Mental health prevalence, healthcare use and access between 2018 and 2022 in Sri Lanka: an analysis of survey data

Article history

The contractual start date for this research was in April 2018. This article began editorial review in March 2023 and was accepted for publication in October 2023. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Global Health Research editors and publisher have tried to ensure the accuracy of the author’s article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright statement

Copyright © 2024 Deivanayagam et al. This work was produced by Deivanayagam et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2024 Deivanayagam et al.

The challenges Sri Lankans have faced and impact on mental health

Sri Lanka has suffered four decades of violent conflict, a tsunami, terrorist attacks and an economic crisis, with unknown mental health consequences. 1 Some stability over the following decade was shaken by the Easter Sunday bombings of churches and luxury hotels in 2019 by Islamic State of Iraq and Syria (ISIS) terrorists, killing more than 250 people, which led to increased hostility and outbreaks of violence against Muslims. 2 In 2020, Sri Lanka was hit by the coronavirus disease 2019 (COVID-19) pandemic, with more than 600,000 confirmed cases and over 16,000 official deaths in Sri Lanka. 3 The adoption of large tax cuts in 2020 combined with COVID-19 control measures, which included closing borders, lockdowns and school closures, had substantial negative economic impacts and starved social sectors of fiscal resources, resulting in substantial job losses and disrupted social services, and contributed to the worst economic crisis since the 1930s, which from late 2021 led to a scarcity of fuel, food and medicines. 4 The COVID-19 Delta wave in mid-2021 also resulted in over 3000 infections and 150 deaths per day in August 2021 and overwhelmed the hospital system for several months. Hospitals in the country in 2022 reported shortages of essential medicines and other health items, while frequent power outages affect the delivery of health services. 5 The United Nations Office for Coordinated Humanitarian Affairs estimated in April 2022 that 5.7 million people needed humanitarian assistance including food, fuel, cooking gas, essential supplies and medicines in 25 districts across the country. 6 In June 2022, the World Food Programme estimated that 3 in 10 households (6.2 million people) were food insecure, and food inflation was alarmingly high at 57.4%. 7

The mental health impacts of conflict, natural disasters and economic crises have been documented in Sri Lanka and globally. The prevalence of mental health disorders in conflict-affected populations is substantially higher than in the average population: 17.3% versus 5.3% for depressive disorders. 8,9 People living in conflict-affected settings are at increased risk of developing depressive, anxiety and psychotic disorders. 10 The negative mental health consequences of the 2004 Indian Ocean tsunami have been investigated particularly amongst mothers and adolescents in Sri Lanka. 11,12 A 2021 study found that the current crisis has led to high levels of distress among young people. 13 The current economic crisis and accompanying political uncertainties, widespread protests and social disruptions have the potential to adversely affect mental health, though this has not been studied extensively. Worsening public mental health in the context of these multiple crises can additionally strain health systems. 14

Regarding previous prevalence estimates, in 2008, the World Health Organization (WHO) estimated that around 3% of the Sri Lankan population experienced some form of mental ill health. 15 According to 2015 WHO estimates, the prevalence of depressive disorder in Sri Lanka was around 4.1% and anxiety disorder 3.4%. 13 Sri Lanka also has a high suicide mortality rate of 14 per 100,000 in 2019, compared to a global average of 9.2 per 100,000. 16 Higher levels of mental distress are an important risk factor for suicide. 16,17

Barriers to healthcare use and access

People living with mental health difficulties may experience individual, interpersonal, social or structural barriers to help-seeking. The literature consistently highlights that, culturally, social stigma associated with mental health is common in Sri Lanka, whereby negative attitudes lead to people being marginalised in society. 1,13 Patients, carers and families face subsequent social and economic exclusion due to mental health difficulties, depriving them of participating in socioeconomic processes such as work and marriage. 16 Stigma experienced by patients and carers is associated with delays in help-seeking. 17 Lack of knowledge about mental health as a result of these social norms means individuals and carers may not recognise difficulties and in turn may not seek help. If they recognise difficulties, they may avoid seeking help due to stigma, denial or fear of societal repercussions. Even if people recognise that they need support, there is another layer of barriers to obtaining care. Some of these barriers in the context of mental health include language,18 acceptability/appropriateness of care,19 lack of family support,20 availability of medication,21 lack of transport22 and financial barriers. 21 It is possible that several of these factors may have manifested during this challenging time period, for example, the high price of fuel and the lack of availability of medicines, impacting on healthcare use and access. Worldwide systematic reviews also highlight that the rate of healthcare use declined following the COVID-19 pandemic23 and economic downturn. 24

