The study uses a matched pair cluster-randomised controlled design to measure the impact of integrating community delivery of antiretroviral (ARV) medication using an existing cadre of community health workers on HIV patients’ adherence to and retention in ART care and on health outcomes in Dar es Salaam, Tanzania.
Studies have shown that integrating HIV services with other health services can have a substantial impact on increasing access to services and reducing loss due to a lack of follow up. This is of particular interest to the Tanzanian government. Individual barriers to adherence and retention included treatment-related misconceptions and misinformation, distance to antiretroviral therapy (ART) clinics, stigma, and financial burdens. The intervention used an existing community health worker cadre to deliver antiretroviral medication to patient homes or other location for stable patients and assessed its impact on adherence and retention.
- Is a differentiated ART care model (antiretroviral community delivery for patients who are clinically stable on ART and standard facility-based care for those who are not) non-inferior to facility care for all in preventing and treating viral failure?
- What is the intervention’s effect on the differentiated care model on patients’ healthcare expenditures?
The implementers measured the impact of integrating ARV community delivery into an existing community-based health programme on patient HIV viral suppression in Dar es Salaam, Tanzania. Two teams of health workers, home-based carers (HBCs) and community-based health care workers, conducted monthly home visits. These included promoting community engagement, providing HIV testing and counselling, ART adherence counselling and support, distributing antiretroviral drugs, and assisting in antiretroviral therapy patient enrolment and follow-up. Intervention health workers received a three-week pre-service training and a refresher course after four months of implementation.
Theory of change
The theory of change is grounded in the assumption that community health worker visits can have a substantial impact on health outcomes for people living with HIV. Visits by HBCs to provide support for adherence and retention, and to deliver antiretroviral medications, will keep patients in treatment, which will result in better adherence, retention, and health outcomes (viral suppression). First, HBC visits will serve as frequent reminders to patients to attend clinic appointments and adhere to ART. Second, home-based distribution of ARVs will increase access to treatment and reduce transportation barriers. Third, HBCs can act as reliable sources of health information while acting as a motivating factor. Finally, HBCs can verify adherence and follow up with patients that have missed appointments.
The evaluators matched health facilities and their catchment areas within three municipalities (Temeke, Kinondoni, Illala) in Dar es Salaam, and randomly assigned one of each pair to treatment (home delivery of ARVs to stable patients and facility care for those not stable) and one to control (facility-based care for all). Baseline and endline data were collected through surveys, administrative data, focus groups and key informant interviews.
They used intent to treat analysis, controls for baseline viral suppression, adjusts for patient characteristics (sex and age), as well as an estimating a complier average causal effect using an instrumental variable of treatment assignment. The evaluators also reported costs and provided an estimation of cost per participant. The unit of observation was the patient.
The authors concluded that the intervention of community delivery of ARVs through an existing community health worker programme was non-inferior to facility-based care.
10.9 per cent of control and 9.7 per cent of intervention were in viral failure at endline. Among those who were virally suppressed at baseline, 4.3 per cent and 4.6 per cent were in viral failure at study exit in the control and intervention arms, respectively. Among those who received ARV community delivery, 5.7 per cent were in viral failure at study close. However, among those who had received community delivery for at least 90 days, viral failure was 7 per cent. The relative risk for a two-sided analysis was 0.89, which was below the non-inferiority margin of 1.45. However, restricted to those with a suppressed viral load at baseline who received home delivery compared the relative risk confidence interval was above and the upper bound above the non-inferiority bound in all models.
While costs of clinic attendance were reduced due to fewer clinic visits, overall health expenditure was not significantly affected.
Implications for policy and practice
Home delivery may improve retention in ART care but some patients may skip annual appointments. Removing eligibility criteria of attending the catchment area clinic may increase patient eligibility, but it would either entail larger costs by requiring workers to travel further to reach patients, or be more logistically complicated by requiring greater communication between clinics.
Implications for further research
The study had a number of limitations. A lower than expected proportion of patients enrolled in the home delivery option and the study was underpowered to assess an effect for the subset of patients who enrolled. In addition, the average number of days enrolled in home delivery was only 226 days and more research is needed to determine longer-term safety of such a program.