Unintended Consequences of Community Health Worker Programs in South Africa

Abstract

Task shifting from trained clinicians to community health workers (CHWs) is a central, primary health care strategy advocated by global health policy planners in resource-poor settings where trained health professionals are scarce. The evidence base for the efficacy of these programs, however, is limited—in particular, research that identifies their potential unintended consequences. Based on sustained ethnographic study of CHWs working for AIDS projects in South Africa at the height of the country’s AIDS epidemic, this article identifies how structural and local factors produced unintended consequences for CHW programs. These consequences were (a) CHWs moonlighting for multiple organizations, (b) CHWs freelancing in communities without regulation, and (c) adverse patient outcomes resulting from uncoordinated care. These consequences stemmed from structural elements of a bureaucratically weak health system and from local grassroots dynamics that jeopardized long-term CHW program sustainability and eroded national health goals.

Keywords

community-based programs health policy HIV/AIDS program evaluations ethnography South Africa

Introduction

Recent public health proposals have called for expanding the role of community health workers (CHWs) in primary health care, particularly in sub-Saharan Africa (Singh & Sachs, 2013; World Health Organization [WHO], 2012). CHWs are unlicensed lay members of the community with some training in health care delivery who are viewed by global health experts as central to achieving comprehensive primary health care worldwide. Affordable approaches to improving the performance of CHWs are a high priority for middle- and low-income countries seeking to comply with global health policy frameworks (Boerma & Siyam, 2013; Cometto & Witter, 2013). Proposals for expanding CHW programs point to the low costs of hiring and training CHWs and to the potential for bringing CHW projects up to scale quickly. However, the evidence base for the efficacy of these programs is limited (Singh & Sachs, 2013)—in particular, research that identifies the dimensions of their successful implementation (Kangovi, Grande, & Trinh-Shevrin, 2015) and their possible unintended consequences (Lewin et al., 2010).

Program designers and evaluators have long been concerned that unintended outcomes of health and social interventions are insufficiently incorporated into program assessments (Funnell & Rogers, 2011; Rossi & Williams, 1972). Unintended outcomes are effects of an intervention that were not anticipated by the project’s designers (Jabeen, 2016; Merton, 1936). One reason that such outcomes are obscured in global health programming is that much research focuses narrowly on project intervention so that planners and evaluators have limited understanding of how programmatic goals translate into real-world outcomes (Rossi, Lipsey, & Freeman, 2003).

Social scientists contend that public health policy designers and evaluators attempt to control for or experiment on the unpredictable social (Biehl & Petryna, 2013), without a strong understanding of how structural and community processes influence the implementation of health projects (Bell, 2012; Colvin, 2015; Green, 2014). Furthermore, planners seldom attend to long-term dimensions of policy reform; instead, they rely on cross-sectional research conducted at a particular moment in the implementation process (Pierson, 2005). Based on sustained ethnographic study of CHWs working for public health AIDS programs in South Africa, this article identifies how structural factors and dynamic local-level processes produced unintended consequences for these programs.

Structural factors in global health include policy frameworks and the organization of national health systems. CHW programs are characterized by a vertical primary health care model (Campbell & Scott, 2011) that responds to infectious diseases through the scaling up of narrow, technical disease interventions (Biesma et al., 2009; Storeng, 2014). These vertical interventions have been criticized for their limited scope in responding to the complex needs of populations (DelVecchio Good, Good, & Grayman, 2010). Vertical disease models are driven by international donors and nongovernment agencies (McKay, 2017; Pfeiffer & Nichter, 2008), often producing fragmented health systems in countries receiving international aid (Babb & Chorev, 2016; Lieberman, 2011). These models struggle to sustain interventions over the long term as international funding priorities change (Hafner & Shiffman, 2013). Health system fragmentation and program instability produce conditions in which health projects can deviate from their intended goals (Biehl & Petryna, 2013).

Structural forces are connected to the social determinants of health (Link & Phelan, 1995; Phelan, Link, & Tehranifar, 2010). According to this framework, individuals and groups use resources including knowledge, power, and social connections to avoid risks and adopt protective strategies (Phelan et al., 2010). Unequal access to resources produces health disparities, where individuals and groups who have limited command over resources may experience morbidity or mortality from preventable disease. These health disparities are a form of structural violence (Farmer, 2003; Galtung, 1969), where unequal power and resources built into the social structure place people in harm’s way. Social scientists (Bureau-Point & Phan, 2016; Campbell, 2003) have shown how CHW programs intended to address health disparities might produce forms of structural violence unintended by CHW program designers as a result of failures in broader health system or design flaws embedded in CHW projects.

Structural forces also shape the experiences of CHWs on the ground, as they experience disjunctures between accomplishing the goals of standardized health interventions designed by global actors and meeting the local health needs of patients (Djellouli & Quevedo-Gómez, 2015; Justice, 1987; Nugus et al., 2018). In terms of local factors, CHW programs assume low-resource communities to be coherent spatial, bureaucratic, and social units (De Wet, 2011). These assumptions are, however, naïve to the complex power relations that organize social life (Campbell & Scott, 2011; Swidler & Watkins, 2009), making program outcomes harder to control. Furthermore, projects are regulated by external international organizations and by local agencies that recruit workers from target villages and neighborhoods. These workers operate in a gray zone (Kalofonos, 2014) between formal work and unpaid volunteerism, defined by low pay or by no remuneration at all (Bureau-Point & Phan, 2016; Maes, 2015; Swidler & Watkins, 2009; Takasugi & Lee, 2012), that reinforces race and gender inequalities in global health (Aronson & Neysmith, 1996; Hallgrimsdottir, Teghtsoonian, & Brown, 2008; Malkki, 2015) and increases the potential for unintended consequences of CHW programs (Takasugi & Lee, 2012).

I spent 21 months in South Africa, from 2007 to 2009, as a participant–observer of CHWs and their supervisors who were associated with programs administering health care to AIDS patients and their children. During the time of my fieldwork, global policy consensus and guidelines for task shifting in response to the AIDS epidemic were created (WHO, 2007). South African government agencies rapidly adopted the language of task shifting in their policy frameworks for CHWs (Department of Health, 2008). Having qualitative data from this critical period, 2007 to 2009, allows me to analyze the policy effects (Pierson, 2005) of formalizing CHW roles in local settings in South Africa. By studying the implications of initial policies on local stakeholders and interest groups, I provide insight into the challenges to long-term policy and program sustainability and the unintended consequences that can arise from implementation.

Analysis of my qualitative data revealed three unintended consequences of these programs: (a) CHWs moonlighting for multiple organizations, (b) CHWs freelancing in communities without regulation, and (c) adverse outcomes for patients resulting from uncoordinated care. Building on scholarship that calls for integrating structural with local dimensions of community health work (Baynes et al., 2017; Kok et al., 2017), this study finds that program design and administration lacked integration with the national health system of South Africa and also attended inadequately to the individual livelihoods of care workers and the embeddedness of health projects in their community settings.

CHWs, Task Shifting, and Programs of AIDS Care in South Africa

CHWs are defined as health workers who receive “standardized training outside the formal nursing or medical curricula to deliver a range of basic health, promotional, educational, and outreach services, and who [have] a defined role within the community system and larger health system” (Naimoli, Frymus, Wuliji, Franco, & Newsome, 2014, p. 3). CHWs have played a central role in mitigating the AIDS epidemic through task shifting (Schneider, Okello, & Lehmann, 2016), which the WHO (2007) has defined as follows:

the extension of the scope of practice of community health workers, including people living with HIV/AIDS, in order to enable them to assume some tasks previously undertaken by senior cadres (e.g. nurses and midwives, non-physician clinicians and medical doctors). (p. 45)

Prior to the availability of antiretroviral therapy (ART), CHW roles focused on supporting primary caregivers by providing home-based care to patients and AIDS education and counseling (Perry, Zulliger, & Rogers, 2014). With the advent of ART, CHW roles expanded to include clinical health services such as voluntary AIDS testing and counseling, treatment preparation, and follow-up support with treatment adherence counseling (Schneider & Lehmann, 2010). Task shifting from doctors to nurses and from nurses to CHWs has been recommended as a cost-saving measure to extend universal ART coverage (Meyer-Rath et al., 2017).