Research questions

The project aims to answer the following questions:

1. How did the prevalence of (common) mental health difficulties amongst adults in Sri Lanka change between 2018–9 (Wave 1) and 2022 (Wave 2)?

2. How do the rates of healthcare use among those with high mental health difficulties compare to those without high mental health difficulties at Waves 1 and 2?

3. How do the rates of healthcare access (unmet need) among those with high mental health difficulties compare to those without high mental health difficulties in Waves 1 and 2?

4. Are the differences in rates of healthcare use and access between those with differing mental health difficulties dependent upon sociodemographic factors?

Sample

The SLHAS is a national, longitudinal, cohort study managed by a consortium of the Institute for Health Policy, the University of Colombo, University of Peradeniya, University of Ruhuna and the University of Rajarata, approved by the Ministry of Health (MOH), Sri Lanka. The project uses data from the SLHAS, an ongoing nationally representative study of adults (aged 18 years and older) living in Sri Lanka. Six thousand six hundred and sixty-eight adults participated in Wave 1, the majority of which (4548) also completed the follow-up in Wave 2. This study uses an analytic sample of 4030 respondents who participated in the healthcare use module in both Wave 1 and Wave 2 and held information for at least one mental health measure in both surveys.

SLHAS Wave 1 was conducted from November 2018 to November 2019. Detailed information on the recruitment and sampling strategy can be found as part of recent work using data collected in the SLHAS. 25 Recruited participants were invited to a field clinic near their residence to complete in-person interviews. In April 2019, during Wave 1 of data collection, Sri Lanka was impacted by terrorist attacks, which caused a 6-week disruption of field work leading to non-coverage of several primary sampling units (PSUs), a fall in Muslim response rates, and the inability to survey one predominantly Muslim PSU owing to security conditions. The affected PSU was replaced with a substitute PSU from the same stratum matching by district, sector and level of socioeconomic development. However, it was not possible to match the ethnic profile of the original PSU.

SLHAS Wave 2 data collection took place between August 2021 and August 2022. Participants completed interviews by telephone, with information being collected on COVID history, vaccination, mental health, healthcare use and access, household well-being and public opinion.

Across both waves, participants provided information about themselves and their households.

Measures

Data analysis

A series of descriptive analyses were conducted to provide an overview of the demographic and socioeconomic characteristics of the cohort, and to quantify the prevalence of mental health difficulties within the cohort at each wave, stratified by socioeconomic subgroups. Subsequent regression analyses examined the changes in prevalence of mental health difficulties over time, again stratified by economic subgroups. Cross-sectional descriptive analyses examined rates of healthcare use and access among those experiencing mental health difficulties in Waves 1 and 2, and regression analyses compared use and access for those experiencing ‘mild’ and ‘high’ mental health difficulties with those experiencing ‘minimal’ mental health difficulties, adjusting for confounding variables in four stages: unadjusted; adjusted for age and sex; further adjustment for education and socioeconomic status; and further adjustment for ethnicity, religion, region and language.

The cohort was weighted to be representative of the population of Sri Lanka, accounting for sampling design and differential non-response within the second wave of data collection.

All analyses were conducted in STATA^® Release 17 (StataCorp LLC, College Station, TX, USA). 30

Demographic characteristics

The sociodemographic characteristics of the cohort are presented in Table 1. The cohort comprised of 4030 participants with a mean age of 49.1 years [standard error (SE): 0.25], 51.2% of which were female. The majority of participants were Buddhist (73.8%), and ethnically Sinhalese (78.3).

