been shown to increase the risk of mental disorders, especially mood disorders [24]. Life shocks seem responsible for a part of existing mental health disparities: exposure to financial hardship and adverse life events such as separation from spouse, personal injury or jail resulted in a greater risk of mental disorders for the most disadvantaged socioeconomic groups [25]. Another part of adulthood disparities will follow from early stages in life. Children and adolescents at socioeconomic disadvantage are two to three times more likely to develop mental disorders. Among several indicators of socioeconomic status, low parental education and household income substantially impact children and youth mental health [26]. LOW AND UNEQUAL ACCESS TO CARE The substantial treatment gap between those needing mental treatment and those receiving it is a major challenge to mental health policy. This gap persists worldwide, including in high-income countries where service uptake has increased since the 1990s. Estimates from 2001-2012 show that only 37% of all individuals with anxiety, mood, and substance use disorders had received treatment in the past year [27]. Worryingly, the literature suggests that countries with a reduction of the treatment gap in the last decades did not achieve the expected decline in the prevalence of common mental disorders. Among several hypotheses explaining this limited impact is a potential “quality gap”, per which the treatment provided would not meet the minimal standards of clinical practice guidelines or be optimally targeted to those in greatest need [28]. The treatment gap is unequally distributed within countries, being the largest for population groups at socioeconomic disadvantage. The literature on disparities in access 1

10 Chapter 1 to care shows that the magnitude and direction of the inequalities vary depending on the socioeconomic measure used (e.g. education, income), the type of care studied (e.g. specialist vs. non-specialist, but also psychiatrists vs. psychologists), and across countries with different levels of coverage and publicly funded services. Studies from high-income countries as the Netherlands, Denmark, Australia and Canada suggest that low-income individuals are more likely to access mental health support provided by human services such as social workers or counsellors in non-speciality settings, general practitioners or public-funded psychiatric services [29-33]. At the same time, studies report higher income individuals as more likely to access care from psychologists both in settings with (Canada and Denmark) and without (Germany) out-of-pocket payments for these services [29, 31, 34]. High-income individuals were also more likely to access psychiatric care in countries with limited health care coverage, like the United States of America [32]. Another relevant and related stream of literature is the one that reports cost-sharing as a financial barrier to access care, even in the most egalitarian countries such as the Netherlands [35-37]. Besides income, education also strongly influences access to mental health care. Evidence from the Netherlands and Canada consistently displayed highly educated individuals accessing psychiatric and psychologist services more often than low-educated ones [38, 39]. While inequalities in access to mental health care remain a challenge in most counties, efforts to promote equitable access to treatment have lately gained space within policy agendas. Much less attention is paid to what happens once individuals start treatment, and the potential “quality gap”. The few studies that examined individual-level disparities in the amount of treatment received suggest that, conditional on accessing to care, low socioeconomic status is associated with a reduced rate of visits [31, 40]. Literature is even more scarce in what concerns inequalities in outcomes. Existing observational studies are small and present mixed findings on the associations between area-based measures of deprivation and treatment outcomes [41, 42]. More robust evidence is only available from experimental settings, with differences by socioeconomic status being observed in randomised clinical trials (RCTs) and their follow-up data. These studies have shown that both pharmacotherapy and psychotherapy approaches had higher magnitude effects for individuals with depression who had higher education and income or those who were employed and homeowners [43, 44]. Importantly, inequalities identified in RCTs settings are likely to be even larger in the real world, where treatment adherence and continuation are lower, and influences of doctor/patient preferences and beliefs are not accounted for through randomisation. Describing disparities in mental care delivery and outcomes is, therefore, a priority to inform policy-makers on closing the mental health gap. The lack of evidence might be partially caused by the challenge of simultaneously observing the need for and the utilisation of mental health services and socioeconomic status. Appropriately assessing patient need is essential in the context of the horizontal equity principle equal treatment

