Performance-based incentives may be appropriate to address challenges to delivery of prevention of vertical transmission of HIV services in rural Mozambique: A qualitative investigation

Background: Performance-based incentives (PBIs) have garnered global attention as a promising strategy to improve healthcare delivery to vulnerable populations. However, literature gaps in the context in which an intervention is implemented and how the PBIs were developed exist. Therefore, we (1) characterized the barriers and promoters to prevention of vertical transmission of HIV (PVT) service delivery in rural Mozambique, where the vertical transmission rate is 12 %, and (2) assessed the appropriateness for a PBI’s intervention and application to PVT. Methods: We conducted 24 semi-structured interviews with nurses, volunteers, community health workers, and traditional birth attendants about the barriers and promoters they experienced delivering PVT services. We then explored emergent themes in subsequent focus group discussions (n = 7, total participants N = 92) and elicited participant perspectives on PBIs. The ecological motivation-opportunity-ability framework guided our iterative data collection and thematic analysis processes. Results: The interviews revealed that while all health worker cadres were motivated intrinsically and by social recognition, they were dissatisfied with low and late remuneration. Facility-based staff were challenged by factors across the rest of the ecological levels, primarily in the opportunity domain, including the following: poor referral and record systems (work mandate), high workload, stock-outs, poor infrastructure (facility environment), and delays in obtaining patient results and donor payment discrepancies (administrative). Community-based cadres’ opportunity challenges included lack of supplies, distance (work environment), lack of incorporation into the health system (administration), and ability challenges of incorrect knowledge (health worker). PBIs based on social recognition and that enable action on intrinsic motivation through training, supervision, and collaboration were thought to have the most potential for targeting improvements in record and referral systems and better integrating community-based health workers into the health system. Concerns about the implementation of incentives included neglect of non-incentivized tasks and distorted motivation among colleagues. Conclusions: We found that highly motivated health workers encountered severe opportunity challenges in their PVT mandate. PBIs have the potential to address key barriers that facility- and community-based health workers encounter when delivering PVT services, specifically by building upon existing intrinsic motivation and leveraging highly valued social recognition. We recommend a controlled intervention to monitor incentives’ effects on worker motivation and non-incentivized tasks to generate insights about the feasibility of PBIs to improve the delivery of PVT services. Two complimentary frameworks were applied to guide the study design and data analysis [6] (Fig. 1). The first was the ecologically embedded determinants of performance research agenda, which is embedded within an ecological framework that places the health worker at the center and moves outwards towards the political and economic environment [19]. The second was the motivation-opportunity-ability framework, which is grounded in human resources and operations management [20] and posits that three domains are required for optimal worker performance [21]. In integrating these frameworks, we operationalized “motivation” as the individual’s desire and willingness to act. “Opportunity” encompassed the many contextual factors that enable action beyond the individual. “Ability” included the skills and knowledge to execute action and overlaps with both the individual level and more distal levels of the ecological determinants of performance framework [6]. The integrated ecological motivation-opportunity-ability framework for health workers delivering prevention of vertical transmission of HIV services, modified from [6] p. 785 Mozambique’s legacy of colonization, war for independence (1964–1974), and civil war (1977–1992) left its health system and infrastructure unprepared for the HIV/AIDS epidemic [22]. Today, Mozambique has only 4 physicians and 41 nurses per 100 000 people, far below the regional average [23]. Task-shifting initiatives such as training mid-level técnicos de cirúrgia (surgical technicians) have helped address the skilled labor shortage for surgical needs [24] but have not alleviated the workload of nurses and midwives who deliver PVT services, which were integrated into antenatal care at the primary care level and free to patients [25]. We conducted this research in 2012 in a rural district in northern Inhambane Province, where CARE International was the PEPFAR-implementing partner. The district had a population of 56 000, few maintained roads, and irregular public transportation. In 2012, there were approximately 2700 pregnancies in the district, with an estimated 53 % of births occurring at health facilities [26]. HIV prevalence among pregnant women attending antenatal care was 10.5 % [27]. The public health system was comprised of one type III health facility in the district capital, one type III peripheral health facility, and four type II peripheral health facilities. The two physicians for the entire district were based at the type III facility in the district capital. A técnico led the largest peripheral health facility and nurses led the others. At the time of the study, the district and large peripheral type III facilities were the only facilities where patients could access ART (when CD4 count ≤350 cells/mm); only antiretroviral