Scaling up of highly active antiretroviral therapy in a rural district of Malawi: an effectiveness assessment

Methods We scaled up and simplifi ed HAART in this programme since August, 2002. We analysed survival indicators, CD4 count evolution, virological response, and adherence to treatment. We included adults who all started HAART 6 months or more before the analysis. HIV-1 RNA plasma viral load and self-reported adherence were assessed on a subsample of patients, and antiretroviral resistance mutations were analysed in plasma with viral loads greater than 1000 copies per mL. Analysis was by intention to treat. Findings Of the 1308 patients who were eligible, 827 (64%) were female, the median age was 34·9 years (IQR 29·9-41·0), and 1023 (78%) received d4T/3TC/NVP (stavudine, lamivudine, and nevirapine) as a fi xed-dose combination. At baseline, 1266 individuals (97%) were HAART-naive, 357 (27%) were at WHO stage IV, 311 (33%) had a body-mass index of less than 18·5 kg/m², and 208 (21%) had a CD4 count lower than 50 cells per μL. At follow- up (median 8·3 months, IQR 5·5-13·1), 967 (74%) were still on HAART, 243 (19%) had died, 91 (7%) were lost to follow-up, and seven (0·5%) discontinued treatment. Low body-mass index, WHO stage IV, male sex, and baseline CD4 count lower than 50 cells per μL were independent determinants of death in the fi rst 6 months. At 12 months, the probability of individuals still in care was 0·76 (95% CI 0·73-0·78) and the median CD4 gain was 165 (IQR 67-259) cells per μL. In the cross-sectional survey (n=398), 334 (84%) had a viral load of less than 400 copies per mL. Of several indicators measuring adherence, self-reported poor adherence (<80%) in the past 4 days was the best predictor of detectable viral load (odds ratio 5·4, 95% CI 1·9-15·6).