11 General Introduction of equals, which argues for “equal treatment for equal medical need, irrespective of other characteristics such as income, race, place of residence, etc.” [45]. However, the need for (mental health) treatment is a complex dimension to capture. Data on clinical diagnoses are essential because diagnosis guides most of the treatment provision but should be complemented with information on symptom severity, distress, functioning and disability, among others. This information is usually only measured in cohorts/surveys of limited sample sizes, which do not include detailed information about treatment practices. On the other hand, electronic health records providing larger samples and rich treatment information will usually miss data to measure need and characterise socioeconomic status comprehensively. MENTAL HEALTH AND INEQUALITIES: TWO CONTEMPORARY POLICY CHALLENGES From a policy perspective, mental health has been underrepresented in most health (and public) policy agendas for decades mainly when compared to physical health. With some of the first pivotal declarations and reports dating from the 1990s and early 2000s, the frequency of calls for policy action started increasing around the 2010s [1]. These calls advocate for an entire transformation of the global mental health agenda: from the basis of recognising mental health as a fundamental human right [46, 47] to the need to act on children´s and youth’s mental health [48, 49], and to transform mental health systems by shifting care towards the community and increasing service coverage to reduce the treatment gap [50]. Nevertheless, it was just during the COVID-19 pandemic that mental health started to get full attention from governments and societies, although still heterogeneously across countries. Several grand challenges persist in transforming this attention into action at both government and societal levels. These challenges are interconnected and include the stigma faced by people with experience of mental health conditions, limited mental health system leadership and governance, and the lack of an integrated multi-sectoral approach requiring cross-ministerial commitment and funding to integrate mental health policy across education, labour and welfare systems [51-53]. Mental health inequalities are also yet to receive the appropriate attention in policy agendas, as most discussion has been devoted to disparities in general health measures. This discussion has been partially supported by research findings comparing inequalities between countries. International comparisons of mortality and morbidity by educational level and income have shown substantial differences and unexpected country rankings, and well as unpredictable trends. Examples of these findings include the fact that countries with more developed welfare such as the Nordic countries do not have smaller inequalities [54]. Or the contrast between countries that have made substantial progress in mortality inequalities by educational level (e.g. Finland, England and Wales, France) and countries that experienced considerable setbacks from very small to substantial disparities, in the 1

12 Chapter 1 same period (Norway and Hungary, 1970 to 2010) [55]. International differences in the equity of health care financing and delivery also has been attracting policy attention since the early 1990s [56]. By then, studies showed that higher-income individuals used specialist doctor appointments substantially more than low-income groups, accounting for these groups differences in need for such services. The inequities were observed in most European countries regardless of their different health care systems. In contrast, the likelihood of using primary care (seeing a GP) was more equitably distributed according to need [57]. Some countries even had a pro-poor distribution on the number of GP appointments, conditional on having at least one visit [56]. Despite some country commitment to reducing health care inequities findings from more than a decade later concluded that the inequity persisted through time, except for a few exceptions [58]. Progress has been equally slow towards reducing health inequalities, even though the ambition to do so has been shared by countries throughout the last decades. One of the ambitions set internationally was “Closing the gap in a generation”, an initiative led by the World Health Organization (WHO) and anchored in the work of the Commission on Social Determinants of Health [59]. The initiative included several countries and partners aiming to shape policy and programs on social determinants of health to improve health equity. Unfortunately, the progress towards closing the gap in a generation has been minimal. Several reasons can be discussed as being behind this lack of progress. First, the widening of inequalities in social determinants of health, particularly those linked to socioeconomic conditions such as income and occupation [54]. A set of crises has recently fuelled the increase in income inequality. Among these crises is the COVID-19 pandemic, which has threatened to reverse part of the last years´ achievements in reducing disparities [60]. Second, how progress has been measured and whether the focus should be on relative (e.g., rate ratios of mortality) or absolute (e.g. rate differences in mortality) changes in inequalities [61]. Progress is admittedly more difficult to achieve when measured relatively in the presence of downward trends, as was the case of mortality and morbidity in Europe pre-pandemic [54]. While such an argument favours looking at absolute inequalities, methodological concerns exist for this approach: particularly in the case of steadily decreasing outcomes, an “arithmetic maturation process” could automatically lead to a decrease in absolute inequalities that does not represent actual progress [61]. Third, and perhaps most important, deliberate efforts towards closing the health gap have been too few and of too small magnitude to deal with the challenge [54]. Most countries have never implemented a strategy primarily aimed at reducing health inequalities. In those that did – like the almost unique case of England - the population-level effects were limited [54]. International comparisons suggest that country differences in mortality inequalities seem to be greatly driven by inequalities in lifestyle behaviours (smoking and excessive alcohol consumption) and poverty. Country-level contextual factors such as

13 General Introduction economic and political conditions or welfare policies would mostly matter under extreme conditions or crises [62]. Therefore, policies promoting the welfare state seem to have limited effectiveness in addressing health-related behaviours, and the inequalities driven by those [54]. Appropriate and effective action looks particularly difficult to achieve because it should follow the motto health-in-all-policies and target the wide social determinants of health with precision and magnitude. Policy efforts have to go beyond the scope of health policy and be coordinated in a whole-government approach. Awareness around the impact of health in the domains of education, employment or social protection has increased with the recent pandemic crisis, and both the topics of health inequalities and mental health gained attention in the cross-ministry policy agendas. To convert this attention to evidence-based decision-making, it remains crucial to continue expanding our knowledge about the groups that are most vulnerable to (mental) health disparities and the pathways running behind these. We must also better evaluate policy distributional effects and learn what policies work towards closing the gap. RESEARCH AND EVIDENCE TO GUIDE POLICY AND PRACTICE The previous sections discuss several gaps requiring additional evidence to inform the strategy for closing the mental health gap. Well-known moral, socioeconomic and political arguments support research use to inform decision- and policy-making. Using the best evidence can enhance the effective use of public resources, increase transparency and accountability, and improve public policy’s effectiveness, efficiency and equity [63]. Why is this not regularly done, even when the proper evidence is in place? Like with closing the health gap, conducting evidence-based policy-making seems a bigger of a challenge than what its advocates like to admit. Everyone recognises its importance, but the commitment of the several parts is not enough to move it forward. One potential reason behind the difficulty of implementing evidence-based policymaking is the type, breadth and quantity of evidence needed. Analysing the evidence creation funnel (Figure 1.) is helpful to understand that primary research, which is usually the main output of academic work for peer-reviewed publications, is only one part of the evidence needed. In fact, primary research has limited potential to respond to questions on impact, effectiveness and/or cost-effectiveness. Secondary research, which synthesises the results of primary research studies in a larger evidence body, is most suitable to provide those answers. There is still tertiary research, which aims at adapting primary and secondary research to the needs of the evidence users and is the least valued in the academic research context [63]. All these three parts of the funnel consist of scientific evidence, which is produced through formal, systematic and rigorous processes. 1