The AIDS epidemic in South Africa is one of the most severe on the continent, with the HIV prevalence rate currently at 18% for persons aged 15 to 49 years (Statistics South Africa, 2017). The spread of AIDS in South Africa was exacerbated by the AIDS denialist stance of the Mbeki administration, which slowed the public health rollout of life-prolonging ART (Karim, Churchyard, Karim, & Lawn, 2009; Mayosi et al., 2012). Since the end of Mbeki’s tenure in 2009, reductions in pediatric AIDS rates have been attributed to the distribution of ART and the prevention of mother-to-child transmission of HIV (Joint United Nations Programme on HIV/AIDS, 2015). South Africa now has 3.3 million patients enrolled in treatment and runs the largest public ART program of any country in the world (Johnson, Dorrington, & Moolla, 2017). But South Africa’s health system continues to face challenges from the legacy of a fragmented health system under apartheid, compounded by severe human resource constraints, particularly in the domain of primary health care, and leadership deficits at all levels (Fried, Harris, Eyles, & Moshabela, 2015).

These health disparities have produced asymmetries in South Africa between private and public health care. The private sector supports 16% of the population, with approximately US$1,370 spent on each patient, and this sector employs 70% of the health professionals; in contrast, the public hospital sector serves 68% of the population and spends US$220 per patient (Marten et al., 2014; van Rensburg, 2014). Efforts to introduce universal health insurance (National Health Insurance [NHI]) previously stalled (Harris, 2017), and there have been severe problems with leadership and accountability in managing the public health sector (Rispel, 2016). In the wake of changes in leadership, pilot implementation of NHI resumed (Plagerson, Patel, Hochfeld, & Ulriksen, 2019) and CHWs are envisaged to play central roles in primary health care (Republic of South Africa, 2017; Schneider, Schaay, Dudley, Goliath, & Qukula, 2015).

In the management of HIV/AIDS, South Africa relies heavily on CHWs to mitigate the wide-ranging consequences of HIV/AIDS (Schneider & Nxumalo, 2017). Schneider, Hlophe, and van Rensburg (2008) estimated that there were approximately 40,000 CHWs in South Africa, a labor force equivalent in size to South Africa’s nursing population, although, as Lund (2010) observed, there is no accurate census of this broad cadre of care workers. CHW labor is largely coordinated by international nongovernmental organizations (NGOs) or domestic community-based organizations affiliated with NGOs (Mayosi & Benatar, 2014). Most CHWs are economically marginalized African women (Daniels, Clarke, & Ringsberg, 2012; Swartz & Colvin, 2015; van de Ruit, 2017b). Furthermore, the blurred boundaries between formal work and volunteering lead to uneven pay scales for CHWs, with some health services offering no remuneration and others paying a modest stipend (Schneider et al., 2008; Swartz & Colvin, 2015). Early iterations of CHW programs in the 1970s and 1980s had a vision that CHWs were to be embedded in their community to perform social roles such as health education and counseling (Campbell & Scott, 2011). But community embeddedness is now no longer emphasized in CHW program design; rather, CHWs living in communities located on the periphery of health service systems are now valued for playing a bridging role between patients and the formal health system (Schneider & Lehmann, 2010).

Although proposals for scaling up ART are being celebrated, epidemiological research on large ART schemes is finding that the expansion of treatment is accompanied by higher rates of patient attrition, which threatens to undermine the goals of ART expansion (Auld et al., 2016; Mutevedzi, Lessells, & Newell, 2013; Nglazi et al., 2011). Higher rates of patient attrition have been attributed to the pressure that increased patient demand places on health systems (Fatti, Grimwood, Mothibi, & Shea, 2011). However, epidemiologists must also consider the organizational aspects of ART that affect patient retention and the potential for unintended consequences attendant on task shifting from trained clinicians to semiskilled care workers. The extensive role played by CHWs in the provision of ART (Mayosi & Benatar, 2014) provides an important context to examine the unintended consequences of CHW interventions.

Setting and Method

My data derive from the ethnographic fieldwork I conducted in South Africa from 2007 to 2009, principally in the province of KwaZulu-Natal, considered to be the epicenter of South Africa’s AIDS epidemic (Naidoo, Jacobson, Neugut, Dlova, & Mosam, 2016). I secured human subjects ethical clearance from the University of KwaZulu-Natal in South Africa (IRB # HSS/0249/07) and from the University of Pennsylvania in the United States (IRB # 806377). Names of people, places, and organizations were removed to protect the confidentiality of all participants.

The goal of this study was to develop a grounded theory (Glaser & Strauss, 1967) of how community health projects produced both intended and unintended outcomes in two racially marginalized settings in South Africa under conditions of economic and health insecurity (McKay, 2017; van de Ruit, 2017a). Theoretical sampling (Starks & Trinidad, 2007) based on residential characteristics of rural and urban settings and the kinds of care work undertaken was used to garner the diverse experiences of CHW performing their work roles. I spent more than 21 months in an urban township, a formerly racially segregated area outside a large metropolitan area. For a period of 12 months, I conducted frequent visits to a rural village.

The urban township is large; its oldest section was originally a mission settlement developed during the colonial period. Newer sections of the township were built during apartheid to house workers close to an industrial area; the area was expanded again to include low-income housing in the postapartheid era. As a result, the township is ethnically and linguistically diverse. While I was there, it had multiple health projects, a public hospital, and multiple state-run health clinics. The rural village, in contrast, is part of the former homeland system of KwaZulu. The village falls under the traditional authority system, in which the local Inkosi (chief) has bureaucratic and cultural authority, and the village is ethnically and linguistically homogeneous. The village was marked by scarcity while I was there; its health and economic infrastructure was limited and often ineffective. Although both locales were racially segregated settings characterized by economic marginalization, differences in the quantity and variation of CHW projects and coordination of health services were apparent. The spatial and organizational similarities and differences enabled me to apply the constant comparative method (Draucker, Martsolf, Ross, & Rusk, 2007) in both data gathering and analysis to understand how structural and grassroots factors led to differential outcomes in the health programs I chose as case studies.

I undertook participant observation and in-depth interviews to learn how intended and unintended consequences arose from the interactions of program supervisors, CHWs, and their patients (Blumer, 1986). My interactions with participants were conducted in English and Zulu. I had a working knowledge of Zulu and had two Zulu-speaking research assistants, one living in a rural village and the other in the urban area, who interpreted formal interviews as well as workshops and trainings. At these meetings, I took detailed field notes, and I accompanied CHWs on their community rounds. I conducted 73 in-depth interviews and facilitated six focus-group discussions that were recorded and later transcribed. The focus groups were concentrated in the rural village, where there were small organic groups such as a basket-weaving group, a women’s church group, and the remnants of a defunct home-based-care group who met and interacted regularly. These organic groups afforded me an opportunity to contextualize the AIDS projects within broader social and cultural groupings in the village. In contrast, the size of the urban township and the multiple economic activities its residents participated in meant that coordinating organic group meetings were more challenging. As a supplement to these primary data, I also analyzed national health care policy guidelines, organizational documents prepared by the CHW projects, and secondary sources of CHW research in sub-Saharan Africa.

Issues of access, power, and informed consent were an ongoing negotiation throughout the study. I gained access to the township and village through officials and respected community organizers and through my research assistants who lived locally. My research assistants played an important role in facilitating my access to communal spaces, gaining trust among my research participants, and guiding me through unfamiliar neighborhoods. My outsider status as a White, middle-class, woman researcher created asymmetries in these economically and racially marginalized settings. Sometimes, I had to dispel misunderstandings: research participants who mistrusted my intentions, project beneficiaries who mistook me for a social worker, or CHWs who worried I would report my findings back to their employers. In some cases, CHWs hoped I was offering employment; thus, I was concerned that their consent to have me shadow them on their daily rounds was based on a misconception about future employment. When I sensed any misconception, I clarified that my research would not bring economic rewards.

I worried, too, that my accompanying CHWs on rounds was altering the dynamic between them and their patients. To address this concern, I undertook two strategies to test the credibility of the exchanges I was observing. First, in the urban research site, I trained my research assistant to take field notes and encouraged her to conduct home visits with CHWs in my absence, so that we could compare exchanges when I was and was not present. Based on these comparisons, I was satisfied that my presence in the urban village was not leading to grossly skewed observations of the care-work situation. Second, toward the end of the study, I developed close relations with a few CHWs with whom I had overcome some initial barriers, and I interviewed them on several occasions to learn about their lives in more depth.

