Focus group discussions.

The data collectors facilitated nine focus group discussions with the CHWs recruited from the seven study sites. Every CHW who formed part of the CHW programme in these sites was invited to participate in the FGD. Where the CHWs exceeded the number suitable for a focus group discussion (7–10 participants), the researchers held two focus groups to accommodate the excess. There were no recorded refusals. Each FGD had approximately 10 participants, all women. Using a structured guide, we explored topics on the nature of their activities while out in the community, their experiences, successes and challenges while providing the services. Each of the discussions was audio-recorded, lasted approximately 2hrs, and was facilitated by the first author while data collectors helped take additional notes relating to the discussion.

Observations.

We designed observation templates and refined them using role-plays involving the data collectors. At each site, the data collectors observed a pair of CHWs for a period of 3 days. The CHWs were observed during household visits, the fieldworkers documented the services they provided and engagements with patients while in the field. Some of the observed visits were supervised by outreach team leaders (OTLs). However, in Ehlanzeni sites, the OTLs were not observed with CHWs, as the supervisors did not go to the households with the CHWs due to transport issues. The data collectors asked the household members for permission before carrying out the observations of the CHWs. There were no reported refusals to participate in the observations.

Interviews.

We conducted 17 semi-structured interviews with facility managers, OTLs and community representatives (CR). These participants were purposively selected based on their knowledge and experience of the CHW programmes in the different sites. We asked them about CHW duties in the households, their experience of CHW services, relationship with the healthcare system and community structures. We asked the OTLs and community representatives about the support they provide to the CHWs, the role they played in the community, successes and challenges of the programme. Each interview lasted approximately 45mins and were audio recorded with participant consent.

Furthermore, sixteen key informants (KI) drawn from Sedibeng, Johannesburg, Tshwane and Ehlanzeni districts were recruited using snow-ball sampling technique to participate in the study. The KIs occupied positions of family medicine practitioner, programme coordinator, nurse and academic/ researcher in the healthcare system, local municipalities and institutions of higher learning (Table 4). We asked questions about the origins of the programme, the practice of community-based health care, and the challenges of rolling out the programmes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. CHWs, supervisors and key informant sociodemographic data.

https://doi.org/10.1371/journal.pgph.0000881.t004

Sedibeng.

Due to increasing cases of chronic diseases and shortage of health workers to provide health care, a community-based health programme was introduced in 2010. Through the assistance of local gatekeepers (i.e., political leaders), health posts (HPs; temporary wooden structures) were introduced in one of the more densely populated areas of Sedibeng district. Nurses and lay health workers (CHWs), selected from local primary health facilities and NGOs respectively, were recruited to work on the programme. The CHWs received training from a team of family medicine specialists to undertake a community diagnosis, involving the collection of individual and household data (e.g., number of householders, income, access to social assistance, self-reported illness etc.), to identify health, social and economic needs.

An unpublished evaluation by members of Sedibeng health team indicated an increase in the number of identified TB, hypertension and child malnutrition cases (Table 5). However, with the introduction of the nation-wide CHW programme in 2011, the programme was experiencing challenges, as funds and other institutional resources were diverted to the nation-wide CHW programme.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Descriptions of the CHW programmes and the communities they serve.

https://doi.org/10.1371/journal.pgph.0000881.t005

At the time of the study, Team 1 provided services to an informal settlement, shacks made from corrugated iron and plastic, were home to local South Africans and immigrants (Table 5). Access to running water, sanitation and electricity was limited. The residents relied on water taps installed in central points of the community. The two CHW teams were operating from a health post and PHC facility and led by junior and retired senior nurses. The CHWs provided range of promotive and preventative services including delivering chronic medication to pensioners. The CHWs, using standardized forms, undertook household registration, contact tracing and made referrals to local PHC facilities. However, the CHWs stored the forms in their homes and the information wasn’t used to inform the type of care provided to the households.

Johannesburg.

In Johannesburg, community orientated care was introduced by family medicine practitioners in one community health centre prior to 2011. The practice was called Chiawelo Community Practice and operated within the community health centre. At the time of the study, it also functioned as one of the local universities teaching and research site. Similar to the Sedibeng team, there were community consultations to gain permission to work in the community. The catchment area was mapped, and community interviews carried out by CHWs to understand community needs. The team registered 22,000 households. All the registered householders were told about the community practice (e.g what it intend to achieve and how they will benefit). Patient care was provided by a medical doctor, clinical associate, nurse and a team of eleven CHWs recruited for the practice. The services were limited to the individuals registered under the community practice.

After several years, the initiative stalled as external funding used to set up the practice had come to an end. The CHWs were absorbed into the national CHW programme, resulting in contractual conflicts between the CHWs and programme leaders, as their integration into the national programme meant a reduction in their monthly stipend. At the time of the study, no evaluations had been published on the outcome of the practice; anecdotal evidence suggested an improvement in chronic care management.

