Acceptability of community-based mentor mothers to support HIV-positive pregnant women on antiretroviral treatment in western Kenya: A qualitative study

Background: Option B+ is a comprehensive antiretroviral treatment (ART) designed for HIV-infected pregnant/ postpartum women. However, barriers to implementing Option B+ and establishing long-term ART adherence while facilitating retention in prevention of mother to child transmission of HIV (PMTCT) services remain. Community-based mentor mothers (cMMs) who can provide home-based support for PMTCT services may address some of the barriers to successful adoption and retention in Option B+. Thus, we evaluated the acceptability of using cMMs as home-based support for PMTCT services. Methods: Gender-matched in-depth interviews were conducted between September-November 2014 for HIV-infected pregnant/postpartum women and their male partners living in southwestern Kenya (n = 40); additionally, we conducted four focus groups involving 30 health workers (n = 70) within four health facilities. Audio-recordings were transcribed, translated, and then coded using a thematic analytical approach in which data were deductively and inductively coded with support from prior literature, identified themes within the interview guides, and emerging themes from the transcripts utilizing Dedoose software. Results: Overall, the study results suggest high acceptability of cMMs among individual participants and health workers. Stigma reduction, improvement of utilization of health care services, as well as ART adherence were most frequently discussed potential benefits of cMMs. Participants pictured a cMM as someone acting as a role model and confidant, and who was over 30 years old. Many respondents raised concerns about breaches of confidentiality and inadvertent disclosure. Respondent suggestions to overcome these issues included the cMM working in different communities than where she lives and attending home-visits with no identifying clothing as an HIV-related health worker. Conclusions: The home-based cMM approach may be a beneficial and acceptable strategy for promoting ART adherence and retention within PMTCT services for pregnant/postpartum women living with HIV. Considering the risks of inadvertent disclosure of HIV-infected status and related negative consequences for pregnant/postpartum women living with HIV, similar cMM program designs may benefit from recognizing and addressing these risks. Trial registration: The MOTIVATE! study was registered on July 7, 2015 at the ClinicalTrials.gov (NCT02491177). Formative data were collected between September and November 2014. Findings from this study were used to refine the data collection instruments and intervention plans in the quantitative stages of the MOTIVATE! trial (Mother and Infant Visit Adherence and Treatment Engagement) [30] initiated in December 2015. The MOTIVATE! trial is evaluating the impact of cMM and text messaging interventions to support comprehensive ART adherence and PMTCT retention for Option B+ implementation in Kenya using a 2 × 2 factorial design (Clinicaltrials.gov Identifier: {“type”:”clinical-trial”,”attrs”:{“text”:”NCT02491177″,”term_id”:”NCT02491177″}}NCT02491177). Eligibility criteria for the randomized trial following the formative qualitative part of the study included HIV-infected pregnant woman 18 years of age or older, having access to a mobile phone, and agreeing to cMM visitation and residing within the catchment area served by the health facility where they enrolled in the study. Details of the MOTIVATE! study protocol and intervention design have been published elsewhere [30]. This formative study was conducted in four health facilities located in Homa Bay, Migori and Kisumu Counties of southwestern Kenya. Approximately a third of all HIV infections in Kenya are located in this region [31, 32]. HIV county prevalence rate remains high: 14.3% in Migori, 19.9% in Kisumu, and 26.0% in Homa Bay County [33]. In 2013, about 15,000 women in these counties received maternal prophylaxis for PMTCT with estimated 62–88% coverage [34]. The main economic activities in these three counties include fishing and subsistence farming. This cMM strategy was developed by the MOTIVATE! study team [30]. Similar to facility-based mentor mothers, cMMs are HIV-infected women with recent (6 months to 2 years) pregnancy experience and on Option B+ ART. They serve as peer mentors to HIV-positive pregnant and postpartum women and aim to improve PMTCT service uptake and retention. Table 1 outlines differences between cMMs and facility-based mentor mothers. The cMMs receive a two-week intensive training based on curricula adapted from UNICEF, Medicine Sans Frontiers, and KMMP [29]. Comparisons between facility-based and community-based mentor mothers The cMMs engage with HIV-infected pregnant woman at earliest identification within the antenatal clinic and follow these women up to at least 1 year postpartum. Follow-up involves up to 4 prenatal and 9 postnatal home visits tailored to complement pre- and postnatal antenatal care visits, infant immunization schedules, and infant HIV testing in line with the Kenyan guidelines for caring for the newborn at home [35]. These guidelines advocate for five visits with 2 prenatal visits occurring 4 weeks apart, 1 delivery day visit and 2 postnatal visits one on day 3 and day 7 after delivery. In order to reinforce information or health topics discussed by the service provider at the health facility, we added home visits at 6, 10, and 14 weeks of infant age, and at months of age 6, 9, and 12 which coincide with immunization dates and important dates for early infant diagnosis and testing and follow up. The topics discussed during home visits include: promotion of antenatal care (ANC), birth planning, adherence to antiretroviral drugs, immediate newborn care, family planning, care of the HIV exposed