Background

The chronic shortage of health care workers is recognized as one of the major bottlenecks to health care provision, and scaling up treatment is no exception [1]. The impact is most devastating in rural areas, where the human-resource crisis is most acute [2]. For the past 3 years, MÉdecins Sans FrontiÈres (MSF) has been supporting a program to provide care and treatment for people with HIV/AIDS in the local service area of Lusikisiki, one of the poorest and most densely populated rural areas of South Africa.

The Lusikisiki subdistrict comprises 150,000 inhabitants serviced by 1 hospital and 12 clinics. The HIV infection prevalence is high, with almost one-third (31%) of women who present at antenatal care clinics testing positive. A lack of staff within the health system is a major problem. With just 5 physicians per 100,000 people, Lusikisiki's physician-to-patient ratio is 14 times lower than the national average [3]. Approximately one-half of all nursing posts remain vacant. In addition, a chronic lack of auxiliary staff adds to nurses' workloads. In this article, we describe how the integration of HIV care and treatment into primary health care in Lusikisiki overcame some of the challenges of working in a resource-limited rural area, to achieve good treatment coverage and clinical outcomes [4].

Approaches to Support Clinic-based Care

The World Health Organization promotes the role of primary health care and community-led care in the delivery of antiretroviral therapy (ART) in resource- limited settings [5]. In keeping with these principles, the delivery of HIV services in Lusikisiki was achieved through decentralization to primary health care, task shifting within services, and strong community support.

Decentralization and task shifting. The provision of treatment at the clinic level inevitably resulted in a significant increase in the number of service users within a system that was already chronically understaffed and poorly equipped. Although overall utilization of clinic services increased almost 2-fold since the start of the program (from 16,465 service users in April 2004 to 28,191 service users in April 2006), the number of professional nurses (30) did not increase. This near doubling of the workload would have been impossible to manage without task shifting.

With appropriate training, mentoring, and supervision, it was possible to delegate the running of the ART program to primary health care nurses and community health workers. Table 1 provides a description of some of the changes in the roles of clinic and support personnel in the Lusikisiki program. As much responsibility as possible was delegated to lower-level health care workers while ensuring that professional medical oversight was provided to maintain quality control.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1.

Enrollment into antiretroviral therapy (ART) programs at clinics, compared with the hospital. The graph shows the no. of patients newly enrolled into ART programs for each quarter (i.e., nos. are not cumulative).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1.

Traditional roles of health staff in HIV/AIDS care, compared with roles of health staff in the Lusikisiki program.

Training and mentoring through mobile teams. All clinics received regular physician support via a mobile team visit to support the overall health services. In addition, these mobile teams provided training and mentoring of nurses in HIV management, including prevention of mother-to-child transmission, management of opportunistic infections, and ART. This was reinforced through on-the-job training, which also included patient management skills, such as examination, history taking, and counseling. Tuberculosis (TB) is the major cause of death among people with HIV/AIDS in South Africa; according to Statistics South Africa (the government statistical agency) [6], >50,000 people died of TB in 2001 alone. In the Lusikisiki program, 82.5% of patients with TB are found to be HIV positive; therefore, a particular emphasis was placed on developing a high suspicion for TB and making a clinical diagnosis when smear results are negative or extrapulmonary TB is present.

Systems improvement and quality control were overseen by the mobile team (comprising 1 physician and 1 nurse), using a program-evaluation tool that looks at specific outcomes of the different components of HIV care on a quarterly basis. The whole clinic team was brought together to identify the strengths and weaknesses and to decide on priorities for the following quarter.

Creating new capacity. The role of adherence counselors extended far beyond counseling. With training and mentoring, these auxiliary/lay workers were able to support many of the key processes required to run a clinic-based HIV service, including service-user support, treatment preparedness, facilitation of support groups, and arrangement of follow-up visits, as well as teaching people receiving ART to package pillboxes, addressing problems in adherence, and collecting and collating statistics. High commitment was maintained by means of weekly meetings and workshops (and, it should be acknowledged, by the tenacious commitment of key individuals). Adherence counselors also worked with other community actors: volunteer workers (community caregivers), other support groups, adherence and clinic committees, and treatment activists. They also undertook all aspects of voluntary counseling and testing, with nurse supervision. Previously, this was solely the responsibility of the nurses, but, with this increased capacity, the number of people tested has increased >4-fold in the past 3 years (4874 tests performed in 2002 vs. 18,809 performed in 2005).