This team worked in a diverse area with formal brick housing, hostels (originally built for working miners, but now occupied by families), as well as informal settlements. Communities with formal housing tended to have access to running water and electricity and were relatively affluent; the hostels and informal settlements were not (Table 5). High levels of crime made it difficult for the CHWs to adequately provide care. A CHW commented “Many criminals live in this hostel. We are always scared when walking in this area. It is worse because we have to enter the households and provide care” (CHW, FGD, Team 3). The CHWs preferred to visit their hostel clients in groups, however, this was not always possible; to reach their allocated number of patients they had to split up.

Tshwane.

COPC informed intervention was first introduced in Tshwane as part of local government health services, with a group family medicine practitioners and researchers leading the initiative. In 2011, the initiative became part of the nation-wide CHW programmes with external funding. Similar to the Sedibeng district, in Tshwane local schools and churches were used as health posts, particularly in areas where there were few PHC facilities. In rolling out programme, the CHWs mapped the community and undertook a community diagnosis. The CHWs were equipped with tablets with a specifically designed app to gather household and individual data. Data from CHWs’ tablets was aggregated by the supervisor and programme leaders for supervision and in-service training purposes.

The programme implementation relied on the CHWs employed by the Department of Health, as part of the nation-wide CHW programme. The CHWs had only been offered 10-day training by the Department and the majority had low literacy levels. The low stipend, lack of working tools and low literacy levels led to a conflict between the CHWs and the Department of Health, and meant the CHWs were often unreliable pool of community workers. Similar to the other health districts that had attempted to implement community-based health services in accordance with COPC principles, implementation of the programme weakened when external funding came to an end.

The two teams were providing services to residents in informal settlement, RDP (These are houses that have been built by the government and are given to low-income families, as part of the Reconstruction and Development Programme introduced in 1994) and formal brick housing. The informal settlements were without running water or electricity, and the majority of the residents were unemployed. In some part of the community, there was sewage in the streets making the area inaccessible by foot and motor vehicle. The CHWs prioritized providing care to patients located in the informal settlements, as many of the residents faced challenges in maintaining medication protocols for chronic conditions. Ensuring patients took their medication as prescribed required frequent visits (Field observation notes, Team 4).

Ehlanzeni.

The national CHW programme was introduced in the district in 2011. CHWs, who remained in the employment of the NGOs, were selected to work on the new programme (Table 5). Their monthly stipend was paid by the provincial Department of Health via the NGO. The two CHW teams were overseen by senior professional nurses based in the local PHC facilities; the CHW also reported to program managers attached to the NGOs. The CHWs undertook a combination of health promotion and preventative roles in the surrounding rural communities. While the CHWs had prior work experience in providing home-based care services, none of the CHWs had received CHW training prior to being inducted into the CHW programmes. The appointed supervisors were unable to provide field support to the CHWs due to lack of transport to accompany the CHWs while visiting households. The supervisors spent most of their time in the facility attending to patients.

In the communities, many households had an unreliable water supply. The majority of the residents didn’t have water boreholes and had to use wheelbarrows loaded with 20 liters containers to ferry water from households with boreholes. The area had poor roads, inaccessible by a motor vehicle (Table 5). Pensioners often struggled to visit the healthcare facility, as they could not walk or get a lift (Observation notes, Team 6). The CHWs complained about the hindrance of local traditional beliefs in their efforts to provide care. For example, during household visits, the researchers observed a diabetic pensioner with a wound on her foot refusing CHW care; she believed her neighbour was the cause of her poor health and misfortune (Observation notes, Team 6). As local women, the CHWs had to compromise and accommodate the patient beliefs.

A defined community and use of evidence.

The Tshwane, Sedibeng and Johannesburg teams gathered data to understand the local context and health needs. CHWs visited households to learn about their perceived needs. Key informant in Sedibeng commented: “The community participated by identifying its needs, the needs identified by the community were then moderated to come up with a list of urgent needs to be addressed” (KI, interview, Sedibeng). Undertaking a community diagnosis also allowed the implementing teams to explain the programme to the community and involve the community. The data generated through this process was collated and used to inform the human resources and service requirements for the intervention.

Across the different sites, there was evidence of continuous data collection to inform service delivery. In Tshwane district, as mentioned above, a mobile technology was installed in tablets given to CHWs to records householders’ health needs, care provided and schedule follow-up visits. However, due to limited funding, the use of mobile health technology could not be sustained, as lost and malfunctioning tablets were not replaced or repaired. A key informant commented: “unfortunately, if a technology device is used every day, those devices last for about 2 years. Therefore, you need to budget for a 2-year replacement cycle. Otherwise, it doesn’t really work” (KI, Interview, Tshwane). Key informants and CHWs mentioned the Department of Health did not replace or refurbish lost or damaged tablets. The CHWs who damaged or lost the tablets regressed to the traditional way of collecting and recording patient data on a piece of paper.

