RESUMO
OBJECTIVE: We sought to validate the D:A:D risk score for chronic kidney disease (CKD) in people living with HIV in a cohort from Harare, Zimbabwe. In addition, we aimed to evaluate proteinuria as a predictive variable in the risk score model, being the first study to do so. DESIGN: Data from people living with HIV attending a clinic in Harare were evaluated. Those with a baseline estimated the glomerular filtration rate >60 mL/min/1.73 m 2 , and at least 2 subsequent estimated glomerular filtration rate measurements were included. A modified version of the D:A:D risk score model was applied to categorize participants as "low," "medium," and "high-risk" of progression to CKD. Potential predictors of renal impairment were assessed by logistic regression in univariate and multivariate models. Proteinuria was evaluated in a nested model using D:A:D risk categories. RESULTS: Two thousand seven hundred ninety-three participants were included. Forty participants (1.4% of the cohort) progressed to CKD during the median follow-up time of 4.2 years. Progression rates were 1%, 3%, and 12% in the low, medium, and high-risk groups, respectively. Proteinuria data were available for 2251 participants. The presence of proteinuria was strongly associated with progression to CKD [(OR 7.8, 95% CI: 3.9 to 15.7), and its inclusion in the risk score improved the discrimination of the model with the c-statistic increasing from 0.658 to 0.853]. CONCLUSION: A modified version of the D:A:D CKD risk score performed well in predicting CKD events among this sub-Saharan African cohort of people living with HIV. Inclusion of proteinuria into the risk score model significantly improved predictability.