The four CHW projects I studied were Project 1, Project 2, Project 3, and Project 4. Initially, I conducted selective sampling by choosing organizations that provided support services to children orphaned by the AIDS epidemic; however, as I became increasingly focused on the role CHWs played in delivering and implementing AIDS care support, I expanded my sample to include home-based care to gain a broader exposure to the situation of CHWs. Three projects were based in an urban setting, and one was in a remote village. In the middle of my study, I added a fifth group, a rural care worker collective (hereinafter, Collective) operating in the rural village. Collective was a grassroots organization formed by CHWs, some of whom worked for projects and some of whom did not. By including Collective in my ethnography, I was able to contrast CHW projects initiated by NGOs, where CHWs were isolated versus a grassroots organization that sought to organize CHWs. This combination of case study organizations let me survey the structural organization of primary care in the South African health system along with a grassroots effort to reform this system. Supplementary Table 1 profiles the five case study organizations.

All of the CHWs I interviewed were economically marginalized African women receiving remuneration on a continuum from no pay at all to a small stipend. Furthermore, consistent with recent studies on CHW work (Daniels et al., 2012; Nxumalo, Goudge, & Manderson, 2016; Swartz & Colvin, 2015), the CHWs were primarily employed by NGOs or nonprofit organizations (NPOs), and the tasks they performed included a range of referral functions between welfare offices and health clinics, on behalf of patients with HIV or families affected by AIDS.

Field notes and transcribed interviews were analyzed with QSR International’s NVivo version 10 software, following inductive analytic techniques influenced by grounded theory. Initially, I undertook open coding with in vivo and process-coding strategies to generate a set of themes, and then supplemented my coding strategy with analytic memo writing. I followed up this first round of analysis with axial coding that focused on the most frequent or salient codes pertaining to CHWs’ situations (Charmaz, 2006). To test the credibility of my interpretation of themes, I performed member checks by sharing my research memo with my interviewees (Lincoln & Guba, 1985). The key themes of this study—notably the economic marginalization of CHWs, inadequate CHW supervision, and few avenues for career advancement—resonated with my research participants, and I later learned that, in the interceding years between my leaving the field in 2009 and my follow-up exchanges in 2012 and 2014, working conditions for CHWs had worsened.

Results

Three unintended consequences of CHW projects emerged as salient in my analysis: (a) CHWs worked for more than one health care project in their communities to stabilize their livelihoods, (b) freelance CHWs operated in communities without adequate supervision or organizational affiliation due to the high failure rate of CHW projects, and (c) both structural and grassroots processes contributed to adverse outcomes for patient care.

CHWs Moonlighting to Supplement Their Inadequate Earnings

Remuneration for CHWs varied inequitably across the four organizations: from the purely voluntary (for which CHWs were unpaid), to small compensation paid out in gift vouchers, to full-time monthly wages ranging from ZAR 700 (US$100) up to ZAR 1,500 (US$214). (The ZAR, South Africa’s currency, had an exchange rate of 1 ZAR to US$0.13 in 2009.) CHW wages fell within the lowest four deciles of the South African labor market, which had declined in real wages over the last two decades (Leibbrandt, Woolard, McEwen, & Koep, 2010). Low remuneration for CHWs incited both high attrition and moonlighting.

When I asked them why they had chosen to join AIDS care projects, most CHWs said they saw it as a pathway to securing full-time work. There was a pool of people willing to take CHW work; yet, they frequently quit the project after discovering that the stipends were low or that the work was short term. Although the literature on CHWs acknowledges that low remuneration and limited upward mobility are key factors in staff attrition (Takasugi & Lee, 2012), no studies address the issue of how CHWs offset job instability and low pay by moonlighting.

Several of the CHWs I interviewed augmented their salaries by working for multiple organizations at the same time. Employment diversification, or moonlighting, was widely practiced but seldom discussed openly because it was frowned on by organizations and caused tension among CHWs competing for the same employment opportunities. The topic of moonlighting came up on community rounds, however, when CHWs described to me the kinds of organizations that supported their patients and the tasks they performed. This practice was more common for CHWs working for the urban projects because there were several initiatives concentrated in their communities. This was less prevalent in the rural area where far fewer agencies were present.

For instance, when I accompanied Childcare Worker 1, on her rounds, she explained her motivations for working for two orphan projects in her community: “I want to speak the truth; it was the stipend. It was ZAR 500 (US$38) that made me work for Project 3. At Project 1 we get ZAR 300 (US$23) after 3 months.” This childcare worker provided care for the same orphan through two different organizations, and she described performing similar family surveillance activities for the organizations. For example, when representing Project 3, she conducted surveys of beneficiary families’ health and welfare needs. Similarly, for Project 1, she conducted an in-depth family assessment covering topics similar to those in the Project 3 survey. Moonlighting had created redundancy in the efforts of health care organizations on account of one CHW’s dual allegiance.

The director of Project 1 acknowledged that moonlighting was a common practice for CHWs in the organization and a cause for concern about duplicating services. But the director admitted to having little control over the practice because Project 1 CHWs were volunteers without a formal stipend; instead of pay, they were given gift vouchers for their service. But CHW moonlighting was not limited to organizations staffed by unpaid volunteers; it also occurred at higher-paying CHW projects such as Project 2, which paid ZAR 1,500, the highest wage of the organizations I studied.

At Project 2, moonlighting by CHWs was against the organization’s policy; yet, the practice among CHWs was nonetheless widespread. Home-Based-Care Worker 1, for example, managed a food distribution project in her community in addition to her duties for Project 2. Food was delivered by an external charity, and Home-Based-Care Worker 1 oversaw its distribution to her community members every month. Overseeing distribution took a day away from her work for Project 2. Not only was she taking time away from her CHW project, but her allocation of the food also lacked transparency and accountability. The food distribution project was meant to target the most destitute members of the community, but this did not appear to be what actually happened. On her rounds, Home-Based-Care Worker 1 reminded all the families with whom she worked to attend the food distribution, even though they did not appear to be the poorest members of the community. Furthermore, friends and family in her patronage network appeared to receive generous portions of food in contrast to their neighbors. This dynamic leading to unfair distribution suggested to me that patients served by Project 2 as well as Home-Based-Care Worker 1’s kin received preferential treatment over other community members due to her moonlighting. An overlap of employment gave her considerable discretionary power in the allocation of scarce resources.

This finding that CHWs moonlighted as a means to secure their livelihoods builds on anthropological scholarship on CHW remuneration and the moral economy of care (Maes, 2015; Maes, Kohrt, & Closser, 2010; Nading, 2013; Swartz & Colvin, 2015). Many CHW projects have policies against CHW remuneration; instead, tangible resources such as knowledge and skills are considered an optimal exchange for CHWs energy and time (Maes, Closser, Vorel, & Tesfaye, 2015). The literature examines contradictions that arise when economic rewards are introduced into economically marginalized settings that reshape relations among CHWs and between CHWs and NGOs (Swartz & Colvin, 2015). Yet, this study, in keeping with anthropological research (Maes, 2015), shows the negative consequences of low or no remuneration for CHWs. Moonlighting created competing priorities for CHWs between their patients and the different projects they served, contributed to program attrition, and created competition among CHWs.

In sum, moonlighting had detrimental consequences for the quality of CHW projects because it duplicated health services in the community and diluted CHWs’ commitment to their primary organization and work role. This practice was not intended by the NGOs running the projects, but it occurred anyway because CHW remuneration was so low. Moonlighting was more prevalent in the urban setting, as there were several projects operating in the area as opposed to the rural village where there were fewer projects.

Freelance CHWs Operating in Their Community Without Regulation

That CHWs operated in their community without organizational affiliation became apparent to me when I attended Collective’s workshops. This was a heated topic among Collective members in their group discussions because it had implications for both patient safety and employment equity. The majority of CHWs in Collective who were unaffiliated with organizations were in their predicament because their former employer had ceased its operation. An additional minority of unaffiliated CHWs were there as members of a nascent organization yet to receive official governmental clearance to begin operation. These unaffiliated freelance CHWs were doing their work without sanction because they saw a need in their communities and wanted to help despite bureaucratic delay in getting approval. Home-based-care workers who formerly belonged to NGOs, which had since shuttered their operations, continued to provide support for families caring for members with HIV/AIDS or tuberculosis—for instance, by bathing patients, watching them while family members were out of the house, and accompanying patients to the clinic. Childcare workers waiting for approval of their project had started offering after-school drop-in services at their home.