prophylaxis was available at the type II peripheral facilities. There were no private health facilities or physicians. A number of curandeiros (traditional healers) practiced in the district. Four cadres of health workers provided PVT services within the district (Table 1). Maternal and child health nurses provided the majority of clinical PVT services at health facilities. Activistas, or community volunteers, provided home care and counseling to individuals living with HIV/AIDS and received supervision and financial support from CARE International. Community health workers (CHWs) provided a broad portfolio of health services to households within 10 km2 of their home [28]. CHWs were trained in late 2011 and began working in early 2012 and received support and supervision from the implementing partner Malaria Consortium in addition to the Ministry of Health [29]. Traditional birth attendants (TBAs) historically assisted with home births but now increasingly focused on referral for health facility deliveries. TBAs were not systematically organized or supervised. Organizational structure and description of key services provided by the four health worker cadres preventing vertical transmission of HIV in rural Mozambique a60 % of the minimum monthly salary, per government recommendations To characterize health workers’ barriers and promoters (objective 1), we recruited members of the four cadres for semi-structured interviews. Maternal and child health nurses were purposively sampled based on their role and type of health facility (district, large peripheral, small peripheral). Key informants from two activista associations identified activistas, who were purposively sampled based on level of engagement. CHWs were identified by their district coordinator and were invited to participate when they visited the health facility to stock-up on supplies. A convenience sample of TBAs attending a training jointly facilitated by the district health authority and CARE International was invited to participate. Sample sizes for each cadre were based upon achieving the saturation needed to outline overarching themes [30], with the intent to expound upon these in subsequent focus group discussions. The interview guide contained questions about participants’ experiences delivering care to HIV-infected women and their HIV-exposed children, as well as their perceptions of the barriers and facilitators mothers face in the uptake of PVT services (Additional File 1). The guide was modified for each health worker cadre and pre-tested with the corresponding cadres in a neighboring district. Two Mozambican research assistants experienced in qualitative research conducted the interviews from September 2012 to January 2013. Interviews were conducted in Xitswa or Portuguese per participant preference and took approximately 60 min. Participants were interviewed in private spaces at health facilities (nurses, activistas), in their communities (CHWs), and training site (TBAs). To share, validate, and expound upon early findings from the interviews (objective 1) and assess the appropriateness of PBIs for PVT service delivery (objective 2), activistas, TBAs, and an array of facility-based health workers were recruited for focus group discussions. Representatives of each health sector at the district and large peripheral facilities and all staff at the small peripheral health facility were invited to participate because key informants strongly felt that all staff contributed to PVT care. All members of the two activista associations were invited to participate. TBAs known to be active in their communities were invited through key informants and snowball recruitment. No focus group was conducted with CHWs because concurrent interviews with HIV-infected mothers did not show women were receiving PVT services from them at that time. We conducted a total of seven focus groups lasting 90–120 min in March 2013. Participants were asked about types of incentives, how goals should be set and assessed, and concerns about implementing PBIs. One focus group was conducted at each of the three health facilities (district type III facility (n = 12); peripheral type III facility (n = 12); type II (n = 3)) in Portuguese. One focus group was conducted with each of the two activista associations in Xitswa at their respective meeting locations (n = 22 each). A sixth focus group with TBAs (n = 6) was conducted in Xitswa at a community meeting location. The final focus group was conducted with representatives from each of the three health facilities and two activista associations (total n = 13). The interviews and focus groups were complimented by participant observation primarily conducted at the three health facilities and at activista meetings between July 2012 and March 2013. Participant observation at the two type III and one type II facilities was conducted at minimum on a biweekly basis at each facility, during facility business hours, and included a few facility-wide meetings on strategies for integrated HIV/AIDS case management. Participant observation at activista association meetings was conducted once per month. Handwritten notes were recorded and were subsequently typed. The interviews and focus groups were audio recorded and accompanied by detailed handwritten notes. The interviews were transcribed into Portuguese, and the detailed notes from the focus groups were typed in Portuguese with the support of the audio recording. All transcripts were translated into English and were coded by two co-authors using the thematic analysis approach [31]. Interview results were shared with focus group participants to prompt further discussion, creating an iterative analysis process. Participant observation data was used to triangulate themes and validate findings [32].