In the other districts, there was no use of mobile technology. Instead, household information was recorded on paper, often filed at the CHW homes, and not used to help inform future interventions in the community. However, some of the informants were concerned that under the national CHW programme the use of evidence to guide practice is diminishing. A key informant in Sedibeng commented: “….. WBOT asks you questions regarding your needs, but only God knows for what, because I haven’t seen anyone acting on the collected data… WBOT is designed not to have impact on the health of the people” (KI, Interview, Sedibeng).

Service integration.

To assess CHWs connection with the health system, we examined four features of the programmes—distance/ location from the clinic, supervision, relationships with health workers and referral systems.

Distance / Location. In the rural setting, many villages were located far from health facilities. To reduce the time spent walking between their homes, the health facility and the community, the CHW were advised to only report to the facility on Fridays to compile their weekly reports. On the other days of the week, they went straight from their homes to visit households: “They realized some of us come from far places, so coming to the clinic just to sign in and later go back to the households was impractical.” (CHW, FGD, Team 7). Though this change helped save the CHWs time, it reduced their access to supervision from the facility-based nurses/supervisors. In contrast, in urban sites, the catchment areas were conveniently located closer to the health facilities, and although supervisors lacked transport, they were able to walk or used their own cars.

Supervision. When CHW teams were led by senior nurses who provided thorough supervision, the CHW were more competent, likely to receive support from the facility-based staff members and be respected by the community. More skilled, senior supervisors discussed complex cases with CHWs (e.g., patients refusing care and not adhering to treatment) and where necessary accompanied the CHW to see the patient. However, this was not the case in the rural teams, where, although the supervisors were senior nurses, it was not possible for them to invest sufficient time and resources in the supervision of the CHWs. Some of the CHWs in Ehlanzeni were without a supervisor for a prolonged period during data collection. “we have not had a supervisor since October last year… when we have problems we tell the facility manager who makes time to assist us” (CHW, FGD, Team 6). During data collection, a supervisor was appointed and had to undergo an induction by the Department of Health. The other supervisor in Team 7 only saw the CHWs when they reported to the facility on Fridays; during the week she didn’t have a vehicle to visit households. The CHWs were allowed to call the supervisor to discuss issues they encountered. However, this was impractical as they were not provided with airtime and couldn’t afford buying airtime for work purposes. The contracting NGO was also not providing daily supervision to the CHWs.

Relationships with health care workers. The CHWs in Team 2, led by a senior supervisor, received sufficient support from facility-based staff members, as the supervisor negotiated for their inclusion in facility activities and borrowed equipment (e.g., blood pressure machines) on their behalf. In order to build a good relationship with facility staff, the supervisors in Johannesburg helped the nurses in the clinic: “to ensure the patients referred by my people (CHWs) are attended timeously, I make time to assist the nurses in the clinic. I sometimes assist X in the triage room or attend women who are visiting the clinic for family planning services” (CHW supervisor, interview, Team 3). However, support for CHWs planned activities was not forthcoming: “I usually go to the facility manager’s office and inform her that I am having a campaign based on HIV, TB, vitamin A or de-worming. She will be happy but there will be no assistance, even if I can ask for assistance from the EPI or TB side. They will say, you are independent, you must know these things” (CHW supervisor, interview, Team 5).

In the rural sites, when the CHW were at the clinic, they used a separate building, not the main clinic; this arrangement isolated the CHWs from the rest of the facility staff members. The team leaders also spent less time with the CHWs, as they only visited the facilities on Fridays.

In the urban sites, the CHWs had a complex work relationship with facility-based health workers. In one of the facilities, nurses used degrading names to refer to the CHW. A CHW commented: “they call us names like “mamoroto” [someone who is working with urine]… They always chase us around saying we must go to the side where patients’ urine is collected” (CHW, FGD, Team 5). This degrading name was conceived from the CHWs role in the collection of patients’ urine as part of their duties within the facility.

Referral systems. CHWs refer cases they cannot resolve at household level to health facilities for further assessment and care. A referral requires cooperation between the CHWs and facility staff (i.e., nurses providing feedback to the CHWs). However, due to CHWs “inferior” status, some nurses did not complete the back referrals, thus hindering CHWs from providing care post referral. A CHW commented: “we always have to beg to receive the report about the patients that we referred to the clinic”. Another CHW added: “getting the completed back referral form from the nurses is never easy, I always have to ask my supervisor to get the form from the clinic staff” (CHW, FGD, Team 1). These issues were more common in teams where the CHWs did not have a senior or present supervisor to mediate between them and facility staff.

Comprehensive and multi-disciplinary approach to care.