14 Chapter 1 In their definition of evidence-informed2 decision-making (EIDM), the WHO argues that scientific evidence is insufficient to address the context in which decisions are made, which is inherently local, political and shaped by institutional constraints and individual interests and preferences [63]. Tacit or colloquial evidence should also play a role. This type of knowledge is “mostly informal, and often includes opinions, values and habits of policy-makers, clinicians, patients or citizens expressed in different forms in formal deliberative dialogues, on websites, in policy documents, reports, and other formats” [63]. In the context of evidence-informed practices, scientific and tacit evidence would have a complementary rather than competitive relationship, in which the latter would complement and enquire about the appropriateness of the former to the (local) contexts in which decisions are made [63]. Figure 1. Evidence creation funnel. Source: Evidence, policy, impact. WHO guide for evidenceinformed decision-making [63]. 2 Recently, WHO has put emphasis on the use of the term evidence-informed over evidence-based policy-making. This accounts for the fact that evidence is often only one of several factors influencing policy-making processes. It also uses decision instead of policy as the former is broader and encompasses all the types of decision-making for which evidence might be relevant.

15 General Introduction A second challenge is related to the causal nature of primary research which is particularly desirable to inform decision-making. Contrary to what happens in precursor movements of evidence-based medicine and health technology assessment, most exposures of interest in health and public policy cannot be studied in RCTs or experiments. This is undoubtedly the case of interventions, programmes and policies targeting social determinants of health, in which promoting randomisation would often be unethical (poverty, unemployment), unfeasible or very costly (neighbourhood environment) or impossible (ethnicity, place of birth). Methodological alternatives to evaluate causal effects of these exposures are therefore extremely important for evidence-based policy. These alternatives are called quasi-experiments3 and are observational studies in which, per opposition to randomisation, the investigator does not control the assignment of individuals to the exposure. Instead, the exposure is somehow – either by “accident of chance” or by design – exogenously attributed to certain (groups of) units and not to others. Conditional on some assumptions, the intervention and control groups resulting from the exogenous assignment can be compared to elicit causal effects. How strong these assumptions are depends on the mechanism behind the assignment. For instance, lotteries or weather shocks are examples of “accidents of chance” producing interesting variations to study cause and effect of getting wealthy or migrating, respectively. In this case, the treatment group is indeed “as if” randomly assigned, as individuals cannot manipulate or select themselves into winning the lottery or suffering from a hurricane. There are other cases in which the assignment is not random, and the possibility of mimicking randomisation comes with additional study design features that require some assumptions. This will often be the case of quasi-experiments trying to leverage program or policy roll-out processes that imply local/staggered implementation which creates opportunities for comparing treated and non-treated units. Quasi-experiments represent a great promise for evidence-based policy on health inequalities, in which most exposures of interest are non-randomisable. Furthermore, most quasi-experimental designs can accommodate either stratified analysis by socioeconomic status or interaction terms between socioeconomic status and exposure. This allows the quantification of distributional effects, examining policy impacts on inequalities [64]. Producing quasi-experimental evidence was one of the objectives of this thesis, to strengthen the scarce body of causal evidence about the distributional effects of mental health policy. We were interested in evaluating exposures that might reduce or (unintendedly) widen the mental health gap. Several interesting interventions and policies could not be studied due to the lack of exogenous variation. Others had some potential but were not pursued due to challenges to the utilization of quasi-experimental 3 Quasi-experiments are often also designated as natural experiments, although some authors will recognise differences between these terms. 1