In South Africa, a state-sponsored employment creation project known as the Expanded Public Works Learnership Programme (Department of Public Works, 2009) encourages community groups to propose community projects and apply for certification. Once registered, a group may apply for 6 months’ funding through the learnership program. Prior to applying, however, they need to secure nonprofit status from the Department of Social Development. CHWs described delays from the time of submitting their paperwork to the receipt of their nonprofit certification, or NPO number. They described the cumbersome documentation required to prove that an organization is legitimate and complained about the apparent arbitrariness behind the allocation of NPO numbers. The NPO application process was, thus, a source of great anxiety for CHWs, who had few livelihood alternatives to their care work. In an interview with Childcare Worker 2, who had initiated her own after-school project for orphaned and poor children in her community, I asked about her experience applying for certification and recorded her reaction in my field notes:

Her manner changes, she becomes emotional and tense—saying that people should not have to fill out forms and wait when people are hungry now and sometimes they die. She says how hard it is, tears gather in her eyes, tears of frustration. She feels she has made empty promises because the letter of application that she wrote has yet to be approved.

Childcare Worker 2’s frustration drove her to freelance. Both failed projects and unregistered community groups had created a pool of CHWs like Childcare Worker 2 who operated in their communities without supervision, training, remuneration, or any form of regulation.

NPO applications posed another obstacle to regulated care work given that such certificates were limited in number. CHWs who successfully applied for funding from the South African government could form community care groups, whereas other community members working in the same area went without funding. I attended a Collective meeting where CHWs described how they attempted to work around the obstacles. My field notes record their exasperation:

Care workers are complaining about the NPO application process. One CHW says we have got people promising to help us, people are playing with us. Another care worker says in an outraged tone that they paid a prominent local politician [to help in getting an NPO number]—until now nothing has happened. Then the first speaker reports she has proof about NPO numbers. She stands up, marches to the desk where the facilitators are standing, and waves a pink slip that looked like a receipt in her hand. (It was a very dramatic moment—none of the participants had used the space like this). She flattened the pink note on the table, and I could see the facilitation team were struggling not to laugh. Seeing that nobody had reached for the note, she held it aloft and, turning to the facilitator, she said she paid ZAR 750 ($57) for an NPO number. She paused and then shaking with emotion says, “Even now we did not get anything.”

This event highlights the vulnerability of CHW projects to corruption. The NPO application process, designed to manage and certify care-work agencies, became a domain of political influence. The CHW who took a stand in protest at the meeting revealed that CHWs were paying politicians to fast-track their applications without getting the desired result. Recent concerns about irregularities in the NPO sector came to the fore when the Department of Social Development deregistered 36,000 NPOs for noncompliance with financial reporting requirements (Child, 2013). The vulnerabilities I witnessed in the NPO application system continued to intensify after my fieldwork ended. Furthermore, recent research has documented the fragility of community-based organizations and their reliance on local and international funding to remain sustainable (Akintola, Gwelo, Labonté, & Appadu, 2016)

In sum, weak regulatory frameworks resulting from health system fragmentation and design flaws in CHW projects created conditions where CHWs operated in communities without any formal supervision and a system for certifying NPOs that was subject to manipulation.

Adverse Patient Outcomes Resulting From Uncoordinated Care

A single patient case illustrates how competition among CHWs and limited regulation of CHWs operating in a health care environment prone to corruption may have serious implications for patient care. The patient was diagnosed with tuberculosis, but this diagnosis came too late to save her life. The details of this patient’s case, taken from my field notes, suggest that her death was preventable and was caused by deficits in the care provided by CHWs as well as from systemic inefficiencies in the primary health care system:

An African teenage woman, is bedridden and living at home. The house is shared by 4 adults: two of the adults, her brother and stepfather, both were unemployed and were home with the patient. The patient’s mother works away from the township and is seldom at home. Home-based-care worker 1 [introduced earlier in relation to moonlighting] took me to visit this patient, who she regularly visits as part of her work for Project 2. Home-based-care worker 1 is accompanied by a directly observed tuberculosis volunteer. Prior to approaching the house, home-based-care worker 1 indicates the patient had a strange rash over her body. On our arrival at the patient’s house at 9 a.m. on a winter’s morning, we are greeted by the patient’s brother, who invites us inside and ushers us into her bedroom. The patient’s face is covered in a rash, her lips are swollen and filled with sores. Her condition appears to be oral thrush.

There were several discrepancies about the quality of care this patient was receiving that I recorded in my notes. One day’s entry reads as follows:

The patient struggles to move, and when she speaks, she has a hacking cough. She asks after the other care workers, saying they had her documents. Home-based-care worker 1 turns to me, explaining that the patient had gone to the hospital about three weeks ago but as yet does not have her results. The letter from the hospital is with two other volunteers as well as the patient’s identity document. A short while later the two other volunteers arrive. The two ladies are well dressed, and home-based-care worker 1 explains they have newly become volunteers and are working freelance in this neighborhood. There is tension between the care workers. Home-based-care worker 1 confronts them about the documents, and the two ladies deny they had the patient’s documents; instead they say they have made an appointment for the patient to see a clinician described as “the rash doctor” in a week. The atmosphere in the room is tense. The visit winds down with the two volunteers saying that the patient would go to the rash doctor.

An entry from my notes taken on another day describes what I later learned had happened to this patient:

A few weeks later I return to work with home-based-care worker 1, covering a different section of her neighborhood. Home-based-care worker 1 hesitantly tells me that the patient passed away 2 weeks prior to this. The day after our visit, home-based-care worker 1 reports, and the patient’s mother returned home, alerted by her husband about our visit, and her mother took her to the clinic. The patient was diagnosed with TB, but her condition had deteriorated so badly she passed away shortly after that. The freelance volunteers continued denying having the patient’s papers.

The freelance CHWs’ control of the patient’s paperwork made me very concerned; at no time should CHWs have patient documents in their possession due to patient privacy rights. It appeared, based on Home-Based-Care Worker 1’s account that these volunteers were operating in the urban township with minimal supervision. At the time of my fieldwork, patients with tuberculosis were eligible for a government disability grant. The grant was a means-tested, sizable monthly benefit available to an individual with a physical or mental disability recognized by the state. This benefit was particularly vulnerable to corruption (City Press, 2009; Ndaliso, 2010). The circumstances of this patient’s case made me concerned that the volunteers already knew of her tuberculosis diagnosis and were possibly fraudulently applying for the disability grant on her behalf. In a context with limited regulation and registration of care work, it would be possible for members of the public to impersonate CHWs.

Blame for this tragic case did not rest solely with the freelance volunteers, however. Home-Based-Care Worker 1’s role as a trained CHW working for Project 2 was also problematic. Her training included appropriate patient referral, patient confidentiality, and respect. Yet, her actions suggested that she had little knowledge of her role and responsibilities despite the training she had received from the organization. In addition, communication between the different groups of CHWs had broken down, and they relayed contradictory information, which created obstacles to the patient getting a timely diagnosis that might have saved her life.

CHWs frequently complained about the limited respect they received from members of their community. I observed several occasions in the urban township when householders were reluctant to admit CHWs for a visit, saying they did not want to be disturbed. One CHW related an instance where the clinic encouraged her to visit a family with a person who was ill at home. When she arrived, however, the family denied there was a sick person in the house. This CHW interpreted their reluctance to let her in as a sign they felt ashamed to admit their kin had HIV. This indeed may have been the case, but their reluctance to work with a CHW may also have come from their negative perceptions about the quality of care the CHW would provide.

The patient’s death could have been prevented had she received prompt care. CHW programs are designed so that patients are referred by CHWs to appropriate health care providers and receive effective follow-up and monitoring services. None of these health services worked in this patient’s case, which was remarkable for its delayed diagnosis, improper handling of patient paperwork, and multiple CHW involvement. The lack of coordination among her competing CHWs left the patient to suffer from neglect. This reflects how structural factors such as a weak regulatory environment and poor coordination within the health system gave rise to local dynamics, wherein CHWs deviated from patient protocols to pursue their livelihood competitively and, in turn, compromised their commitment to patient care.

Discussion

In keeping with social science scholarship on community care work (Maes & Kalofonos, 2013; Swartz & Colvin, 2015), CHWs who participated in this study were, on the whole, deeply committed to their work, saw the consequences of ill health and poverty in their communities, and sought to bring relief to their patients. Their commitment to their work, however, was thwarted by structural factors and social processes in local villages and neighborhoods that led to unevenness in how CHW programs were implemented.