Engaging the community in HIV/AIDS care is a proven way to enhance program quality, in terms of clinical outcomes, adherence rates, and retention [7]. In Lusikisiki, the community interacted with HIV services in a number of ways. General support groups provided peer support for disclosure and testing and performed home visits when problems were identified. ART support groups prepared people for treatment, provided support for adherence and managing adverse effects, and traced and supported individuals who defaulted. A clinic committee represented service users in the case of complaints, advocated for better infrastructure and drug supply, and monitored HIV program and condom distribution in the community. An adherence committee followed up with nonadherent patients and served as arbitrator if a clinic team could not decide on the readiness of a person for ART. Finally, individual service users provided important support to other members of the community through learning about HIV and sharing their experiences.

Initially, some nurses doubted the capacity of community health workers, but their participation in clinical discussions and patient management was encouraged by the supervising physician. Eventually, nurses began to appreciate the benefits of having some of their workload shared by community health workers.

Outcomes of Integrating HIV Services into Clinic Care

The approach taken in Lusikisiki has allowed for rapid scale-up of treatment coverage in a short time with good outcomes. A cohort analysis of people who have been receiving treatment for >12 months shows satisfactory immunological recovery and viral suppression (table 2). The data presented allow comparison of outcomes in the hospital and the clinics. Hospital-based ART is generally promoted in South Africa, as in many countries, with a progressive down-referral of patients to the clinics in some areas. Our aim was to show that satisfactory outcomes can be achieved when treatment is initiated at the clinic level. Routine program monitoring showed no difference between the sex ratio among patients attending the clinics and that among patients attending the hospital. The proportion of patients enrolling who had CD4 cell counts ≪50 cells/mm³ was 19.2% (95% confidence interval [CI], 17.1%–21.3%) in the clinics and 26.3% (95% CI, 23.1%-29.0%) in the hospital (P = .0002). This indicates that the patients enrolling in the hospital had disease that was at a more advanced stage.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2.

One-year outcomes at clinics and the hospital in Lusikisiki, South Africa.

The greater proximity and acceptability of services at the clinic level has led to faster enrollment of people receiving treatment and better patient retention. Only 2% of people were lost to follow-up in the clinics, compared with 19% at the hospital. This higher dropout rate at the hospital may be explained by several factors: a higher early mortality, people having to travel farther, less preparation of ART recipients, and less-effective follow-up of individuals who default (adherence counselors are not employed at the hospital). There is no statistical difference between the recorded mortality rates in the hospital and the clinics; however, the percentage of individuals remaining in care was lower in the hospital (67%) than in the clinics (81%). Although the mortality rate among those lost to follow-up cannot be known, it is expected to be high.

Enrollment initially increased at a similar pace at the clinics and hospital. After 1 year, enrollment at the hospital reached a plateau and then began to decline, suggesting a saturation of services. In contrast, enrollment in the clinics continued to increase (figure 1). This is likely a result of clinics offering multiple service points and clinic services being integrated into general consultations.

By mid-2006, there were 2200 people receiving ART in Lusikisiki. According to modeling done for 2005, the program had achieved universal coverage for the subdistrict [4, 8]. At this level of coverage, people arriving at the clinics with HIV/AIDS were far less sick than previously noted. In the inception phases of the program, many people were very ill. This “catch-up of the backlog” is reflected in the statistics: in early 2004, 50% of service users at the hospital and 40% of those at clinics arrived with CD4 cell counts ≪50 cells/mm³; by the end of 2005, the number of patients with CD4 cell counts ≪50 cells/mm³ had decreased to 16% at both the hospital and clinics. Because people were arriving with a better immune status, clinical management was less time consuming, so that more patients could be seen.