16 Chapter 1 designs. These challenges hinder researchers’4 widespread use of quasi-experiments and also a larger uptake of these studies by policy-makers. Addressing them within research and policy environments is crucial to quasi-experiments finally being recognised and adequately positioned as an essential primary research tool in the context of the evidence creation funnel and EIDM. A third reason goes beyond the comprehensiveness and quality of the evidence needed and recognises that simply the existence of evidence will “virtually never automatically drive tangible policy and practice change” [63]. This finding is driven by the researchto-policy gap that persists between sterile research environments and the context in which decisions are made, subjected to political cycles and power interests, budgetary limitations and institutional habits. The research-to-policy gap results from a lack of alignment between the research process and the policy/action cycle, and differences between researchers and policy-makers in their characteristics, beliefs and preferences. Lack of institutional capacity and resources to promote adequate knowledge translation make this gap wider. Knowledge translation can be defined as “the exchange, synthesis, and effective communication of reliable and relevant research results”. It focuses on promoting interaction among the producers and the users of research, removing the barriers to research use, and tailoring information to different target audiences [65]. Various models of knowledge translation may rely more on the researcher side (push efforts), the policy-maker side (user-pull efforts), or both (exchange efforts). In its most advanced form, knowledge translation requires integrated efforts encompassing all the previous models, promoted by dedicated platforms created with this purpose [66]. In sum, achieving societal impact through research goes beyond the interdisciplinary approach, the appropriateness of the study design, the robustness of the methods, and the data quality. While all these are needed – and were explicitly pursued in the chapters of this thesis – an extra effort of dissemination and knowledge translation is also key in bringing the findings to those who may use them in their decision-making. Such efforts were, therefore, also part of the activities pursued during this PhD. AIM AND OUTLINE OF THE THESIS This thesis aimed to produce evidence to guide mental health policy in reducing inequalities in mental health care and outcomes. This objective was achieved by focusing on two gaps that persist within the existing knowledge. First, knowing what inequalities are in place, particularly those beyond access to care. Second, examining the effects of 4 The extent to which quasi-experiments are used and the research approaches to their use differs considerably between disciplines, with microeconomics having a longer tradition and proficiency in applying them than most health sciences.

17 General Introduction interventions potentially impacting such inequalities. Therefore, this thesis responds to two research objectives: 1. Characterise inequalities in providing care to people with mental health conditions. 2. Identify causal effects of mental health related interventions, programs and policies with impact on the most vulnerable patient groups, using quasiexperimental methods. Chapters 2 to 5 in this thesis answer policy-relevant questions at the intersection between mental health and inequalities. Methodologically, the chapters differ in the evidence they provide, ranging from purely descriptive findings (Chapter 2) to effects obtained from robust quasi-experimental studies (Chapters 3 and 4) but also associations that account for potential endogeneity from time-invariant unobservables (Chapter 5). Chapter 2 describes income inequalities along the mental health treatment pathway in the Netherlands, providing pivotal findings to move policy agendas beyond their usual focus on access to care. Chapter 3 focuses on a crucial timing of individuals’ life course with regards to their mental health treatment: the transition to adulthood. It uses a difference-in-discontinuity design to study whether a 180 euro increase in the Dutch annual deductible was a financial barrier to 18-year-olds accessing mental health care, and whether such an effect differed by youngsters’ household income. Chapter 4 consists of an instrumental variable analysis to study the effects of supported housing, a programme targeted to a highly vulnerable group, both psychosocially and in socioeconomic terms. The study uses a leniency design based on the Dutch institutional feature that centralises all applications in assessors with discretionary power. It studies the effects of supported housing eligibility on health, care, employment and income of the individual and also the parents. Changing country, Chapter 5 examines the relationship between low supply of psychotherapy and high prescription of antidepressants in Portugal. Using a withinbetween random effects model it provides empirical evidence to inform a long-lasting policy discussion on the lack of psychological resources available within the Portuguese National Health Service. Last, Chapter 6 consists of an opinion piece written together with a fellow junior colleague, as we both shared PhD struggles of conducting quasi-experimental studies to inform and evaluate public policy. It reflects on the potential of quasi-experiments in the context of public health research and describes the most common challenges – and potential solutions – to implementing them. Additional considerations about the challenges and opportunities of bringing (interdisciplinary) research and policy together are outlined in the thesis Discussion, in Chapter 7. Beyond the usual conclusions about the PhD work published, the Discussion chapter provides the candidate’s reflections on several PhD experiences that were not translatable into publications. This includes the experience of bridging public health and health economics in an interdisciplinary 1

18 Chapter 1 trajectory, and the lessons taken from efforts to conduct policy-focused research with societal impact: communication of research findings, co-creation of projects with policymakers, answering to policy-making needs in times of crisis, and a traineeship in an international policy organisation.