Mental health prevalence

The prevalence of high common mental health difficulties in Wave 1 was 2.9% (n = 115) and 6.1% (n = 247) in Wave 2. During Wave 1, 25.1% of the cohort experienced mild mental health difficulties, and 72.0% experienced minimal difficulties. During Wave 2, 41.5% experienced mild mental health difficulties, while 52.4% experienced minimal difficulties. Prevalence estimates of common mental health difficulties differed by age, sex, ethnicity, SES, education, region and religion, and are presented in Table 1.

Change in prevalence of mental health difficulties over time

Table 2 displays multivariable regression analyses comparing prevalence of mental health difficulties between waves. The levels of high mental health difficulties changed significantly between the two waves [β = 0.23, 95% confidence interval (CI) 0.20 to 0.26; p < 0.05], highlighting the increased prevalence of common mental health distress in Wave 2 compared to in Wave 1. This change in prevalence was greater for women, older people, people in lower SES quintiles, people with below O-level education, non-Buddhist people and non-Sinhala people.

Healthcare use and access

Figure 2 displays the overlap of high mental health difficulties and the use of outpatient and inpatient medical services during Waves 1 and 2. Among the people experiencing high common mental health difficulties, rates of outpatient use in the previous 12 months decreased by 3.1% between waves, from 35.4% to 32.3% in Waves 1 and 2, respectively (see Table 3). However, rates of inpatient use in the previous 12 months increased by 7.7% between waves (Wave 1: 16.0%; Wave 2: 23.7%).

FIGURE 2.

Inpatient and outpatient use of medical care for Sri Lankans reporting ‘moderate to severe’ mental health difficulties in SLHAS 2018–9 and 2021–2.

When adjusted for confounders, those experiencing high mental health difficulties were significantly more likely to access outpatient services in both Wave 1 [adjusted odds ratio (aOR) = 2.15, 95% CI 1.27 to 3.63] and Wave 2 (aOR = 1.96, 95% CI 1.34 to 2.85) than those with minimal difficulties (see Table 4). Similarly, those experiencing high mental health difficulties were significantly more likely to use inpatient services than those with minimal difficulties in both waves (Wave 1: aOR = 2.03, 95% CI 1.10 to 3.76; Wave 2: aOR = 2.76, 95% CI 1.80 to 4.22).

In addition, while 27.7% of the cohort experiencing high mental health difficulties did not access medical services when required in the 12 months prior to Wave 1, they were not statistically significantly less likely to access services than those with minimal mental health difficulties (aOR = 2.53, 95% CI 0.71 to 9.06). In the month prior to Wave 2 responses, 8.6% of participants experiencing high mental health difficulties did not access needed medical services, again not significantly more than those with minimal mental health difficulties (aOR = 1.02, 95% CI 0.71 to 1.48).

Primary findings and concordance with previous research

Prevalence of high mental health difficulties rose significantly over the study period. This rise was not equal across different sociodemographic groups, with a disproportionate increase in mental health difficulties amongst older adults, those in the lowest socioeconomic classes and those not from minority ethnic groups. In addition, the gap between inpatient use for those living with the highest and lowest levels of mental health difficulties widened across the study period, though the outpatient use gap remained stable. Finally, while those with high mental health difficulties reported high levels of unmet need for health care, when compared to those with the lowest levels of mental health difficulty, the difference was not found to be significant, likely due to a restricted sample size.

These results are largely concordant with previous research. Prevalence of mental health difficulties increased throughout the study period, mirroring a qualitative rapid needs assessment in Sri Lanka, which found reduced mental health and psychosocial well-being at the household level since the start of the economic crisis. 31 Globally, studies consistently report a worsening of population mental health following periods of public health crises such as the COVID-19 pandemic;32 economic disruption;33 and exposure to ethnically and politically driven riots and protests. 34 In addition, the overall decrease in use of healthcare services estimated in this study, particularly in terms of outpatient services, is consistent with findings from worldwide systematic reviews focused on rates of healthcare utilisation following the COVID-19 pandemic23 and economic downturn. 24

Strengths

A key strength of this study is that it utilises a large, nationally representative sample. To our knowledge, this study is the first population prevalence estimate for common mental health difficulties in Sri Lanka across the adult population. Moreover, this study provides longitudinal prevalence estimates, highlighting changes in prevalence between two critical periods for Sri Lankans pre- and post-COVID-19. Furthermore, surveys were completed at critical stages in times when the population was facing significant social, political and economic unrest. Therefore, the surveys have captured mental health difficulties during a particularly challenging period of time for Sri Lankans, when they have experienced significant barriers to meeting their basic needs such as access to food and fuel in 2022.

