Health care providers’ perceptions of barriers to perinatal mental healthcare in South Africa

Background: Perinatal mental disorders are a leading contributor to morbidity and mortality during pregnancy and postpartum, and are highly treatable when identified early. However, many women, especially in low and middle-income countries, lack access to routine identification and treatment of mental illness in public health settings. The prevalence of perinatal depression and anxiety disorders, common mental disorders, is three times higher for South African women relative to women in high-income countries. The public health system has begun to integrate mental health into maternal care, making South Africa a relevant case study of perinatal mental healthcare. Yet studies are few. We sought to investigate healthcare providers’ perceptions of the barriers to early identification and screening of common perinatal mental disorders in public health facilities in South Africa. Methods: Employing qualitative methods, we used purposive sampling to identify study participants, supplemented by snowball sampling. From September 2019–June 2020, we conducted in-depth interviews with 24 key informants in South Africa. All interviews were recorded and transcribed verbatim. We used a thematic approach to generate initial analytical themes and then conducted iterative coding to refine them. We adapted a delivery systems’ framework to organise the findings, depicted in a conceptual map. Results: Reported barriers to early identification and treatment of mental illness in the perinatal period encompassed four levels: (1) structural factors related to policies, systems and resources; (2) socio-cultural factors, including language and cultural barriers; (3) organisational factors, such as lack of provider preparation and training and overburdened clinics; and (4) individual patient and healthcare provider factors. Conclusion: Barriers act across multiple levels to reduce quality mental health promotion and care, thereby creating an environment where inequitable access to identification of mental disorders and quality mental health services was embedded into systems and everyday practice. Integrated interventions across multiple levels are essential to improve the early identification and treatment of mental illness in perinatal women in South Africa. We provide recommendations derived from our findings to overcome barriers at each of the four identified levels. We employed a qualitative design and methods to explore the complex nature of barriers to mental health services. Qualitative methods allow for building trust with key informants and fostering a meaningful research relationship to better ensure the trustworthiness and reliability of qualitative data [40]. The research conformed with the Consolidated Criteria for Reporting Qualitative Research (COREQ) Checklist. We included a detailed account of the COREQ checklist, as applied to our research, in Table A of the Supplementary Materials section [41] . We used purposive sampling to identify initial key informants for the study, supplemented by snowball sampling to locate additional key informants [42]. We selected eligible key informants based on three criteria: 1) knowledge about the concern under investigation; 2) willingness to discuss it; and 3) representation of a range of perspectives [42]. We recruited English speaking key informants via an email invitation that included detailed information about the study. We received ethics approval from the Institutional Review Board of the University of Massachusetts Boston. All key informants provided verbal informed consent for study participation and digital recording. We maintained key informants’ confidentiality and anonymity throughout. We stored digital recordings and data in an encrypted, password-protected, secure location requiring authentication. A total of 24 key informants with expertise in maternal health and/or mental health agreed to participate. Data were collected through in-depth interviews using a semi-structured question guide to capture key informants’ perceptions of barriers to perinatal mental ill-health identification and treatment in practice. The interviews were conducted by the first author, from September 2019–June 2020 via Zoom (20); telephone (2) and Qualtrics (2), an online survey tool. The same question guide was used for all data collection modes. The question guide included four main categories of questions related to: the perinatal mental health landscape and service delivery; mental health policies and practices concerning screening and early identification of CPMDs; implementation of mental health policies; and recommendations for addressing barriers to implementation. Key Informant interviews spanned 45–80 min in duration. We conducted interviews until saturation was reached, i.e., no new information emerged of significance to the study aim [42]. Research has indicated that some participants prefer the use of Zoom to in-person interviews, as the benefits of Zoom include establishing rapport, convenience and user-friendliness [43]. This accurately characterises the researchers’ experiences in the current study. Additionally, Qualtrics was found to be a useful data collection tool as an alternative to Zoom for two respondents. They worked in busy clinical settings during the early days of the COVID-19 pandemic and thus communicated their interest in participating and their preference for the flexibility of the online survey format, in lieu of a scheduled, longer Zoom interview. These realities highlight the preferences for remote research interviews during COVID-19 among respondents based in clinical settings. Nonetheless, preferences might differ for other types of participants, work settings and at other times. The key informants included four medical doctors (three of them obstetricians); four psychologists; four mental health counselors/social workers; three psychiatrists (two focused on perinatal psychiatry); three birth and postpartum doulas; three nurse-midwives; and a government maternal health professional. We also interviewed two mental health academics, one with a clinical background in nursing, to gain their policy perspectives. All key informants were based in South Africa, working, or with extensive expertise in, maternal mental healthcare in the public health system. We listened to the recordings and transcribed them verbatim. We then reviewed the transcripts and began open coding, identifying emerging themes and related subthemes on healthcare provider perspectives. The authors met frequently to discuss the emerging themes, revise and refine them. After we identified and developed the themes, we then mapped the findings onto the theoretical multilevel conceptual framework, adapted from previous delivery systems models, (Fig. 1), originating from a systematic review that examined barriers to mental health service access for women with perinatal mental illness [44]. To ensure quality, rigour, and reliability in data collection, analyses and reporting, we maintained an audit trail, kept field notes, used purposive sampling, relied on thick description that retains the context of the data, employed a code-recode strategy, peer debriefings (discussion) and reflexivity among the co-authors [45]. Adapted model depicting multilevel conceptual framework for barriers to mental health services in the perinatal period [44] The conceptual framework organised and elucidated themes and subthemes on healthcare provider reported factors—occurring at the individual, organisational, sociocultural and structural-level—that were perceived barriers to early identification of perinatal mental illness cited by respondents. We then performed a cyclical, iterative review of themes and subthemes to further refine the “story” of the data (p. 87), represented in the adapted conceptual map (Fig. ​(Fig.2)2) [44, 46]. Conceptual map of barriers by level that impact early identification and screening for perinatal mental health services in South Africa