19 General Introduction REFERENCES 1. Patel V, Saxena S, Lund C, et al. The Lancet Commission on global mental health and sustainable development. Lancet. Oct 27 2018;392(10157):1553-1598. 2. Collaborators GBDMD. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry. Feb 2022;9(2):137-150. 3. Arias D, Saxena S, Verguet S. Quantifying the global burden of mental disorders and their economic value. EClinicalMedicine. Dec 2022;54:101675. 4. Vigo D, Thornicroft G, Atun R. Estimating the true global burden of mental illness. Lancet Psychiatry. Feb 2016;3(2):171-8. 5. Gore FM, Bloem PJ, Patton GC, et al. Global burden of disease in young people aged 10-24 years: a systematic analysis. Lancet. Jun 18 2011;377(9783):2093-102. 6. Erskine HE, Moffitt TE, Copeland WE, et al. A heavy burden on young minds: the global burden of mental and substance use disorders in children and youth. Psychol Med. 2015;45(7):1551-1563. doi:10.1017/ s0033291714002888 7. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. Jun 2005;62(6):593-602. 8. Muntaner C, Eaton WW, Miech R, O’campo P. Socioeconomic position and major mental disorders. Epidemiologic reviews. 2004;26(1):53-62. 9. Patel V, Lund C, Hatherill S, et al. Mental disorders: equity and social determinants. Equity, social determinants and public health programmes. 2010;115:134. 10. Stewart-Brown S, Samaraweera PC, Taggart F, Kandala N-B, Stranges S. Socioeconomic gradients and mental health: implications for public health. The British Journal of Psychiatry. 2015;206(6):461-465. 11. Yu Y, Williams DR. Socioeconomic status and mental health. Handbook of the sociology of mental health. Springer; 1999:151-166. 12. Sareen J, Afifi TO, McMillan KA, Asmundson GJG. Relationship Between Household Income and Mental Disorders: Findings From a Population-Based Longitudinal Study. Archives of General Psychiatry. 2011;68(4):419-427. doi:10.1001/archgenpsychiatry.2011.15 13. M S, A L, G C. Social determinants of mental health: A review of the evidence. 2016;30:259-292. 14. Rehkopf DH, Buka SL. The association between suicide and the socio-economic characteristics of geographical areas: a systematic review. Psychol Med. Feb 2006;36(2):145-57. 15. Fryers T, Melzer D, Jenkins R. Social inequalities and the common mental disorders. Social Psychiatry and Psychiatric Epidemiology. 2003/05/01 2003;38(5):229-237. doi:10.1007/s00127-003-0627-2 16. Lorant V, Deliège D, Eaton W, Robert A, Philippot P, Ansseau M. Socioeconomic Inequalities in Depression: A Meta-Analysis. American Journal of Epidemiology. 2003;157(2):98-112. doi:10.1093/aje/kwf182 17. Huang X, Ribeiro JD, Musacchio KM, Franklin JC. Demographics as predictors of suicidal thoughts and behaviors: A meta-analysis. PLoS One. 2017;12(7):e0180793. 18. Milner A, Spittal MJ, Pirkis J, LaMontagne AD. Suicide by occupation: systematic review and meta-analysis. Br J Psychiatry. Dec 2013;203(6):409-16. 19. Ying YH, Chang K. A study of suicide and socioeconomic factors. Suicide Life Threat Behav. Apr 2009;39(2):214-26. 20. Cai J, Wei Z, Chen M, et al. Socioeconomic status, individual behaviors and risk for mental disorders: A Mendelian randomization study. Eur Psychiatry. Apr 18 2022;65(1):e28. 21. Eid L, Heim K, Doucette S, McCloskey S, Duffy A, Grof P. Bipolar disorder and socioeconomic status: what is the nature of this relationship? Int J Bipolar Disord. 2013;1:9. 22. Yanan L, Lei Z, Ping H, Lihua P, Chao G, Xiaoying Z. Individual-level and area-level socioeconomic status (SES) and schizophrenia: cross-sectional analyses using the evidence from 1.9 million Chinese adults. BMJ Open. 2019;9(9):e026532. doi:10.1136/bmjopen-2018-026532 23. Werner S, Malaspina D, Rabinowitz J. Socioeconomic status at birth is associated with risk of schizophrenia: population-based multilevel study. Schizophr Bull. Nov 2007;33(6):1373-8. 1