Limitations

This study explores common mental health difficulties measured by the PHQ and DASS – depression, stress and anxiety; however, it does not provide insight into more complex or less common mental health disorders. Additionally, it was not possible to capture whether people with mental health difficulties used or accessed mental health care, distinct from health care more broadly. Although data on self-reported reasons for using medical care were collected in Wave 1, the reported percentage of visits was too low to include in the analysis, possibly due to systematic under-reporting. Similarly, the differences in questions asked between Wave 1 and Wave 2, and the proportion of people asked, limit conclusions that can be drawn around healthcare use and access. As such, only the prevalence estimates are longitudinal, while the estimates of use and access to medical care are cross-sectional, and therefore could not be compared between time periods.

It is unknown whether negative attitudes towards mental health may have led to underestimation of mental distress in this study. As noted above, due to the COVID-19 pandemic, Wave 1 was undertaken in-person and Wave 2 via telephone, which may have further deterred disclosure of mental health difficulties during this time period. Although this has not yet been analysed among Sri Lankan adults, there is evidence that the PHQ performs well in phone modality. 39 It would be beneficial for future studies to investigate issues related to measuring mental health, applicability and cultural safety of mental health measures.

Equality, diversity and engagement

The sampling design28 prioritised coverage of all demographics including socioeconomically disadvantaged communities by maximising the number of clusters in relation to the overall sample size, explicitly including socioeconomic indicators in the stratification of PSUs within the sampling frame and ensuring that all districts and sectors within districts were covered in the final sample. To include participants from all ethnicities who speak different languages, interviews were carried out by a team of field staff who spoke Sinhala, Tamil and English. If the selected participant spoke a language not spoken by the interviewer, the interview was rescheduled to another time by another interviewer. Following the ISIS terrorist attacks in April 2019, there was a fall in Muslim response rates, and it became unfeasible to survey one predominantly Muslim PSU due to security reasons. Despite replacing this PSU with a substitute from the same stratum matching by district, sector and level of socioeconomic development, it was not possible to match the ethnic profile of the original PSU. During the analysis stage, under-representation of specific ethnic groups or geographical regions in the final sample was directly dealt with by unweighting that took into account these characteristics. Regarding diversity of the research team, there was a range of expertise, spanning epidemiology, social science, public health and statistics. Three of the paper’s authors are early career researchers.

Community engagement and involvement

The SLHAS has maintained contact with its participants by sending all participants a newsletter updating them on progress and developments in the study, and including postcards for participants to send back feedback, which many have. The SLHAS also makes efforts to widely communicate its research findings and publications to the general public via social media and the local news media.

Acknowledgements

Ravindra Prasan Rannan-Eliya and Kishan Patel are joint last authors of this paper. The authors thank their colleagues that facilitated the SLHAS, Dr Alexis Palfreyman, Safiya Riyaz and Kavitha Vijayaraj for support in the design of the tools and training of data collection staff and Professor David Osrin, who was principal investigator of the NIHR Global Health Research Group (17/63/47).

Contributions of authors

Thilagawathi Abi Deivanayagam (https://orcid.org/0000-0002-1130-2333) (NIHR Academic Clinical Fellow in Public Health) contributed to study design and data acquisition, co-ordinated the analysis and drafting process and drafted the manuscript.

Sorcha Ní Chobhthaigh (https://orcid.org/0000-0002-5215-6352) (PhD Student in Public Health) contributed to study design and data acquisition, and drafted the manuscript.

