Abstract
BACKGROUND: Mortality rates among patients initiating antiretroviral therapy (ART) in sub-Saharan Africa continue high. Also HIV treatment services from the region are affronting the challenges of been attending more patients than never. In this scenario, there are no integrated scoring systems capable of an adequate risk identification/ prognostic stratification among patients requiring ART; in order of optimize actual programmes outcomes. Several independent risk factors at baseline are associated with a poor prognosis after ART initiation. These include: male sex, low body mass index, anemia, low CD4 count and stage-4 WHO disease. The aim of this research was evaluate prospectively a new scoring system composed by these factors. METHODS: An open cohort study was conducted in 1769 patients from May 2008 to December 2010 at two HIV clinics of Zimbabwe. A new clinical model (MASIB score) was applied at ART initiation and patients were followed for 4 months. After that, validation characteristics of the score were examined. RESULTS: Patients selected in this cohort exhibited similar baseline characteristics that the patients selected in previous cohorts from the region. Overall performance for mortality prediction of MASIB score was accurate, as reflected by the Brier score test result 0.084 (95%CI: 0.080-0.088). Calibration was adequate taking in consideration a p>0.05 in the Hosmer Lemeshow test and discrimination was also good (Area Under Curve: 0.915, 95%CI: 0,901- 0,928). CONCLUSION: The new model developed exhibited adequate validation characteristics supporting the clinical use. Further evaluations of this model in others scenarios from the sub-Saharan region are needed.