20 Chapter 1 24. Barbaglia MG, Have Mt, Dorsselaer S, Alonso J, de Graaf R. Negative socioeconomic changes and mental disorders: a longitudinal study. Journal of Epidemiology and Community Health. 2015;69(1):55-62. doi:10.1136/jech-2014-204184 25. Hashmi R, Alam K, Gow J. Socioeconomic inequalities in mental health in Australia: Explaining life shock exposure. Health Policy. Jan 2020;124(1):97-105. 26. Reiss F. Socioeconomic inequalities and mental health problems in children and adolescents: a systematic review. Soc Sci Med. Aug 2013;90:24-31. 27. Evans-Lacko S, Aguilar-Gaxiola S, Al-Hamzawi A, et al. Socio-economic variations in the mental health treatment gap for people with anxiety, mood, and substance use disorders: results from the WHO World Mental Health (WMH) surveys. Psychol Med. Jul 2018;48(9):1560-1571. 28. Jorm AF, Patten SB, Brugha TS, Mojtabai R. Has increased provision of treatment reduced the prevalence of common mental disorders? Review of the evidence from four countries. World Psychiatry. Feb 2017;16(1):9099. 29. Bartram M, Stewart JM. Income-based inequities in access to psychotherapy and other mental health services in Canada and Australia. Health Policy. Jan 2019;123(1):45-50. 30. Bartram M. Income-based inequities in access to mental health services in Canada. Can J Public Health. Aug 2019;110(4):395-403. 31. Packness A, Waldorff FB, Christensen RD, et al. Impact of socioeconomic position and distance on mental health care utilization: a nationwide Danish follow-up study. Soc Psychiatry Psychiatr Epidemiol. Nov 2017;52(11):1405-1413. 32. Alegría M, Bijl RV, Lin E, Walters EE, Kessler RC. Income Differences in Persons Seeking Outpatient Treatment for Mental Disorders: A Comparison of the United States With Ontario and the Netherlands. Archives of General Psychiatry. 2000;57(4):383-391. doi:10.1001/archpsyc.57.4.383 33. Meadows GN, Enticott JC, Inder B, Russell GM, Gurr R. Better access to mental health care and the failure of the Medicare principle of universality. Medical Journal of Australia. 2015;202(4):190-194. 34. Epping J, Muschik D, Geyer S. Social inequalities in the utilization of outpatient psychotherapy: analyses of registry data from German statutory health insurance. International Journal for Equity in Health. 2017;16(1):18. 35. Lopes FV, Herl CJR, Mackenbach JP, Van Ourti T. Patient cost-sharing, mental health care and inequalities: A population-based natural experiment at the transition to adulthood. Social Science & Medicine. 2022;296:114741. 36. Ravesteijn B, Schachar EB, Beekman AT, Janssen RT, Jeurissen PP. Association of cost sharing with mental health care use, involuntary commitment, and acute care. JAMA psychiatry. 2017;74(9):932-939. 37. Lambregts TR, van Vliet R. The impact of copayments on mental healthcare utilization: a natural experiment. Eur J Health Econ. Jul 2018;19(6):775-784. 38. Have M, Oldehinkel A, Vollebergh W, Ormel J. Does educational background explain inequalities in care service use for mental health problems in the Dutch general population? Acta Psychiatr Scand. Mar 2003;107(3):178-87. 39. Leah S. Steele CSDEL, Kenneth LKL. Education Level, Income Level and Mental Health Services Use in Canada: Associations and Policy Implications. Healthcare Policy. 2007;3(1):96-106. 40. Steele LS, Glazier RH, Lin E. Inequity in mental health care under Canadian universal health coverage. Psychiatric services. 2006;57(3):317-324. 41. Delgadillo J, Dawson A, Gilbody S, Böhnke JR. Impact of long-term medical conditions on the outcomes of psychological therapy for depression and anxiety. Br J Psychiatry. Jan 2017;210(1):47-53. 42. Poots AJ, Green SA, Honeybourne E, et al. Improving mental health outcomes: achieving equity through quality improvement. Int J Qual Health Care. Apr 2014;26(2):198-204. 43. Mills JA, Suresh V, Chang L, et al. Socioeconomic Predictors of Treatment Outcomes Among Adults With Major Depressive Disorder. Psychiatr Serv. Sep 1 2022;73(9):965-969. 44. Falconnier L. Socioeconomic status in the treatment of depression. Am J Orthopsychiatry. Apr 2009;79(2):148-58. 45. Doorslaer E, Wagstaff A. Measuring and Testing for Inequity in the Delivery of Heath Care. Journal of Human Resources. 2000;35:716-733. doi:10.2307/146369 46. Organization WH. The World Health Report 1995 - Bridging the gaps. 1995.