In exploring the features of comprehensive care and use of multidisciplinary teams to deliver care, the use of experts to provide advice and mentorship, CHWs limited knowledge and skills, and unmet health needs emerged as subthemes.

Expert advice and mentorship. In the urban sites, the CHWs received expert advice and mentorship from clinicians (e.g. medical doctors) and social services professionals in order to provide health promotion, prevention, screening services and referral for a wide range of health conditions. In Tshwane district, the CHWs met with members of a multi-disciplinary team weekly to debrief and receive case guidance. A key informant commented: “Last week, I attended a meeting in the X clinic where the CHWs and OTL had a review meeting. The multi-disciplinary team consisting of a dietician, doctors, nurses and so forth attended the meeting. They reviewed the difficult cases and advised the CHWs accordingly” (KI, interview, Tshwane). Similarly, the Sedibeng and Johannesburg teams had the support of medical doctors, nurses and clinical associates. The multi-disciplinary approach benefited the CHWs, as they had access to clinical knowledge shared during the meetings. In these meetings, the CHWs presented complex cases they dealt with in the field and received expert advice or mentorship.

Limited knowledge and skills. However, there were concerns that the CHWs did not possess the knowledge and skills required to implement support interventions aligned to COPC. A key informant in Tshwane district commented: “you cannot expect people who are at that level of training, who are not managed properly and without team leaders to function well”. To overcome this challenge, for example, the Johannesburg team used external funds to employ, train and retain CHWs with desired knowledge and skills mix. In Sedibeng, the CHWs experienced numerous challenges when offering services to those without residency status: “Working in this community is never easy for us. The CHWs do visit the householders to provide services, however some the householders do not have identity documents needed to register them for social services” (CR, interview, Team 1). The CHWs reported these cases to their senior team leaders who couldn’t assist as they did not go to the fields with them to offer onsite supervision.

Unmet health needs. The identification of unmet need also appeared to present challenges for the healthcare system. The Sedibeng team during household visits found suspected cases of cervical cancer, and those who needed cataract operations. However, care for such conditions was only available at distant, tertiary hospitals, and required numerous visits. The majority of the householders who received care from the CHWs were pensioners and unemployed women with dependents. They did not have the funds to visit the facilities. As a result, some of the participants deemed the attempt at COPC informed intervention as ineffective, as the identification of community health needs was not matched with available care.

Community participation.

One of the key principles of COPC is for health programme developers to engage communities and ensure the services to be provided are relevant to the needs of the community. It is important that there are open channels of communication such that if there is a misunderstanding, issues can be clarified.

Setting up programmes. Each programme consulted local stakeholders such as political and traditional leaders, school managers and local NGOs. In Tshwane, the team worked with local stakeholders to secure space in their schools and churches to be used as health posts. A participant commented: “After we have introduced and involved the community in our work, implementation went well. The ward councillor even helped us to find a suitable space to use as a health post” (KI, interview, Tshwane). Another participant added: “Since we informed them about our programme. We were protected, nothing bad has happened to us because the community know us. If I complain that somebody snatched my phone, they will make sure that they will get whoever took it” (KI, interview, Tshwane). In Johannesburg at the start of the programme, there was resistance from patients and community, but continuous engagement led by a community liaison officer who was employed to support the programme helped ease the tension between the health team and community. Similarly, in Sedibeng (Team 1), local leaders upon realizing patients visiting the health post queued in the sun or cold weather, volunteered their time to build a temporary shelter for the patients to use while waiting for their turn.

Maintaining on-going relationships with service users. In Sedibeng, when the CHW team 1 struggled to persuade patients to visit the local health facility for further health assessment and treatment, community leaders made time to meet and listen to the CHWs problems (e.g., patients refusing CHWs entry into their households), and addressed the issues with the householders. In the different sites, the CHWs also struggled to access some households to provide care due to the stigma associated with their services. Due to their previous role during the height of the HIV pandemic, their service was associated with providing HIV and TB care. A CHW commented: “..the clinic sometimes send us to clients who are defaulting on their medication, but because they [clients] don’t want to be seen being visited by us, they chase us away” (CHW, FGD, Team 3).

Engagement with the broader community. In Ehlanzeni (Team 6), a community representative who also served as a member of the clinic committee ensured the clinic was always represented during community meetings she addressed. This leader’s initiative helped ensure service delivery and health concerns of the residents were discussed in the same platform. However, community members were only interested in topics such as access to employment, water and decent housing. Many of the communities were politically volatile, with residents were regularly protesting for decent housing and running water. In Sedibeng district, a health post was burned down by members of the community protesting against continuous electricity cuts in the area, while in another unrelated event district health officials were held hostage by CHWs protesting against poor work conditions (e.g. low stipend). These on-going service delivery protests often disrupted the engagements between the health team and community: “We often have to stop work, because of service delivery protests” (KI, Tshwane).