South Africa has an estimated 75 % treatment gap for common mental disorders with only 25% of people receiving treatment of any kind for their condition (1). To address this gap, the South African Department of Health has adopted the integration of mental health care into routine services, including screening for mental health disorders, at a primary health care (PHC) level using a task sharing approach as a policy imperative (2). While this may increase service availability, a lack of perceived need or awareness of signs and symptoms and lack of awareness of service availability impacts demand for mental health services (3, 4). Additionally, stigma and misinformation about mental health and treatment pose barriers to help seeking, particularly in African contexts where mental illness is highly stigmatised (5). Integration of mental health services at a community level is fundamental to strengthening health systems (6) including interventions that increase mental health literacy to improve help seeking for mental health disorders (5, 7) and mental health screening to improve detection of mental health problems at a community level (4).
Screening tools for mental health have shown to improve detection of disorders at a community level and are an important first step along the treatment cascade to reduce the treatment gap (4, 7). It is important that screening tools developed are culturally appropriate (8) particularly when used at a community level (9, 10). Checklist based screening tools are mostly developed in high income contexts and often lack cultural sensitivity (8). Detection at a community level by lay workers using an alternative prototype matching approach has been found to promote mental health service use in developing contexts (3, 4, 7). Prototype matching involves detecting a mental health condition by matching a patient’s symptoms against a prototype paragraph-length description of the condition (11).
Community health workers (CHWs) provide an important link between communities and PHC facilities and have been found effective in detection and linkage to care for physical (12-14) (15) and mental health conditions in low- and middle-income countries [LMICs] (4, 7). The South African Department of Health (DoH) has implemented a PHC re-engineering strategy which includes the development of PHC community health teams formally known as Ward based PHC outreach teams with a focus on health promotion and prevention at community level (16). The teams are made up of CHWs supervised by an Outreach Team Leader (OTL), usually an enrolled or professional nurse, and are linked to PHC facilities (17). Their central role is health promotion, screening and linkage to care at a PHC facility level (18). Up until now, they have mainly focused on physical health conditions.
This study is part of the Southern African Research Consortium for Mental health INTegration (SMhINT) project in collaboration with the KwaZulu-Natal Department of Health (KZN DoH), which has been using implementation science to evaluate the implementation of a collaborative care package for integrated primary mental health care (known as the Mental health INTegration [MhINT] package) for widespread scale-up using a learning health system approach (19). The implementation strategies used in MhINT to implement and scale up the collaborative care package include training, supporting tools and materials, and the use of continuous quality improvement [CQI] (20, 21).
The need to strengthen identification of common mental disorders at a PHC level was identified through the first stage evaluation of the original MhINT package (22, 23). This prompted the development and validation of the Brief Mental Health Screening Tool [BMH] (24) which has since been adopted by the KwaZulu-Natal DoH in the battery of screening tests offered at the PHC facility level as part of routine care. The BMH is a 7-item screening tool comprised of brief versions of the Alcohol Use Disorders Identification Test (AUDIT), the Patient Health Questionnaire (PHQ-9) and the General Anxiety Scale (GAD-7), and was validated for use within PHC facilities. In this regard the gold standard was professional nurse diagnosis using Adult Primary Care (APC) which is a nationally adopted integrated set of chronic care guidelines used by PHC nurses in South Africa (25).
A lack of a clear mental health pathway for care at a community level was also identified through the first stage evaluation of MhINT, and like the PHC context, no existing tools for identifying people with mental health problems were available in the community package of services that routinely include identification tools for physical health conditions such as TB, diabetes and HIV. Additionally, a need to expand detection of common mental disorders (CMDs) to a community level was identified by the DoH and a lack of a standardised screening tool for this purpose was noted. Initial consideration of the BMH tool was eschewed given that it focused on symptoms and commonly used psychiatric labels which was felt to be outside the scope of CHWs. Further, given low levels of mental health literacy in South African contexts (5), the DoH was concerned that this approach could result in labelling and stigmatisation of people who screened positive for a CMD in the community. The risk of this outcome is lessened at a PHC facility level where the BMH is administered by a nurse and confidentiality and patients’ rights are regulated by legal and professional ethical codes of conduct. The South African DoH and the KZN DoH thus requested the SMhINT team to develop and validate a mental health tool for use by CHWs to identify household members with possible mental health problems during their routine household visits. This gave rise to the development of the Community Mental Health Education and Detection (CMED) Tool(10) which was informed by the development of the Community Informant Detection Tool (CIDT) used to promote help-seeking among people with mental health disorders in Nepal as part of the Programme for Improving Mental Health Care (PRIME) (9). Apart from helping to identify individuals at a household level who may have mental health problems, the CMED was extended to include psychoeducation about these mental health problems as well as elements of a healthy lifestyle to educate households on what can be done to improve one’s emotional health at home(10). The addition of mental health education reduces the potential to promote labelling in contexts where mental health literacy is poor.
Using a prototype matching approach, the CMED incorporates local idioms in five case vignettes (prototype paragraphs) and related illustrations to facilitate identification of individuals with possible characteristics associated with depression, anxiety, psychosis, harmful alcohol and drug use (10). The prototype paragraph avoids the use of labels and diagnostic categories using vignettes that describe functional impairment of some of the most common mental disorders in everyday language. Each vignette includes a story about a character and is labelled according to the protagonists’ name e.g. Nontobeko is the protagonist in the story about a person with depression symptoms, the vignette is labelled as ‘The story of Nontobeko’ (Figure 1) and not by an associated mental health condition, e.g., depression to avoid labelling and possible stigmatisation(10). Following reading of the vignette, an interactive psychoeducation component has been added to raise mental health awareness about each condition. This is followed by two structured questions that guide the CHW in matching symptoms with the prototype vignette and determines if the family member requires a referral for further care (Figure 1). A flowchart assists the CHW in determining which vignette to read in the household (Figure 2).
The aim of the current study was to assess the accuracy of the newly developed Community Mental Health Education and Detection (CMED) Tool against the BMH which constituted the next level of screening in the system in identifying adult individuals with possible mental health problems in the household at a community level.