21 General Introduction 47. (WHO) P-AHOPWHO. Caracas Declaration. presented at: Conference on the restructuring of psychiatric care in Latin America within the Local Health Systems; 1990; Venezuela. 48. Patel V, Flisher AJ, Hetrick S, McGorry P. Mental health of young people: a global public-health challenge. Lancet. Apr 14 2007;369(9569):1302-1313. 49. Kieling C, Baker-Henningham H, Belfer M, et al. Child and adolescent mental health worldwide: evidence for action. Lancet. Oct 22 2011;378(9801):1515-25. 50. Lancet Global Mental Health G, Chisholm D, Flisher AJ, et al. Scale up services for mental disorders: a call for action. Lancet. Oct 6 2007;370(9594):1241-52. 51. Oecd. Fitter Minds, Fitter Jobs. 2021. 52. Oecd. A New Benchmark for Mental Health Systems. 2021. 53. Thornicroft G, Sunkel C, Alikhon Aliev A, et al. The Lancet Commission on ending stigma and discrimination in mental health. Lancet. Oct 22 2022;400(10361):1438-1480. 54. Mackenbach JP. Re-thinking health inequalities. Eur J Public Health. Aug 1 2020;30(4):615. 55. de Gelder R, Menvielle G, Costa G, et al. Long-term trends of inequalities in mortality in 6 European countries. Int J Public Health. Jan 2017;62(1):127-141. 56. Doorslaer Ev, Jones AM. Income-related inequality in health and health care in the European Union. Health Economics. 2004;13(7):605-608. doi: 57. Van Doorslaer E, Wagstaff A, Van der Burg H, et al. Equity in the delivery of health care in Europe and the US. Journal of health economics. 2000;19(5):553-583. 58. Devaux M. Income-related inequalities and inequities in health care services utilisation in 18 selected OECD countries. The European Journal of Health Economics. 2015/01/01 2015;16(1):21-33. doi:10.1007/s10198013-0546-4 59. CSDH. Closing the gap in a generation: health equity through action on the social determinants of health. Final Report of the Commission on Social Determinants of Health. 2008. 60. Wouterse B, Geisler J, Bär M, van Doorslaer E. Has COVID-19 increased inequality in mortality by income in the Netherlands? J Epidemiol Community Health. Apr 2023;77(4):244-251. 61. Mackenbach JP, Martikainen P, Menvielle G, de Gelder R. The arithmetic of reducing relative and absolute inequalities in health: a theoretical analysis illustrated with European mortality data. J Epidemiol Community Health. Jul 2016;70(7):730-6. 62. Mackenbach JP. Health Inequalities in Europe: New Insights from Comparative Studies. Erasmus University Medical Center; 2016. 63. Evidence, policy, impact. WHO guide for evidence-informed decision-making. 2021. 64. Hu Y, van Lenthe FJ, Hoffmann R, van Hedel K, Mackenbach JP. Assessing the impact of natural policy experiments on socioeconomic inequalities in health: how to apply commonly used quantitative analytical methods? BMC Medical Research Methodology. 2017/04/20 2017;17(1):68. doi:10.1186/s12874-017-0317-5 65. World Health O. World report on knowledge for better health : strengthening health systems. Geneva: World Health Organization; 2004. 66. Lavis JN, Lomas J, Hamid M, Sewankambo NK. Assessing country-level efforts to link research to action. Bull World Health Organ. Aug 2006;84(8):620-8. 1

CHAPTER 2 Income inequalities beyond access to mental health care: a Dutch nationwide record-linkage cohort study of baseline disease severity, treatment intensity, and mental health outcomes Francisca Vargas Lopes, Bastian Ravesteijn, Tom Van Ourti, Carlos Riumallo-Herl Lancet Psychiatry. 2023 Aug;10(8):588-597

24 Chapter 2 A BSTRACT Background Existing literature shows low and unequal access to mental health treatment globally, resulting in policy efforts to promote access for vulnerable groups. Yet, there is little evidence about how inequalities develop once individuals start treatment. The greater use of mental health care among individuals with low income, such as in the Dutch system, might be driven by differences in need and might not necessarily lead to better treatment outcomes. In this study, we aimed to examine income inequalities in four stages of the mental health treatment pathway while adjusting for need. Methods We constructed a nationwide retrospective cohort study, examining all patients aged older than 18 years with a first specialist mental health treatment record in the Netherlands between 2011 and 2016, excluding those who did not receive any treatment minutes. We linked patient-level data from treatment records to administrative data on income, demographics from municipal registries, and health insurance claims. We used multivariate models to estimate adjusted associations between household income quintile (standardised for household size) and outcomes characterising four stages of mental health treatment: severity at baseline assessment based on the Global Assessment of Functioning (GAF) score, treatment minutes received, functional improvement by the end of the initial record, and additional treatment in a subsequent record. Estimates were adjusted for patient need (97 categories of primary diagnosis and severity at baseline assessment measured by GAF) and demographic covariates. Findings Our study population consisted of 951 530 adults with a first specialist mental health treatment record in the Netherlands between Jan 1, 2011, and Dec 31, 2016. Patients in our cohort were on average aged 45.0 years (range 19–107) and mostly female (529 859 [55.7%] women and 421 671 [44.3%] men; no ethnicity data were available). First, we found that patients with the lowest income had the greatest initial therapist-assessed disease severity (5.545 GAF points), which was 0.353 GAF points (95% CI 0.347–0.360) lower than those in the highest income quintile. Second, we found that the negative association between income and treatment minutes was reversed once we adjusted for diagnosis and severity at baseline, with patients with the lowest income receiving 1.8% fewer treatment minutes (95% CI 1.1–2.4) than those in the highest quintile. Third, those in the highest income quintile were 17.3 percentage points (pp) (95% CI 17.0–17.6) more likely to have functional improvements by the end of the initial record, compared with 25.8% of patients with an improvement in the lowest income quintile. Fourth, while 35.7% of patients in the lowest income quintile received additional treatment in a subsequent record, this was only 3.0 pp (95% CI 2.7–3.3) lower for those in the highest quintile. None of these patterns were explained by diagnosis, severity at baseline, or treatment minutes received.