Planners working in international development characterize community interventions as complex due to their embeddedness within systems and subsystems (Funnell & Rogers, 2011). Such complexity is based on the “interactions and relationships between elements of systems and on the local rules that guide those interactions” (Rogers, Westhorp, & Walker, 2015, p. 393), producing variability in both the character and form of projects introduced into unique social contexts (Cohn, Clinch, Bunn, & Stronge, 2013). Complex systems have a high propensity to yield unexpected outcomes (Morell, 2010). This study identified three unintended consequences of CHW programs arising from structural and local grassroots processes: (a) CHWs moonlighting for multiple organizations, (b) CHWs freelancing in communities without regulation, and (c) adverse patient outcomes resulting from uncoordinated care.

With respect to the structural aspects of the health system and the organization of AIDS care, the projects discussed here operated in silos independently of each other and were not well integrated into the health care system. South Africa is like other African countries, in that global health interventions take the form of discrete projects focused on specific diseases such as AIDS or malaria. These projects tend to be funded and managed separately from the public health system (Biesma et al., 2009; Hafner & Shiffman, 2013; Storeng, 2014). Supplementary Table 1 lists the funding sources for the projects I studied. The high failure rate of CHW projects and the informal operations of CHW start-ups in communities added volatility and fragmentation to the health care system. South Africa is no exception and has an array of actors, including the Ministry of Health, international funders, and local NGOs, independently managing health projects (Lieberman, 2011). Public health research focusing on the sustainability of CHW projects has observed the challenges that discrete projects pose for strengthening health systems (Prince & Otieno, 2014; Rosenthal, 2016), creating a domain of care-work activities not regulated by the health care system.

With respect to local-level dynamics, social scientists have emphasized variability in how local communities interpret and enact policy guidelines and carry out standardized projects (Mol, 2002; Mykhalovskiy & Weir, 2004). Furthermore, the specificities of local settings, in their political dynamics, histories of oppression, and forms of authority, will influence the form and function of health projects (Adams, Burke, & Whitmarsh, 2014). This study identifies how CHW moonlighting and characteristics of corruption are shaped by the specificities of local conditions.

The decentralized nature of community care has implications for CHW moonlighting and for social relations within local settings (Campbell & Scott, 2011). Beginning with moonlighting, it is a form of economic diversification, in the context of economic and health uncertainties (Trinitapoli & Yeatman, 2011) and structural violence (Farmer, 2003) wrought by racial oppression under apartheid and health risks associated with the AIDS epidemic. However, moonlighting was more prevalent in the urban setting, as there were several community health projects operating in the same geographic locale, highlighting the importance of understanding how local dynamics influence project outcomes.

There is a growing international literature that examines the scope and consequences of moonlighting among health care workers (Jan et al., 2005; Muula & Maseko, 2006). The main drivers of moonlighting are linked to remuneration in the worker’s primary occupation. These studies have shown that moonlighting has serious consequences for health worker performance and patient safety. However, very little research in South Africa has studied moonlighting and its consequences for health workers. Rispel, Blaauw, Chirwa, and de Wet’s (2014) research on nursing in South Africa’s public health sector reported that 40% of public sector nurses surveyed were moonlighters. Furthermore, Rispel, Chirwa, and Blaauw (2014) showed that nurses who moonlighted were more likely to take off sick days when they were not sick and that they paid less attention while at their primary job. Anthropologists have documented the obstacles that low pay and limited resources create for CHWs (Maes, 2015; Takasugi & Lee, 2012); however, no studies on CHWs have documented how these organizational design features create conditions conducive to moonlighting. My study findings are the first to document dual jobs among CHWs in South Africa and to show that moonlighting within communities has detrimental consequences for the quality of CHW projects by duplicating health services, creating conflict and competition, and diluting CHWs’ commitment to their primary organization and work role.

Turning to power arrangements and the potential for corruption at the local level are critical domains where unintended consequences can occur (Djellouli & Quevedo-Gómez, 2015). From my fieldwork, the example of a CHW bribing a local politician to gain certification for her start-up illustrates how accountability mechanisms are vulnerable to systems of patronage. In addition, the case of the CHW who managed a food distribution project in addition to her primary work for Project 2 shows how discrete projects in communities can become malleable to other local agendas. These instances show the dangers of deemphasizing community embeddedness in CHW program design (Campbell & Scott, 2011; Schneider et al., 2008), as it erodes CHWs’ accountability to the broader community and their patients.

What is at stake in South Africa are detrimental consequences for patient care brought on by health system fragmentation and variability in local-level implementation. The evidence of possible adverse events or patient harms produced by CHW interventions is not well documented in the CHW evaluation literature (Lewin et al., 2010). In the patient case I described, a young woman died because she did not get timely and appropriate care. Her assigned CHW did not refer her to the clinic, her paperwork was handled improperly, and communication between CHWs from different organizations broke down. Her mismanaged care is one example of adverse events attributable to structural factors in the organization of health services delivery and the variability in project performance at the grassroots level.

Moreover, the patient was potentially a victim of corruption linked to the government disability grant. The National Planning Commission (2013) of South Africa has noted how accountability in the health care system is particularly weak. In their study of corruption in the health sector, Rispel, de Jager, and Fonn (2016) estimated, based on their analysis of provincial health expenditures from 2012 to 2013, that 6.3% of expenditures in South Africa were for irregular spending that did not comply with rules and regulations. The authors noted that this irregular expenditure was nearly double the estimates of irregularities in the U.S. health system. The recently reported Life Esidimeni tragedy, in which 144 mental patients died in torturous conditions as a result of corruption in the Gauteng provincial department of health (Nicholson, 2018), demonstrates the dire consequences for patients of corruption in the health care system.

One limitation to this study is that the data were collected before publicly funded ART programs were rolled out to scale starting in 2010. This means that my study reflects the social dimensions of AIDS when CHWs assisted families in applying for welfare benefits, made referrals between clinics and hospitals, and offered bereavement counseling. Despite this limitation, my research provides a detailed ethnographic characterization of how programmatic side effects degrade patient care. The ability of CHWs to deliver both technical and social forms of care is critical to the success of CHW interventions (Campbell & Scott, 2011). A recent study from Brazil describes how a once-successful ART program that was initially centrally controlled by the state in a strong partnership with NGOs (Frasca, Fauré, & Atlani-Duault, 2018) was decentralized to the detriment of patient care, thus reflecting the dangers of multiple stakeholders working in a fragmented health care model.

Conclusion

Efforts in South Africa to strengthen the health bureaucracy and redress racial health disparities are included in legislation for National Institutes of Health (NIH; Republic of South Africa, 2017). Currently in its pilot phase (Plagerson et al., 2019), the first step is to strengthen primary health care, described by the national Department of Health as primary health care reengineering (Republic of South Africa, 2017). CHWs are envisaged as playing a critical role in this policy reform (Schneider et al., 2015). This new proposal imagines teams of CHWs—managed by a nurse who will perform preventive primary health care that includes patient support for both infectious and chronic disease—under the auspices of the health ministry rather than run by discrete NGO or community projects. The proposal for reorganizing and formalizing CHW roles holds a great deal of promise for improving CHW livelihoods and their career longevity. However, my findings show that, for this new proposal to succeed, weaknesses in the health system must be addressed, including strengthening accountability throughout the health system, integrating discrete CHW projects, and introducing robust CHW supervision and mentorship.

CHW evaluation studies, heretofore, have focused primarily on positive outcomes of CHW roles and functions in health systems where health professionals are scarce. However, more research is needed to determine negative, unintended consequences of CHW programs for patients and the organization of the health system more broadly. Future research should extend work on the practice of moonlighting as a livelihood strategy to other contexts and its implications there for patient care. An additional line of inquiry into moonlighting might examine whether and how adequate remuneration and other material benefits limit this activity. Future research should also examine how the weak accountability of CHWs operating in their local communities with assumed, but unauthorized, power has detrimental consequences for patient care and the sustainability of health projects. These new lines of research engage conceptions of social structure and process in developing and measuring CHW program efficacy and quality of patient care.

