University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Task shifting of antiretroviral delivery from health care workers to persons living with HIV/AIDS: clinical outcomes of a community-based program in Kenya.
Global Health Sciences Literature Digest
Published May 12, 2011
Journal Article

Selke HM, Kimaiyo S, Sidle JE, Vedanthan R, Tierney WM, Shen C, Denski CD, Katschke AR, Wools-Kaloustian K. .Task shifting of antiretroviral delivery from health care workers to persons living with HIV/AIDS: clinical outcomes of a community-based program in Kenya. J Acquir Immune Defic Syndr. 2010 Dec 1;55(4):483-90.

In Context

Task shifting has been advocated as a strategy to address the shortage of health care workers that impedes the scale-up of antiretroviral therapy (ART) in resource-limited settings. WHO recommends task shifting from physicians to non-physician clinicians, nurses, and community health workers, including persons living with HIV (PLHIV), to provide HIV services in both the clinic and the community.(1) The evidence to support the effectiveness of task shifting, however, is limited.(2) Mobile health technology may be an effective tool as a component of an overall task shifting strategy.


To assess the impact on clinical outcomes of patients enrolled in an HIV care delivery system which utilized PLHIV as Community Care Coordinators (CCCs), aided by an electronic decision support tool, to deliver medications and provide follow-up care to patients on ART in the community.


HIV clinic in a rural health center and the surrounding community in Western Kenya


HIV-infected adults clinically stable on ART

Study Design

Prospective cluster randomized controlled trial


Primary study outcomes were 1) adherence to drugs and to clinic visits, and 2) clinical outcomes, specifically viral load, intercurrent opportunistic infections, hospitalization, loss to follow-up, change to second-line therapy and mortality. Additional outcomes included pregnancy, mean CD4 count and decline in Karnofsky score.


Details of the development of the community-based ART delivery model, the CCC curriculum, the PDA programming and the onsite mentoring program have been published in an earlier paper.(3)

From March 2006 to April 2008, HIV-infected patients enrolled at the Mosoriot HIV clinic who were at least 18 years old, clinically stable on ART for a minimum of three months, demonstrated good adherence to treatment, had household members who were aware of the patient’s HIV status, lived in Kosaira Division served by the clinic, and were willing to participate were eligible for enrollment. Patients were excluded if they had an active WHO stage 3 or 4 condition, were pregnant (by patient self-report), had been hospitalized in the previous three months, or were unable to understand the informed consent process due to incapacity. The Mosoriot HIV clinic serves the Kosaira Division, which is composed of 24 sub-locations serving a community of 60,000. There were approximately 4,000 HIV-infected adults, over half of whom were on treatment. Each sub location could be managed by a CCC and as such were the unit of randomization.

The intervention consisted of monthly home assessments by community care providers using a Personal Digital Assistant (PDA) pre-programmed to provide decision support. Clinic visits were scheduled every three months. During the monthly home visits, CCCs obtained and entered data into their PDAs regarding the patients symptoms, vital signs and adherence assessment. Pre-programmed alerts were triggered if specified parameters were met. These alerts included reminding the CCC to return to evaluate the patient the next day, transporting the patient to the clinic for urgent evaluation or calling the clinical officer for consultation. They also dispensed a one-month supply of medications from a prefilled kit. The community care coordinators were selected from patients in the HIV clinic, had a secondary education, were clinically stable, and considered by clinic staff to be good role models and mentors. They received training and clinical mentoring.

Standard of care consisted of monthly clinic visits. Clinic visits included triage and vital signs obtained by a nurse, medical history, acute concerns, medication issues and adherence, prescriptions for ART and opportunistic infection (OI) prophylaxis conducted by a clinical officer or physician (10% of visits), and provision of a one-month supply of medications from the pharmacy. This model requires contact with a minimum of three health care providers as well as transportation and wait times.

Data were collected through medical chart review and adherence assessment, and HIV viral load and CD4 were obtained at enrollment, six and 12 months. Analysis was intent to treat.


A total of 208 patients were enrolled, 96 in the intervention group and 112 in the control group. There were no significant differences in baseline clinical status between the two groups. At the 12-month endpoint there were similar losses to follow-up (5.2% vs. 4.5%), with 87 patients in the intervention group and 102 patients in the control group with outcome data.

The intervention group had significantly fewer clinic visits (6.2) than the control group (12.4). Despite fewer clinic visits, clinical outcomes for the two groups were not significantly different. At 12 months, there were no significant differences between the intervention and control groups in terms of new WHO stage 3 or 4 events, rate of OIs, complete change in ARV regimen, decline in Karnofsky score, pregnancy incidence, mean CD4 cell count, or detectable viral load. Both groups had high self-reported levels of medication adherence. The CCC group undertook 64% more acute clinic visits than originally scheduled due to the clinic referral triggered by the programmed alerts for identification of acute medical and social issues during home evaluations.


Community-based care by PLHIV supported by mobile computer-based decision support resulted in similar clinical outcomes as standard care with half the number of clinic visits.

Study Quality

Limitations included small sample size. The authors note they had conducted sample size calculations for a power of >95%. However, despite smaller enrollment they retained >80% power to detect projected differences in adherence rates between the groups. This did not allow an evaluation of non-inferiority of the community-based model compared to the clinic model. In addition, as adherence was one of the eligibility criteria, the evaluation could not measure the impact of the intervention on adherence. Overall quality of this pilot study was otherwise good, though outcome assessment was not blinded and method of allocation was unclear.

Programmatic Implications

This pilot study supports the potential to use PLHIV as part of an HIV care model in the community. Mobile health technologies show promise for assisting in the expansion of ART in resource-limited settings with limited human resources. However, larger scale studies will be needed to confirm the findings.


  1. WHO Library Cataloguing-in-Publication Data. Task shifting: rational redistribution of tasks among health workforce teams, global recommendations and guidelines. NLM classification: WC 503.6
  2. Jaffar S, Amuron B, Foster S, Birungi J, Levin J, Namara G, Nabiryo C, Ndembi N, Kyomuhangi R, Opio A, Bunnell R, Tappero JW, Mermin J, Coutinho A, Grosskurth H; Jinja Trial Team. Rates of virological failure in patients treated in a home-based versus a facility-based HIV-care model in Jinja, southeast Uganda: a cluster-randomised equivalence trial. Lancet. 2009;374:2080-2089
  3. Wools-Kaloustian KK, Sidle JE, Selke HM, Vedanthan R, Kemboi EK, Boit LJ, Jebet VT, Carroll AE, Tierney WM, Kimaiyo S. A model for extending antiretroviral care beyond the rural health centre. J Int AIDS Soc. 2009 Sep 29;12(1):22.