Assessment of a Pilot Antiretroviral Drug Therapy Programme in Uganda: Patients' Response, Survival, and Drug Resistance

Major challenges exist in the development of infrastructure needed to ensure uninterrupted drug stocks and the appropriate, safe, and effective use of HIV drugs in resource-poor countries. The UNAIDS HIV Drug Access Initiative (DAI) was a pilot program developed in 1997 to increase access to AIDS care and drugs in Uganda, Côte d'Ivoire, Chile, and Vietnam. The program began in June 1998. The initiative established national treatment guidelines, developed information materials, and trained and educated health-care providers on AIDS care. Patients and their families were responsible for paying for all their medical care, drugs, and laboratory tests. The US Centers for Disease Control and Prevention (CDC), in collaboration with the Ministry of Health and UNAIDS, assessed the virological and immunological response to treatment, sustainability of treatment, survival, and emergence of drug resistance among patients using the DAI.

Five health-care facilities were accredited to provide antiretroviral drug treatment through the initiative; all were in or near the capital city of Kampala. The experience and capacity of the participating centers varied greatly, although all represented the highest level of medical care available in Uganda.

The 2-year pilot phase of the antiretroviral component of the DAI was Aug 1, 1998 to July 31, 2000. The physician and patient chose the antiretrovirals, frequency of follow-up visits, and laboratory monitoring on the basis of the patient's clinical status, availability and cost of drugs, personal finances, and laboratory test results.

In August 1998, four nucleoside reverse-transcriptase inhibitors (NRTIs) and one protease inhibitor were available through a private drug company. Drug availability increased over time; by February 2000, there were six NRTIs, four protease inhibitors, and three non-nucleoside reverse-transcriptase inhibitors (NNRTIs). During the pilot period, the approximate 30-day cost of therapy was US$214-406 for dual nucleoside reverse transcriptase inhibitors (2NRTI), US$440-660 for HAART with an NNRTI, and US$531-708 for HAART including a protease inhibitor. Patients generally purchased a 1- to 4-week supply of antiretroviral drugs at a time, depending on their finances.

Physicians recorded clinical information on standardized medical record forms. Adherence to antiretroviral therapy was recorded by the physician for the time since the last clinic visit as "about as prescribed," "less often than prescribed," "more often than prescribed," or "not at all." Starting on June 7, 1999, the Uganda Virus Research Institute/CDC-Uganda began to provide free testing for viral load and CD4 cell count for patients in the DAI. Each patient's status was determined from clinical records and staff reports. Phenotypic resistance was judged to be present as determined by biologically based, drug-specific cut-offs. The assessment of immunological and virological response to antiretroviral therapy was restricted to patients 13 years of age or older who had not received previous antiretroviral therapy and who had baseline measurements of CD4 cell count and viral load. Besides antiretroviral regimen, baseline CD4 cell count and duration of therapy were the only significant independent variables included in these models.

In total, 912 patients were enrolled in the DAI at the five centers. Of the 399 patients with documented prescribed antiretroviral medications at the initial visit, 204 (51 percent) were prescribed HAART, 189 (47 percent) 2NRTI, and six (2 percent) NRTI monotherapy. There were no differences between those started on HAART and those started on 2NRTI with regard to CD4 cell count, viral load, age, proportion of antiretroviral-naive patients, or proportion of women.

Of 399 patients only 111 had an initial visit recorded, but for 37 of 111 (33 percent) the initial visit was 91 days or less before the end of the pilot period and these patients were still judged to be in care. The remaining 288 (72 percent) had 1,307 follow-up visits (median 4 per patient). The median time between visits was 32 days. The median time of observation for patients receiving HAART was 94 days, and for 2NRTI was 152 days. Among those who started HAART, 14 (7 percent) were changed to 2NRTI at a median of 107 days, and for those who started on 2NRTI, 45 (24 percent were changed to HAART at a median of 173 days.

Data on self-reported adherence were collected from 221 patients during 967 patient visits. Since the time of the last visit, patients reported taking antiretroviral drugs "about as prescribed" in 850 (88 percent) instances, "less often than prescribed" in 79 (8 percent) incidences, "not at all" in 37 (4 percent) incidences, and "more often than prescribed" in one (<1 percent) instance. By the end of the pilot period, 74 (16 percent) of the 476 patients were known to have died. Of these, 42 (57 percent) died within 3 months of accessing the DAI. The cause of death was reported to be AIDS-related for 66 (89 percent). Patients with a CD4 cell count of less than 50 cells/µL at the start of treatment had a higher hazard of death than those with at least 50 cells/µL. There were no detected differences for men versus women, antiretroviral experienced versus treatment-naive patients, and age 35 years or younger versus older than 35 years. The effect of use of 2NRTI versus HAART and viral of load on survival was not significant in proportional hazards models after adjustment for CD4 cell count.

In general, the authors maintained, "the pilot program showed that, through modest increases of existing resources, an effective system for drug procurement, distribution, and accountability could be implemented and maintained. This accomplishment led to an uninterrupted supply of drugs that supported sustainable management of patients, despite the often-stated financial, logistical, and technical impediments to treatment access."

Back to other CDC news for July 18, 2002

Previous Updates | Search the CDC archive