Supplemental Material

Supplemental_Table_1 – Supplemental material for Unintended Consequences of Community Health Worker Programs in South Africa

Supplemental material, Supplemental_Table_1 for Unintended Consequences of Community Health Worker Programs in South Africa by Catherine van de Ruit in Qualitative Health Research

Footnotes

Acknowledgements

I am grateful to Cal Biruk, Charles Bosk, Sara Compion, Mary Dixon-Woods, Julie Szymczak, Lauren Wynn, and three anonymous reviewers for providing invaluable feedback at various stages in drafting this article. Thank you to Robert Brown for copy-editing an earlier version of this article. The research was supported by the National Science Foundation and the National Research Foundation of South Africa.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was supported by the National Science Foundation: Doctoral Improvement Grant (SES # 0728096) and the National Research Foundation of South Africa.

ORCID iD

Catherine van de Ruit

Supplemental Material

Supplemental Material for this article is available online at . Please enter the article’s DOI, located at the top right hand corner of this article in the search bar, and click on the file folder icon to view.

Author Biography

Catherine van de Ruit is a sociologist researching global health policy, and an assistant professor in the Department of Health and Exercise Physiology at Ursinus College.

References

Adams

Burke

N. J.

Whitmarsh

(2014). Slow research: Thoughts for a movement in global health. Medical Anthropology, 33, 179–197. doi:10.1080/01459740.2013.858335

Akintola

Gwelo

N. B.

Labonté

Appadu

(2016). The global financial crisis: Experiences of and implications for community-based organizations providing health and social services in South Africa. Critical Public Health, 26, 307–321. doi:10.1080/09581596.2015.1085959

Aronson

Neysmith

S. M.

(1996). “You’re not just in there to do the work”: Depersonalizing policies and the exploitation of home care workers’ labor. Gender and Society, 10, 59–77. doi:10.1177/089124396010001005

Auld

A. F.

Shiraishi

R. W.

Couto

Mbofana

Colborn

Alfredo

. . . Jobarteh

(2016). A decade of antiretroviral therapy scale-up in Mozambique: Evaluation of outcome trends and new models of service delivery among more than 300,000 patients enrolled during 2004–2013. Journal of Acquired Immune Deficiency Syndromes, 73(2), e11–e22. doi:10.1097/QAI.0000000000001137

Babb

Chorev

(2016). International organizations: Loose and tight coupling in the development regime. Studies in Comparative International Development, 51, 81–102. doi:10.1007/s12116-016-9217-7

Baynes

Semu

Baraka

Mushi

Ramsey

Kante

A. M.

Phillips

J. F.

(2017). An exploration of the feasibility, acceptability, and effectiveness of professional, multitasked community health workers in Tanzania. Global Public Health, 12, 1018–1032. doi:10.1080/17441692.2015.1080750

Bell

(2012). Cochrane reviews and the behavioral turn in evidence-based medicine. Health Sociology Review, 21, 313–321. doi:10.5172/hesr.2012.21.3.313

Biehl

J. G.

Petryna

(2013). Critical global health. In Biehl

J. G.

Petryna

(Eds.), When people come first: Critical studies in global health (pp. 1–20). Princeton, NJ: Princeton University Press.

Biesma

R. G.

Brugha

Harmer

Walsh

Spicer

Walt

(2009). The effects of global health initiatives on country health systems: A review of the evidence from HIV/AIDS control. Health Policy and Planning, 24, 239–252. doi:10.1093/heapol/czp025

10.

Blumer

(1986). Symbolic interactionism: Perspective and method. Berkeley: University of California Press.

11.

Boerma

Siyam

(2013). Health workforce indicators: Let’s get real. Bulletin of the World Health Organization, 91, Article 886. doi:10.2471/BLT.13.126656

12.

Bureau-Point

Phan

(2016). The structural violence of patient participation in Cambodian HIV services. Qualitative Health Research, 26, 1263–1274. doi:10.1177/1049732315583269

13.

Campbell

(2003). Letting them die: Why HIV/AIDS prevention programmes fail. Bloomington: Indiana University Press.

14.

Campbell

Scott

(2011). Retreat from Alma Ata? The WHO’s report on task shifting to community health workers for AIDS care in poor countries. Global Public Health, 6, 125–138. doi:10.1080/17441690903334232

15.

Charmaz

(2006). Constructing grounded theory: A practical guide through qualitative research. London: Sage.

16.

Child

(2013, January 28). Big back-track on NGOs. Times Live. Retrieved from https://www.timeslive.co.za/news/south-africa/2013-01-28-big-back-track-on-ngos/

17.

City Press. (2009, September 19). Departments clamp down on TB fraud. News24. Retrieved from https://www.news24.com/Archives/City-Press/Departments-clamp-down-on-TB-fraud-20150430

18.

Cohn

Clinch

Bunn

Stronge

(2013). Entangled complexity: Why complex interventions are just not complicated enough. Journal of Health Services Research & Policy, 18, 40–43. doi:10.1258/jhsrp.2012.012036

19.

Colvin

C. J.

(2015). Anthropologies in and of evidence making in global health research and policy. Medical Anthropology, 34, 99–105. doi:10.1080/01459740.2014.963196

20.

Cometto

Witter

(2013). Tackling health workforce challenges to universal health coverage: Setting targets and measuring progress. Bulletin of the World Health Organization, 91, 881–885. doi:10.2471/BLT.13.118810

21.

Daniels

Clarke

Ringsberg

K. C.

(2012). Developing lay health worker policy in South Africa: A qualitative study. Health Research Policy and Systems, 10, Article 8. doi:10.1186/1478-4505-10-8

22.

DelVecchio Good

Good

Grayman

. (2010). Complex engagements: Responding to violence in postconflict Aceh. In Fassin

Pandolfi

(Eds.), Contemporary states of emergency: The politics of military and humanitarian interventions (pp. 241–266). New York: Zone Books.

23.

Department of Health. (2008). Concept paper for the community health worker policy framework review: Deepening access to quality primary health care services in homes and communities. Pretoria: Department of Health.

24.

Department of Public Works. (2009). Expanded public works programme: Five-year report 2004/2005–2008/2009. Pretoria: Government Printers.

25.

De Wet

. (2011). Redefining volunteerism: The rhetoric of community home-based care in (the not so new) South Africa. Community Development Journal, 47, 111–125. doi:10.1093/cdj/bsq066

26.

Djellouli

Quevedo-Gómez

M. C.

(2015). Challenges to successful implementation of HIV and AIDS-related health policies in Cartagena, Colombia. Social Science & Medicine, 133, 36–44. doi:10.1016/j.socscimed.2015.03.048

27.

Draucker

C. B.

Martsolf

D. S.

Ross

Rusk

T. B.

(2007). Theoretical sampling and category development in grounded theory. Qualitative Health Research, 17, 1137–1148. doi:10.1177/1049732307308450

28.

Farmer

(2003). Pathologies of power: Health, human rights, and the new war on the poor. Berkeley: University of California Press.

29.

Fatti

Grimwood

Mothibi

Shea

(2011). The effect of patient load on antiretroviral treatment programmatic outcomes at primary health care facilities in South Africa: A multicohort study. Journal of Acquired Immune Deficiency Syndromes, 58(1), e17–e19. doi:10.1097/QAI.0b013e318229baab

30.

Frasca

Fauré

Atlani-Duault

(2018). Decentralisation of Brazil’s HIV/AIDS programme: Intended and unintended consequences. Global Public Health, 13, 1725–1736. doi:10.1080/17441692.2018.1455888

31.

Fried

Harris

Eyles

Moshabela

(2015). Acceptable care? Illness constructions, healthworlds, and accessible chronic treatment in South Africa. Qualitative Health Research, 25, 622–635. doi:10.1177/1049732315575315

32.

Funnell

S. C.

Rogers

P. J.

(2011). Purposeful program theory: Effective use of theories of change and logic models. San Francisco: Jossey-Bass.

33.

Galtung

(1969). Violence, peace, and peace research. Journal of Peace Research, 6, 167–191. doi:10.1177/002234336900600301

34.

Glaser

B. G.

Strauss

A. L.

(1967). The discovery of grounded theory: Strategies for qualitative research. New Brunswick, NJ: Aldine Transaction.

35.

Green

(2014). What kind of research does public health need? Critical Public Health, 24, 249–252. doi:10.1080/09581596.2014.917813

36.

Hafner

Shiffman

(2013). The emergence of global attention to health systems strengthening. Health Policy and Planning, 28, 41–50. doi:10.1093/heapol/czs023

37.

