Estimating prevalence of accumulated HIV-1 drug resistance in a cohort of patients on antiretroviral therapy

    loading  Checking for direct PDF access through Ovid


ObjectivesEstimating the prevalence of accumulated HIV drug resistance in patients receiving antiretroviral therapy (ART) is difficult due to lack of resistance testing at all occasions of virological failure and in patients with undetectable viral load. A method to estimate this for 6498 EuroSIDA patients who were under follow-up on ART at 1 July 2008 was therefore developed by imputing data on patients with no prior resistance test results, based on the probability of detecting resistance in tested patients with similar profiles.MethodsUsing all resistance test results available, predicted intermediate/high-level resistance to specific drug classes [nucleoside reverse transcriptase inhibitors (NRTIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs) and protease inhibitors (PIs)] was derived using the Stanford algorithm v5.1.2. Logistic regression models were then employed to estimate predicted probability of resistance to each drug class for given values of current viral load, history of virological failure and previous virological suppression. Based on these predicted probabilities and patients’ covariate profiles, estimates of prevalence in 5355 patients with no prior test results were obtained. Overall prevalence of resistance was estimated by pooling these data with those observed in the remaining 1143 tested patients.ResultsPrevalence of NRTI, NNRTI and PI resistance was estimated as 43% (95% confidence interval: 39%–46%), 15% (13%–18%) and 25% (22%–28%), respectively.ConclusionsThis method provides estimates for the proportion of treated patients in a cohort who harbour resistance on a given date, which are less likely to be affected by selection bias due to missing resistance data and will allow us to estimate prevalence of resistance to different drug classes at specific timepoints in HIV-infected populations on ART.

    loading  Loading Related Articles