Delan Devakumar (https://orcid.org/0000-0002-8695-6817) (Professor of Global Child Health) contributed to study design and data acquisition.

Kishan Patel (https://orcid.org/0000-0002-5693-6831) (Research Fellow) contributed to study design and data acquisition, performed statistical analysis and drafted the manuscript.

Ravindra Prasan Rannan-Eliya (https://orcid.org/0000-0002-5013-2816) (Executive Director at the Institute of Health Policy, Sri Lanka) conceived, designed and managed the data acquisition.

SLHAS Collaborators conceived, designed and managed the data acquisition.

All authors contributed to editing the manuscript and read and approved the final version of the manuscript.

Disclosure of interests

Full disclosure of interests: Completed ICMJE forms for all authors, including all related interests, are available in the toolkit on the NIHR Journals Library report publication page at https://doi.org/10.3310/HJWA5078.

Primary conflicts of interest: Ravindra Prasan Rannan-Eliya has received grants from the Institute for Health Policy and Swiss National Science Foundation (SNSF).

Data-sharing statement

Any queries on the new data created in the preparation of this article should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review.

Ethics statement

Ethical approval for the SLHAS data collection was obtained from the Sri Lanka Medical Association (ERC/18-022). Data for this study were requested from the Institute of Health Policy, Colombo, Sri Lanka.

Information governance statement

University College London (UCL) is committed to handling all personal information in line with the UK Data Protection Act (2018) and the General Data Protection Regulation (EU GDPR) 2016/679. Under the Data Protection legislation, UCL is the Data Controller.

Funding

This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Global Health Research programme as award number 17/63/47 using UK aid from the UK Government to support global health research. The views expressed in this publication are those of the author(s) and not necessarily those of the NIHR or the UK government. Research is published in the NIHR Global Health Research Journal. See the NIHR Funding and Awards website for further award information.

The SLHAS Wave 1 data collection was supported by the Swiss Agency for Development Cooperation (SDC) and the Swiss National Science Foundation (SNSF) through the Swiss Programme for Research on Global Issues for Development (r4d programme) by the grant ‘Inclusive social protection for chronic health problems’ (Grant number 400640_160374), and the Institute for Health Policy Public Interest Research Fund (Grant number PIRF-2018-02).

Department of Health and Social Care disclaimer

This publication presents independent research commissioned by the National Institute for Health and Care Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, MRC, NIHR Coordinating Centre, the GHR programme or the Department of Health and Social Care.

This article was published based on current knowledge at the time and date of publication. NIHR is committed to being inclusive and will continually monitor best practice and guidance in relation to terminology and language to ensure that we remain relevant to our stakeholders.

About this article

The contractual start date for this research was in April 2018. This article began editorial review in March 2023 and was accepted for publication in October 2023. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The Global Health Research editors and publisher have tried to ensure the accuracy of the author’s article and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this article.

Copyright

Copyright © 2024 Deivanayagam et al. This work was produced by Deivanayagam et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

Members of SLHAS collaborators author group and their affiliations

1. Professor Vajira Dissanayake, Dean, Faculty of Medicine, University of Colombo, Colombo, Sri Lanka.

2. Dr Prasadini Perera, Senior Lecturer, Faculty of Allied Health Sciences, University of Peradeniya, Peradeniya, Sri Lanka.

3. Professor HMM Herath, Professor, Faculty of Medicine, University of Ruhuna, Galle, Sri Lanka.

4. Professor Bilesha Perera, Professor, Faculty of Medicine, University of Ruhuna, Galle, Sri Lanka.

5. Dr Nilmini Wijemunige, Fellow, Institute for Health Policy, Colombo, Sri Lanka.

6. Dr Anuji Gamage, Senior Lecturer, Faculty of Medicine, General Sir John Kotelawala Defence University, Rathmalana, Sri Lanka.

7. Dr Shanti Dalpatadu, Senior Fellow, Institute for Health Policy, Colombo, Sri Lanka.

8. Dr Sarath Samarage, Senior Fellow, Institute for Health Policy, Colombo, Sri Lanka.