Hallgrimsdottir

H. K.

Teghtsoonian

Brown

(2008). Public policy, caring practices and gender in health care work. Revue Canadienne de Santé Publique/Canadian Journal of Public Health, 99(Suppl. 2), S43–S47. Retrieved from https://www-jstor-org-443.web.bisu.edu.cn/stable/41995018

38.

Harris

(2017). Achieving access: Professional movements and the politics of health universalism. Ithaca, NY: Cornell University Press.

39.

Jabeen

(2016). Do we really care about unintended outcomes? An analysis of evaluation theory and practice. Evaluation and Program Planning, 55, 144–154. doi:10.1016/j.evalprogplan.2015.12.010

40.

Jan

Bian

Jumpa

Meng

Nyazema

Prakongsai

Mills

(2005). Dual job holding by public sector health professionals in highly resource-constrained settings: Problem or solution? Bulletin of the World Health Organization, 83, 771–776. doi:10.1590/S0042-96862005001000014

41.

Johnson

L. F.

Dorrington

R. E.

Moolla

(2017). Progress towards the 2020 targets for HIV diagnosis and antiretroviral treatment in South Africa. Southern African Journal of HIV Medicine, 18(1). Retrieved from https://www.ajol.info/index.php/sajhivm/article/view/161463

42.

Joint United Nations Programme on HIV/AIDS. (2015). Progress report on the global plan towards the elimination of new HIV infections among children and keeping their mothers alive. Geneva: Author.

43.

Justice

(1987). The bureaucratic context of international health: A social scientist’s view. Social Science & Medicine, 25, 1303–1306. doi:10.1016/0277-9536(87)90128-6

44.

Kalofonos

(2014). “All they do is pray”: Community labour and the narrowing of “care” during Mozambique’s HIV scale-up. Global Public Health, 9(1–2), 7–24. doi:10.1080/17441692.2014.881527

45.

Kangovi

Grande

Trinh-Shevrin

(2015). From rhetoric to reality—Community health workers in post-reform U.S. health care. New England Journal of Medicine, 372, 2277–2279. doi:10.1056/NEJMp1502569

46.

Karim

S. S. A.

Churchyard

G. J.

Karim

Q. A.

Lawn

S. D.

(2009). HIV infection and tuberculosis in South Africa: An urgent need to escalate the public health response. The Lancet, 374, 921–933. doi:10.1016/S0140-6736(09)60916-8

47.

Kok

M. C.

Ormel

Broerse

J. E. W.

Kane

Namakhoma

Otiso

. . . Dieleman

(2017). Optimising the benefits of community health workers’ unique position between communities and the health sector: A comparative analysis of factors shaping relationships in four countries. Global Public Health, 12, 1404–1432. doi:10.1080/17441692.2016.1174722

48.

Leibbrandt

Woolard

McEwen

Koep

(2010). Trends in South African Income Distribution and Poverty since the Fall of Apartheid. OECD Social, Employment and Migration Working Papers, no. 101. Paris: OECD Publishing. doi: 10.1787/5kmms0t7p1ms-en

49.

Lewin

Munabi-Babigumira

Glenton

Daniels

Bosch-Capblanch

van Wyk

B. E.

. . . Zwarenstein

(2010). Lay health workers in primary and community health care for maternal and child health and the management of infectious diseases. Cochrane Database of Systematic Review, 2010(3), 1–211. doi:10.1002/14651858.CD004015.pub3

50.

Lieberman

E. S.

(2011). The perils of polycentric governance of infectious disease in South Africa. Social Science & Medicine, 73, 676–684. doi:10.1016/j.socscimed.2011.06.012

51.

Lincoln

Y. S.

Guba

E. G.

(1985). Naturalistic inquiry. London: Sage.

52.

Link

B. G.

Phelan

(1995). Social conditions as fundamental causes of disease. Journal of Health & Social Behavior, 35, 80–94. doi:10.2307/2626958

53.

Lund

(2010). Hierarchies of care work in South Africa: Nurses, social workers and home-based care workers. International Labour Review, 149, 495–509. doi:10.1111/j.1564-913X.2010.00100.x

54.

Maes

(2015). “Volunteers are not paid because they are priceless”: Community health worker capacities and values in an AIDS treatment intervention in urban Ethiopia. Medical Anthropology Quarterly, 29, 97–115. doi:10.1111/maq.12136

55.

Maes

Closser

Vorel

Tesfaye

(2015). Using community health workers: Discipline and hierarchy in Ethiopia’s women’s development army. Annals of Anthropological Practice, 39(1), 42–57. doi:10.1111/napa.12064

56.

Maes

Kalofonos

(2013). Becoming and remaining community health workers: Perspectives from Ethiopia and Mozambique. Social Science & Medicine, 87, 52–59. doi:10.1016/j.socscimed.2013.03.026

57.

Maes

K. C.

Kohrt

B. A.

Closser

(2010). Culture, status and context in community health worker pay: Pitfalls and opportunities for policy research. A commentary on Glenton et al. (2010). Social Science & Medicine, 71, 1375–1380. doi:10.1016/j.socscimed.2010.06.020

58.

Malkki

L. H.

(2015). The need to help: The domestic arts of international humanitarianism. Durham, NC: Duke University Press.

59.

Marten

McIntyre

Travassos

Shishkin

Longde

Reddy

Vega

(2014). An assessment of progress towards universal health coverage in Brazil, Russia, India, China, and South Africa (BRICS). The Lancet, 384, 2164–2171. doi:10.1016/S0140-6736(14)60075-1

60.

Mayosi

B. M.

Benatar

S. R.

(2014). Health and health care in South Africa—20 years after Mandela. New England Journal of Medicine, 371, 1344–1353. doi:10.1056/NEJMsr1405012

61.

Mayosi

B. M.

Lawn

J. E.

van Niekerk

Bradshaw

Abdool Karim

S. S.

Coovadia

H. M.

(2012). Health in South Africa: Changes and challenges since 2009. The Lancet, 380, 2029–2043. doi:10.1016/S0140-6736(12)61814-5

62.

McKay

(2017). Medicine in the meantime: The work of care in Mozambique. Durham, NC: Duke University Press.

63.

Merton

R. K.

(1936). The unanticipated consequences of purposive social action. American Sociological Review, 1, 894–904. doi:10.2307/2084615

64.

Meyer-Rath

Johnson

L. F.

Pillay

Blecher

Brennan

A. T.

Long

. . . Rosen

(2017). Changing the South African national antiretroviral therapy guidelines: The role of cost modelling. PLoS ONE, 12(10), e0186557. doi:10.1371/journal.pone.0186557

65.

Mol

(2002). The body multiple: Ontology in medical practice. Durham, NC: Duke University Press.

66.

Morell

J. A.

(2010). Evaluation in the face of uncertainty: Anticipating surprise and responding to the inevitable. London: Guilford Press.

67.

Mutevedzi

P. C.

Lessells

R. J.

Newell

(2013). Disengagement from care in a decentralised primary health care antiretroviral treatment programme: Cohort study in rural South Africa. Tropical Medicine & International Health, 18, 934–941. doi:10.1111/tmi.12135

68.

Muula

A. S.

Maseko

F. C.

(2006). How are health professionals earning their living in Malawi? BMC Health Services Research, 6, Article 97. doi:10.1186/1472-6963-6-97

69.

Mykhalovskiy

Weir

(2004). The problem of evidence-based medicine: Directions for social science. Social Science & Medicine, 59, 1059–1069. doi:10.1016/j.socscimed.2003.12.002

70.

Nading

A. M.

(2013). “Love isn’t there in your stomach”: A moral economy of medical citizenship among Nicaraguan community health workers. Medical Anthropology Quarterly, 27, 84–102. doi:10.1111/maq.12017

71.

Naidoo

Jacobson

J. S.

Neugut

A. I.

Dlova

N. C.

Mosam

(2016). HAART in hand: The change in Kaposi’s sarcoma presentation in KwaZulu-Natal, South Africa. South African Medical Journal, 106, 611–616. doi:10.7196/SAMJ.2016.v106i6.10333

72.

Naimoli

J. F.

Frymus

D. E.

Wuliji

Franco

L. M.

Newsome

M. H.

(2014). A community health worker “logic model”: Towards a theory of enhanced performance in low- and middle-income countries. Human Resources for Health, 12(1), Article 56. doi:10.1186/1478-4491-12-56

73.

