Addressing the treatment gap for perinatal depression within an integrated primary health care model: development and feasibility study in the Dr Kenneth Kaunda District, North West Province.

Abstract

Background: Perinatal depression (PND) is a common mental disorder (CMD) with onset either during pregnancy or in the postnatal period, with potentially harmful inter-generational impacts on families, and by extension on communities. In South Africa a combination of high prevalence rates for PND, an estimated treatment gap of 75 percent for CMDs, and a large medically uninsured population poses a public health and social burden. Compounding the issue, there is a lack of awareness of and minimal attention paid to PND in scarce-resourced primary health care (PHC) settings in South Africa. Consequently, screening, referral and treatment for PND is low to absent, as are targeted pharmacological and psychosocial therapies for PND. Internationally, evidence supports the concepts of both collaborative care and task-sharing to address PND in low-and middle-income countries (LMIC). In South Africa however, despite support for the integration of mental health services into general health care, promotion of perinatal mental health care, and endorsement of task-sharing in mental health care, promoted by a national mental health policy framework, there is an absence of clear strategies to address PND in the mandated maternity care guidelines in PHC. In response to this service and evidence gap, the aim of this study was to co-develop and evaluate the feasibility of a culturally and contextually appropriate integrated model of care for PND with PHC service users and service providers. The research aimed to contribute towards the body of evidence towards the development of integrated, PHC-based, task-shared collaborative care for PND in South Africa and other LMIC. The study was guided by explanatory models of illness and the UK Medical Research Council framework for complex interventions. Methods: Set in an urban 24-hour service community health centre in the Dr Kenneth Kaunda District, North West province, and nested within the larger PRogramme for Improving Mental health CarE (PRIME) project, the study was undertaken using a phased approach. The first step was an in-depth review of the literature on task-shared care, integrated or collaborative care for PND, particularly in LMIC, and the platforms, models and cadres used in task-shared care for PND. These essential components for task-shared PND care in LMIC were identified and guided the development of the interview schedules for both service users and service providers. In-depth semi-structured interviews were conducted with 20 service users to understand their perceptions, attitudes to task-shared care and recommendations to address PND, using thematic analysis to analyse the data. This work comprised the first phase of the study. In the second phase, nurses (n=10), HIV counsellors (n=20) and operational managers (n=4) were interviewed to gauge their clinical understanding of PND, attitudes to task-shared mental health care and recommendations to address PND. This was followed by a participatory workshop which included nurses, managers and specialists to co-develop a model of care for PND. An additional six key informants were interviewed for institutional perspectives and guidance on the model. Framework analysis was used to analyse the data in this phase. In the third phase, a quasi- experimental cohort design was used to recruit perinatal care attendees (n=54) to evaluate depression outcomes, feasibility and acceptability of the model. Primary care nurses consulting women attending antenatal and postnatal services were trained to identify women with depressive symptoms using a short maternal depression screening tool, and clinical assessment. Pregnant and postnatal women (6-48 weeks postpartum) who had mild/moderate depressive symptoms were referred to an existing 9-session manualized counselling intervention addressing common triggers of depressive symptoms, based on cognitive behavioural approaches, provided by a co-located non-specialist counsellor. Women with

moderate/severe depressive symptoms were referred to both the counsellor and upwards for specialist assessment and treatment. Participants were administered a questionnaire including the Patient Health Questionnaire 9 (PHQ9). Service users (n=31) identified by nurse clinicians and referred for counselling and/or further treatment were assigned to the intervention arm, and service users (n=23) not identified with PND by the nurse, but who screened positive on the PHQ-9 were assigned to the control arm. Participants were interviewed at baseline and four months after baseline to assess change in PHQ-9 scores. Qualitative process evaluations were also conducted with five service user participants and eight health workers after the four-month assessment to identify evidence of feasibility and acceptability, challenges and recommendations. Results: The qualitative results from the first phase indicated support for task-shared care and produced service-user recommendations to address the need for psychoeducation, support groups and counselling, either at community or facility-levels for PND. The second phase service provider engagement (participatory workshop) culminated in the co- development of a task-shared, collaborative care model for PND, with strengthened referral pathways, based on the nurse clinician screening, diagnosing, and referring onward to either a facility-based non-specialist counsellor, a doctor or a mental health specialist. In the third phase, an evaluation of the task-shared, collaborative care model with strengthened referral pathways to a co-located psychosocial intervention delivered by non-professional mental health workers at PHC level, indicated a clinically significant decline in depression scores (10- point reduction) in the intervention arm from baseline (M=14.3, SD 2.9), and at four- month follow-up (M=4.3, SD 4.5). Qualitative data indicated that participants experienced

the counselling intervention as beneficial and acceptable. The non-specialist, co-located counselling was viewed by most nurses as beneficial although there were recommendations to streamline the screening and diagnosis process. Process indicators suggest that the model is feasible and acceptable. Conclusion and recommendations: This study has contributed new applied knowledge regarding the development and evaluation of a task-shared, integrated, collaborative care model for PND at PHC level in South Africa, providing evidence of feasibility and acceptability of the model of care. The favourable results suggest the potential for a larger effectiveness study, based on the recommendations and lessons garnered from this study. At the time of this report, the policy developments within the mental health landscape demonstrate a level of awareness among a minority of policy-makers, researchers and health care providers of the need to promote perinatal mental health. However, the lessons from this study suggest that key policy level changes are required which include but are not confined to the adoption and reporting of mental health data elements and indicators for PND, and adaptations to the maternity guidelines to include detection in the form of brief screening, assessment, diagnosis and referral for PND. With reference to task-shared mental health care, the inclusion of social workers in counselling treatment plans, and the identification of appropriate cadres, trainers, training, and supervision for non-specialist mental health counsellors are critical factors that require concerted political will and effort.