Use of a prescription refill-based measure of antiretroviral therapy adherence to predict subsequent virological rebound in patients with stable undetectable HIV viral loads

V. CAMBIANO, F. LAMPE, A. RODGER, C. SMITH, R. LODWICK, J. HOLLOWAY, M. JOHNSON, A. PHILLIPS

Research output: Contribution to journalMeeting abstractpeer-review

Abstract / Description of output

Aims: To assess whether a simple, routinely available measure of adherence to antiretroviral therapy (ART) – the proportion of days covered by drug prescriptions in the previous 6 months – predicted viral rebound at the next HIV viral load (VL), in patients who were previously virologically suppressed. Methods: The analysis was performed on a cohort of HIV-infected individuals. Each drug coverage–VL episode consisted of a 6 month period immediately prior to a VL<50 copies/mL (time-zero) over which drug coverage was assessed. It was required that the patient had been continuously on ART throughout the period, with VL<50 copies/mL. The next VL after time-zero was used to assess the outcome (whether rebound, defined as >200 copies/mL). Drug coverage was calculated as the proportion of days that the individual had a valid prescription for at least three antiretroviral drugs (ARVs). Each patient could contribute more than one episode to the analysis. Poisson regression was used to describe the effect of prescription coverage on the probability of rebound. Results: Three hundred and seventy-six (2.4%) VL rebounds occurred in 15660 ‘drug coverage – VL episodes’, after a median of 2.7 years on ART (interquartile range [IQR]:1.3–4.6). The median time from time-zero to the subsequent VL measurement was 94 days (IQR: 73–119). Coverage was 100% for 37% of episodes, 96–99% for 18% of episodes and below 60% for only 5% of episodes. The risk ratio (RR) of rebound associated with a 10% increment in prescription coverage was 0.91 (95% CI: 0.87– 0.95), which was unaffected by adjusting for the potential confounding variables (RR = 0.93; 95%CI: 0.88–0.97). The results were similar when coverage by at least one drug was considered sufficient. When restricted to regimens in common use (PI/r or NNRTI: 8466 drug-coverage episodes and 185 rebound events) the adjusted RR was 0.92 (95% CI: 0.85–0.99). Conclusions: ARVs prescription coverage assessed at the time of a VL measure in patients with undetectable VL seems to be clinically useful for predicting VL rebound on the next measured VL
Original languageEnglish
Pages (from-to)21
Number of pages1
JournalHIV Medicine
Volume10
Issue number1
Publication statusPublished - 1 Apr 2009

Fingerprint

Dive into the research topics of 'Use of a prescription refill-based measure of antiretroviral therapy adherence to predict subsequent virological rebound in patients with stable undetectable HIV viral loads'. Together they form a unique fingerprint.

Cite this