Background: The World Health Organization defines HIV virologic failure as 2 consecutive viral loads >1000 copies/mL, measured 3-6 months apart, with interval adherence support. We sought to empirically evaluate these guidelines using data from an observational cohort.
Setting: The Uganda AIDS Rural Treatment Outcomes study observed adults with HIV in southwestern Uganda from the time of antiretroviral therapy (ART) initiation and monitored adherence with electronic pill bottles.
Methods: We included participants on ART with a detectable HIV RNA viral load and who remained on the same regimen until the subsequent measurement. We fit logistic regression models with viral resuppression as the outcome of interest and both initial viral load level and average adherence as predictors of interest.
Results: We analyzed 139 events. Median ART duration was 0.92 years, and 100% were on a nonnucleoside reverse-transcriptase inhibitor-based regimen. Viral resuppression occurred in 88% of those with initial HIV RNA <1000 copies/mL and 42% if HIV RNA was >1000 copies/mL (P <0.001). Adherence after detectable viremia predicted viral resuppression for those with HIV RNA <1000 copies/mL (P = 0.011) but was not associated with resuppression for those with HIV RNA >1000 copies/mL (P = 0.894; interaction term P = 0.077).
Conclusions: Among patients on ART with detectable HIV RNA >1000 copies/mL who remain on the same regimen, only 42% resuppressed at next measurement, and there was no association between interval adherence and viral resuppression. These data support consideration of resistance testing to help guide management of virologic failure in resource-limited settings.