Impact of Hyperbilirubinemia on Persistence and Adherence of Atazanavir among HIV Patients
Methods: This retrospective analysis used medical and pharmacy claims and lab data from 7/1/03-8/31/12 for commercial members from a large US managed care database. Included patients had ≥1 pharmacy claims for ATV (first claim = index date), were age 18+, had ≥6 months of continuous enrollment prior to (baseline period, or b/l) and 12 months after (follow-up period, or f/u) the index date, and no b/l ATV use. HBR was defined by total bilirubin of ≥Grade 3 (>2.5x ULN) within 90 days of index; Non-HBR was defined as ≤Grade 2 (>1.5-2.5x ULN) in the 12-month f/u. Adherence was assessed using a medication possession ratio (MPR). Persistence was defined as the days on ATV until the earliest of discontinuation or end of the study period.
Results: Approximately one-third of the 3,268 ATV patients who met criteria had lab data available (131 HBR and 1,061 non-HBR). Mean age was 43.1 years, 82.6% were male, and 36.3% were treatment-naive. Patients in the HBR cohort were more likely to be male, have a bachelor degree or higher and be Caucasian compared to patients in the non-HBR cohort. Mean (SD) MPR (0.93 (0.10) vs. 0.92 (0.11); p=0.22) and persistence on ATV therapy (260 days (133) vs. 241 (143); p=0.15) in the f/u period were similar in the HBR and non-HBR cohorts. After controlling for demographics, socioeconomic status, b/l Charlson comorbidity index, and concomitant HIV medications while receiving ATV in multivariate analyses, differences in adherence and persistence between the two cohorts remained non-significant.
Conclusion: In this large US managed care database analysis of approximately 1,200 HIV-infected patients on ATV, patients with HBR ≥Grade 3 did not show significant differences in adherence or persistence of ATV in the first year of therapy versus those who developed ≤Grade 2 HBR. HBR does not appear to influence persistence with ATV in a real-world setting.
Bristol-Myers Squibb: Employee, Salary
T. Grasso, Bristol Myers Squibb: Employee, Salary
Y. Fan, None
J. Burke, None
See more of: Poster Abstract Session