Efficacy of antiretroviral therapy programs in resource-poor settings: A meta-analysis of the published literature

Background. Despite the advent of effective combination antiretroviral drug therapy (ART) for the treatment of human immunodeficiency virus (HIV) infection, many doubt the feasibility of ART treatment programs in resource-poor settings. We performed a meta-analysis of the efficacy of ART programs in the developing world. We searched the Medline database with the index terms HIV, antiretroviral therapy, CD4 count, viral load, experience, and outcomes. A total of 201 abstracts were reviewed, and 25 articles were selected for detailed review. Ten observational studies with details on patient outcomes were ultimately included in the analysis. Methods. Three readers independently extracted data from the articles. The details recorded included patient demographic characteristics, baseline CD4 cell counts, baseline HIV RNA viral loads, ART histories, outcomes, and timing of the outcome measure. Results. The proportion of subjects with an undetectable HIV viral load provided the measure of treatment efficacy. A random-effects model weighted the proportion of patients with undetectable viral load at various time points during ART. The proportion was 0.697 (95% CI, 0.582 - 0.812) at month 6 and 0.573 (95% CI, 0.432 - 0.715) at month 12 of ART. The provision of medications free of charge to the patient was associated with a 29% - 31% higher probability of having an undetectable viral load at months 6 and 12 than was the requirement that patients pay part or all of the cost of therapy. Conclusions. ART treatment programs in resource-poor settings have efficacy rates similar to those reported for developed countries. The provision of medications free of charge to the patient is associated with a significantly increased probability of virologic suppression at months 6 and 12 of ART.