RESUMO
BACKGROUND: The aim of this study was to assess the cost-effectiveness of HIV treatment alternatives - with and without highly active antiretroviral therapy (HAART) - within alternative strata based on the CD4+ T-cell count at the initiation of treatment in a low-resource setting. METHODS: A retrospective observational study was conducted following 286 HIV-positive individuals admitted to the principal teaching hospital in Casablanca, Morocco, between 1995 and 2002. Patients were stratified by CD4+ T-cell count and regression models were fitted to determine risk of opportunistic infection. Data on healthcare resource use were derived from patient records and were evaluated from the hospital perspective. RESULTS: HAART led to a significant reduction in the number of HIV-related opportunistic infections (P<0.0001), extended survival (61.3 versus 55.2 months; P<0.0001) and reduced hospital stays (P<0.0001) in comparison with care in the absence of HAART. When medical care and drug costs were considered together, HAART was more costly than providing treatment for opportunistic infections. The incremental cost-effectiveness ratio was lower than gross domestic product (GDP) per capita for patients starting HAART with a CD4+ T-cell count <200 cells/mm3, but this increased to nearly three times GDP per capita when HAART was initiated at CD4+ T-cell counts above this threshold. CONCLUSIONS: HAART is more cost-effective than treating HIV-related opportunistic infections and, contrary to conclusions drawn in developed countries, HAART is more cost-effective when the CD4+ T-cell count drops to <200 cells/mm3.