25 Income inequalities beyond access to mental health care Interpretation Disparities favourable to patients with a higher income persist through the different stages of mental health treatment. These differences highlight the limitations of solely focusing on improving access to care to reduce the mental health gap. Our findings call for a better understanding of the role of social environment and quality of care as complementary mechanisms explaining inequalities during mental health treatment. 2

26 Chapter 2 INTRODUCTION Even though the need for mental health treatment is often concentrated among individuals with low incomes [1-5], people with higher incomes generally use specialist mental health care more, often because financial barriers prevent low-income people from accessing such care [6-9]. Advocates of mental health care reform worldwide have looked to the Netherlands, which has universal and comprehensive mental health care coverage and lower out-of-pocket costs, as a model of ensuring access to adequate mental health treatment for all [10-12]. However, no previous work has evaluated whether the mental health outcomes of such a system are equitable across the income distribution. The current literature shows that there is low and unequal access to mental health care. However, the direction of the disparities varies across countries, measures of socioeconomic status, and types of services [6-9, 13]. Studies conducted in Australia, Canada, Denmark, and the Netherlands found that low-income people access mental health treatment more often than higher income groups through social or primary healthcare workers. In contrast, access to psychologist services was more frequent in those with a high income, both when copayments applied [7] and when financial barriers were low. [8, 14] Access to psychiatric care does, however, depend on coverage. Comprehensive coverage, in countries such as Denmark, was associated with a higher use of specialist care by low-income individuals than in other countries [7], but in countries with low coverage, such as the USA, care was concentrated among high-income individuals [9]. When using education as a measure of socioeconomic status, studies consistently reported individuals with higher educational attainment to have greater access to most types of care than individuals with lower educational levels [15, 16]. Similarly, a multicountry study revealed a modest positive association between education and access to specialist care, but a non- monotonic association between income and treatment rates, with lower specialist treatment rates for middle-income respondents compared with both high- income and low-income respondents [17]. Most of the current literature is focused on access to mental health treatment and there is less evidence available on inequalities among patients in mental health care. This absence of evidence is because of the difficulty in disentangling need for and the use of mental health care across the income distribution during different treatment stages. Most existing studies were limited in the outcomes they observed, such as access or number of visits, relied on samples not larger than several thousands of patients, or could only control for crude measures of need. The few studies that evaluated the amount of care received concluded that individuals of a high socioeconomic status had a higher rate of contact with services [7, 18]. In a cross-country comparison, only a few individuals received minimally adequate treatment overall, with a considerable gradient in favour of high- income countries [19].

27 Income inequalities beyond access to mental health care Although obtaining access to appropriate care is a prerequisite to reduce mental health inequalities, it is probably insufficient. Some evidence based on randomised controlled trials showed differences in treatment effectiveness by socioeconomic status. A meta- analysis combining individual patient data found that improvement after any treatment was positively associated with employment and home ownership [20]. Notably, associations of improvement with financial strain and lower education disappeared when controlling for clinical prognosis factors [20]. Relatedly, data from randomised controlled trials of several pharmacotherapies and psychotherapies for depression showed stronger positive effects among individuals with higher education and income [21, 22]. These findings from experimental settings highlight the need for population-level evidence on inequalities beyond access to care, and particularly on disparities in mental health outcomes. In this study, we aimed to examine income inequalities in four stages of mental health treatment, using nationwide data for all first specialist treatment records in the Netherlands. METHODS Study design We performed a retrospective cohort study examining adults living in the Netherlands who started specialist mental health treatment between 2011 and 2016. In the Netherlands, specialist mental health care typically follows a referral from a primary care physician to a specialist care institution where the main practitioner is responsible for defining a patient’s primary diagnosis and completing the treatment record. The treatment is tailored to each patient’s case and could include care provided by clinical psychologists, psychiatrists, or multidisciplinary teams. Specialist mental health care is covered by a standard benefit package in a setting of mandated universal health insurance. Patients with low income receive government subsidies to reduce the premium costs paid to health insurers. Health insurers pay care providers on the basis of treatment records, which contain standardised information on the services provided. A single treatment record consists of all the care received for up to 1 year after its opening. Treatment continuation after 364 days or starting treatment with a different provider or diagnosis requires a new record. If applicable, a proportion of the provider cost paid by the insurer is billed to the patient as a deductible: in each year patients pay out-of-pocket for any health care received, excluding primary care but including mental health services, until they reach the deductible threshold. The total annual compulsory deductible ranged from €170 (US$237) in 2011 to €385 (US$426) in 2016 [23]. Other out-of-pocket payments and respective changes throughout the study period are described in the Appendix A. 2