infant (including infant prophylaxis and early infant diagnosis), feeding methods and weaning. The cMMs collect data at each visit on mobile phones which have a data collection software (Open Data Kit) installed on them. This application is used not only to collect data, but also prompts the cMMs to cover specific health topics. The health topics to be discussed at each visit are dependent on the gestation of the pregnancy or the age of the infant. As this was a formative study investigating the feasibility and acceptability of community-based mentor mothers, we adapted a thematic analysis approach, and grounded our findings in the data rather than using an existing theoretical framework. A convenience sample of participants (n = 70) meeting eligibility criteria set for the qualitative phase of the MOTIVATE! Trial participated in in-depth interviews (IDI) between September and November 2014. Women were eligible if they were HIV-infected and pregnant/postpartum, 18 years or older and willing to participate. Male partners of HIV-infected pregnant woman were eligible if they were aware of partners’ HIV status, 18 years or older and willing to participate. Health care workers currently employed at one of the 20 study sites or part of the county health management team were considered to be eligible. Reporting of this study adheres to the consolidated criteria for reporting qualitative research (COREQ) guidelines. Gender-matched in-depth interviews were conducted involving 40 HIV-infected pregnant women and those within 6-weeks-postpartum (N = 20), along with their male partners (N = 20), half of which were HIV-infected living in seroconcordant relationships and half which were HIV-negative living in serodiscordant relationships. Health workers, registers, and patient files at health care facilities providing Option B+ services were consulted and used to identify potential participants. A short private session was used to evaluate interests in study participation and to identify eligibility among potential participants, plain-language explanations were provided regarding the separation of study participation from regular medical care and the option to decline to participate in the study. Male partners were contacted regarding potential participation in an interview if female participant agreed to it and has previously disclosed her HIV status to her male partner. Private settings were used for conducting the individual interviews offered in English or Dholuo language, according to the participant’s preference, and lasted approximately 1–1.5 h. Demographics, employment, pregnancy and HIV-related information were collected for each participant. Various health care providers (n = 30) were purposively selected for inclusion in the focus groups (7–8 from each of the four selected facilities). These included facility nurses, community health workers, health educators, mentor mothers, counselors for HIV, laboratory technicians, facility in-charges, program technical advisors, and administrative staff all selected for maximum variation in occupational perspectives. The research coordinator invited potential participants to focus groups at their health facility. Characteristics of participants were collected, including demographics (gender, age, educational level, marital status, the number of living children, clinic name) and employment characteristics (length of time in the current employment, current employment description). Focus groups lasted approximately 2–3 h. Guides for in-depth interviews and the focus group discussions were based on published literature and the research team’s prior experience and research regarding pregnancy and HIV in similar settings [36, 37]. Topics developed within the two guides pertained to the acceptability of comprehensive ART for pregnant HIV-infected women, barriers and facilitators to HIV-care adherence and retention, and the acceptability of the cMM approach. English and Dholuo interview and focus group discussion guides are available as Additional files 1 and 2 respectively. This manuscript focuses on findings on the cMM approach. Findings pertaining to challenges in the health facilities on providing Option B+ are published elsewhere [38]. One male and one female experienced interviewers who were fluent in English and Dholuo, completed training in both qualitative research methods and the study topics. Digital audio- recordings were collected for the in-depth interviews and focus groups and then stored under password-protection in a secure location with regards to the participants’ privacy. These recordings were translated into English if necessary, and transcribed verbatim by trained qualitative staff with identifying data removed. Interviewers/moderators took notes for purposes of assistance with transcription. Dedoose qualitative software program (Sociocultural Research Consultants, LLC) was utilized, with coding and data analysis following a thematic analytic approach [39, 40]. Thematic analysis was chosen as our method of choice owing to its flexibility as it is characteristically independent of theory and epistemology [41]. A framework for qualitative coding was developed using existing literature, interview guides, and an emergent theme approach with data derived from the transcripts. A multi-step inductive process of coding was used in which transcripts were initially broadly-coded by three trained qualitative coders who then followed with fine-coding allowing for the emergence of additional topics, the refinement of major themes and identification of sub-themes. The coders established consistency of coding beginning by coding the same transcripts together and with frequent discussions between coders throughout the process. Data saturation was discussed and jointly concluded that the data saturation was reached. Findings were disseminated at participating health facilities and communities.