National Planning Commission. (2013). National development plan 2030. South African Government. Retrieved from https://www.gov.za/issues/national-development-plan-2030

74.

Ndaliso

(2010, June 14). Five more SASSA officials suspended over disability grant fraud. The Witness. Retrieved from https://www.news24.com/Archives/Witness/Five-more-Sassa-officials-suspended-over-disability-grant-fraud-20150430

75.

Nglazi

M. D.

Lawn

S. D.

Kaplan

Kranzer

Orrell

Wood

Bekker

L. G.

(2011). Changes in programmatic outcomes during 7 years of scale-up at a community-based antiretroviral treatment service in South Africa. Journal of Acquired Immune Deficiency Syndromes, 56(1), e1–e8. doi:10.1097/QAI.0b013e3181ff0bdc

76.

Nicholson

(2018, July 18). Life Esidimeni: Compensation paid but justice delayed. Daily Maverick. Retrieved from https://www.dailymaverick.co.za/article/2018-07-18-life-esidimeni-compensation-paid-but-justice-delayed/

77.

Nugus

Désalliers

Morales

Graves

Evans

Macaulay

A. C.

(2018). Localizing global medicine: Challenges and opportunities in cervical screening in an indigenous community in Ecuador. Qualitative Health Research, 28, 800–812. doi:10.1177/1049732317742129

78.

Nxumalo

Goudge

Manderson

(2016). Community health workers, recipients’ experiences and constraints to care in South Africa—A pathway to trust. AIDS Care, 28(Suppl. 4), 61–71. doi:10.1080/09540121.2016.1195484

79.

Perry

H. B.

Zulliger

Rogers

M. M.

(2014). Community health workers in low-, middle-, and high-income countries: An overview of their history, recent evolution, and current effectiveness. Annual Review of Public Health, 35, 399–421. doi:10.1146/annurev-publhealth-032013-182354

80.

Pfeiffer

Nichter

(2008). What can critical medical anthropology contribute to global health? Medical Anthropology Quarterly, 22, 410–415. doi:10.1111/j.1548-1387.2008.00041.x

81.

Phelan

J. C.

Link

B. G.

Tehranifar

(2010). Social conditions as fundamental causes of health inequalities: Theory, evidence, and policy implications. Journal of Health and Social Behavior, 51, S28–S40. doi:10.1177/0022146510383498

82.

Pierson

(2005). The study of policy development. Journal of Policy History, 17(1), 34–51. doi:10.1353/jph.2005.0006

83.

Plagerson

Patel

Hochfeld

Ulriksen

M. S.

(2019). Social policy in South Africa: Navigating the route to social development. World Development, 113, 1–9. doi:10.1016/j.worlddev.2018.08.019

84.

Prince

R. J.

Otieno

(2014). In the shadowlands of global health: Observations from health workers in Kenya. Global Public Health, 9, 927–945. doi:10.1080/17441692.2014.941897

85.

Republic of South Africa. (2017). White paper 2017: National Health Insurance for South Africa; towards universal health coverage. Pretoria: National Department of Health.

86.

Rispel

(2016). Analysing the progress and fault lines of health sector transformation in South Africa. South African Health Review, 2016(1), 17–23.

87.

Rispel

L. C.

Blaauw

Chirwa

de Wet

(2014). Factors influencing agency nursing and moonlighting among nurses in South Africa. Global Health Action, 7, Article 23585. doi:10.3402/gha.v7.23585

88.

Rispel

L. C.

Chirwa

Blaauw

(2014). Does moonlighting influence South African nurses’ intention to leave their primary jobs? Global Health Action, 7, 25754. doi:10.3402/gha.v7.25754

89.

Rispel

L. C.

de Jager

Fonn

(2016). Exploring corruption in the South African health sector. Health Policy and Planning, 31, 239–249. doi:10.1093/heapol/czv047

90.

Rogers

Westhorp

Walker

(2015). Dealing with complexity in a realist synthesis: Community accountability and empowerment initiatives. In Bamberger

Vaessen

Raimondo

Rogers

Westhorp

Walker

(Eds.), Dealing with complexity in development evaluation: A practical approach (pp. 385–404). Thousand Oaks, CA: Sage.

91.

Rosenthal

(2016). “Doing the best we can”: Providing care in a Malawian antiretroviral clinic. Medical Anthropology, 35, 132–146. doi:10.1080/01459740.2015.1076409

92.

Rossi

P. H.

Lipsey

M. W.

Freeman

H. E.

(2003). Evaluation: A systematic approach. Thousand Oaks, CA: Sage.

93.

Rossi

P. H.

Williams

(1972). Evaluating social programs: Theory, practice, and politics. New York: Seminar Press.

94.

Schneider

Hlophe

van Rensburg

(2008). Community health workers and the response to HIV/AIDS in South Africa: Tensions and prospects. Health Policy and Planning, 23, 179–187. doi:10.1093/heapol/czn006

95.

Schneider

Lehmann

(2010). Lay health workers and HIV programmes: Implications for health systems. AIDS Care, 22, 60–67. doi:10.1080/09540120903483042

96.

Schneider

Nxumalo

(2017). Leadership and governance of community health worker programmes at scale: A cross case analysis of provincial implementation in South Africa. International Journal for Equity in Health, 16(1), Article 72. doi:10.1186/s12939-017-0565-3

97.

Schneider

Okello

Lehmann

(2016). The global pendulum swing towards community health workers in low- and middle-income countries: A scoping review of trends, geographical distribution and programmatic orientations, 2005 to 2014. Human Resources for Health, 14(1), Article 65. doi:10.1186/s12960-016-0163-2

98.

Schneider

Schaay

Dudley

Goliath

Qukula

(2015). The challenges of reshaping disease specific and care oriented community based services towards comprehensive goals: A situation appraisal in the Western Cape Province, South Africa. BMC Health Services Research, 15(1), Article 436. doi:10.1186/s12913-015-1109-4

99.

Singh

Sachs

J. D.

(2013). 1 million community health workers in sub-Saharan Africa by 2015. The Lancet, 382, 363–365. doi:10.1016/S0140-6736(12)62002-9

100.

Starks

Trinidad

S. B.

(2007). Choose your method: A comparison of phenomenology, discourse analysis, and grounded theory. Qualitative Health Research, 17, 1372–1380. doi:10.1177/1049732307307031

101.

Statistics South Africa. (2017). Mid-year population estimates (Statistical release P0302). Pretoria: Author. Retrieved from https://www.statssa.gov.za/publications/P0302/P03022014.pdf

102.

Storeng

K. T.

(2014). The GAVI Alliance and the “Gates approach” to health system strengthening. Global Public Health, 9, 865–879. doi:10.1080/17441692.2014.940362

103.

Swartz

Colvin

C. J.

(2015). “It’s in our veins”: Caring natures and material motivations of community health workers in contexts of economic marginalisation. Critical Public Health, 25, 139–152. doi:10.1080/09581596.2014.941281

104.

Swidler

Watkins

S. C.

(2009). “Teach a man to fish”: The sustainability doctrine and its social consequences. World Development, 37, 1182–1196. doi:10.1016/j.worlddev.2008.11.002

105.

Takasugi

Lee

A. C. K.

(2012). Why do community health workers volunteer? A qualitative study in Kenya. Public Health, 126, 839–845. doi:10.1016/j.puhe.2012.06.005

106.

Trinitapoli

Yeatman

(2011). Uncertainty and fertility in a generalized AIDS epidemic. American Sociological Review, 76, 935–954. doi:10.1177/0003122411427672

107.

van de Ruit

(2017a). Politics in the making of HIV/AIDS in South Africa. Contemporary Sociology, 46, 587–589. doi:10.1177/0094306117725085gg

108.

van de Ruit

(2017b). Social work professionalism during and after apartheid in South Africa. Sociology of Development, 3, 273–294. doi:10.1525/sod.2017.3.3.273

109.

van Rensburg

H. C

. (2014). South Africa’s protracted struggle for equal distribution and equitable access—Still not there. Human Resources for Health, 12(1), Article 26. doi:10.1186/1478-4491-12-26

110.

World Health Organization. (2007). Task shifting: Rational redistribution of tasks among health workforce teams: Global recommendations and guidelines. Geneva: Author.

111.

World Health Organization. (2012). WHO recommendations: Optimizing health worker roles to improve access to key maternal and newborn health interventions through task shifting. Geneva: Author.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.14 MB