Improving quality of mental health care in low‐resource settings: lessons from PRIME

Kilbourne et al1 present a useful framework for measuring and improving the quality of mental health care. They identify several barriers to this undertaking, highlight examples of innovations that can overcome these barriers in several countries, and offer recommendations for improving the quality of mental health care. The vast majority of the examples that are cited are from high‐income country settings. It is worth reflecting on whether similar challenges exist in low‐ and middle‐income countries (LMIC), and indeed what solutions may be found in these diverse low‐resource settings.

Several key challenges can be identified for the improvement of the quality of mental health care in LMIC settings. First is the pervasive reality of limited resources, which has a major bearing not only on treatment coverage, but also quality. As just one structural indicator, there are 1.4 and 4.8 mental health workers per 100,000 population, respectively, in the African and South East Asian World Health Organization (WHO) regions, compared to 43.5 in the European region2. A second challenge is the lack of standardized service quality monitoring tools in LMIC, although instruments like the WHO Assessment Instrument for Mental health Systems (WHO‐AIMS)3 and the WHO QualityRights tool4 are assisting with this. Third is the weak health system environment, including problems of suboptimal and at times dysfunctional general health management information systems. And fourth are the diverse cultural environments and pathways to care, which make assessment of processes and outcomes of care highly challenging.

In the Programme for Improving Mental health carE (PRIME), we have faced all of these challenges in various forms, while attempting to integrate mental health care into diverse low‐resource primary care systems, and improve quality of care. PRIME is a research programme consortium working in Ethiopia, India, Nepal, South Africa and Uganda5. Our aim is to address the following “how” questions: how can we deliver evidence‐based psychosocial and pharmacological interventions in a manner that is integrated into existing primary care systems and sensitive to local cultural needs; how can we ensure high‐quality care while utilizing a task‐sharing approach, employing general primary care workers to deliver mental health care; and how can we ensure continuous quality improvement in challenging low‐resource settings?

Commencing in 2011, we worked closely with Ministry of Health partners to establish one district demonstration site in each country. Our early engagement with partners in each district site entailed the use of Theory of Change methods, to collaboratively map out hypothesized causal pathways from entry into the system to achieving the desired patient and population level outcomes6. In many respects the Theory of Change maps that we developed were underpinned by Donabedian's “structure, process, outcome” framework, so central to Kilbourne et al's paper.

In addition to facilitating local stakeholder partnerships, the Theory of Change approach enabled the PRIME country teams to identify a set of structure, process and outcome indicators. The indicators were then integrated into four main study designs to assess the implementation and impact of the PRIME mental health care plans in each district7. These studies included repeat community surveys to assess changes in population treatment coverage over a 3‐year period; repeat facility detection surveys to assess improvements in the capacity of primary health care workers to identify depression and alcohol use disorders; cohort studies of individuals living with psychosis, epilepsy, depression and alcohol use disorders, to assess improvements in individual level clinical symptoms, functioning and economic circumstances (in some countries including nested randomized controlled trials); and finally case studies to assess structural measures such as medication supply, human resources, facilities and process measures such as numbers of patients treated and referred. The findings of each of these studies are currently being analyzed.

The process of implementing these mental health care plans has highlighted a number of context‐specific quality of care challenges. In response to these, we have developed several local quality improvement initiatives, which are ongoing. These include: specific measures to improve detection of depression and reduce drop‐out from care in people with psychosis in Ethiopia; improving screening for depression, pharmacological management and health management information systems indicators in India; improving individual patient follow‐up in Nepal; facilitating the transition of primary care clinics to chronic disease management and patient‐centred care in South Africa; and building the capacity of records staff, health workers and facility managers to collect and use health management information systems data for mental health care in Uganda. PRIME country teams have simultaneously enrolled in online quality improvement courses hosted by the Institute for Healthcare Improvement, to build their own capacity to develop quality improvement measures and interventions.

We have also developed or adapted several tools to assist in the improvement of the quality of care. One example is the Enhancing Assessment of Common Therapeutic Factors (ENACT) scale in Nepal, which is used by mental health specialist supervisors to routinely assess clinical competence of non‐specialist health workers in the delivery of mental health care8. Another is the adaptation and use of the Institute for Healthcare Improvement's Plan Do Study Act (PDSA) cycle in South Africa.

Although the LMIC settings in which PRIME is working are very diverse, and differ substantially from the high‐income country settings referred to in Kilbourne et al's paper, there are several principles which the authors highlight that are equally relevant to these diverse settings.

First is the importance of using indicators that include structure, process and outcome variables to develop an integrated means of assessing quality of care. As mentioned earlier, the Donabedian framework of structure, process and outcome underpins the PRIME district mental health care plans and the Theory of Change approach that we used. This was essential to map out the steps in the pathway from engagement in the district sites to impact at the patient, system and population level. The challenge remains one of ensuring that routine health management information systems data can be used to monitor services, without the presence of a research infrastructure, and this is one of our key areas of work as PRIME winds up its programme in 2019.

A second principle is the importance of developing common metrics. Although PRIME countries differ substantially, we have been able to apply cross‐country study designs and quality improvement measures which have common elements and approaches9. A third principle is the importance of health systems investing in routine quality improvement measures. These were not present in any of the sites when we began our work in 2011.

Our hope is that by including and piloting these measures in our diverse LMIC settings we will be able to demonstrate both feasibility and impact. We therefore join Kilbourne et al in making the case for further investment in integrated quality improvement measures for mental health care, particularly in low‐resource settings.

Crick Lund
Department of Psychiatry and Mental Health, Alan J. Flisher Centre for Public Mental Health, University of Cape Town, South Africa; and Department of Population and Health Services Research, Centre for Global Mental Health, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK

This paper is an output from the PRIME Research Programme Consortium, funded by the UK Department of International Development for the benefit of developing countries. The views expressed are not necessarily those of the UK government.