Patterns of Lapses and Recoveries During a Quit Attempt Using Varenicline and Behavioral Counseling Among Smokers With and Without HIV

Addressing tobacco use among HIV+ smokers is a priority. Lack of knowledge about how HIV+ smokers respond to tobacco use treatments limits our ability to effectively treat this population of smokers. Using data from 2 clinical trials that provided 12 weeks of varenicline and behavioral counseling, 1 with smokers with HIV (n = 89) and 1 with smokers without HIV (n =179), we used mixed logistic regression modeling to compare point-prevalence abstinence rates and adherence to the initial target quit date (TQD) and Cox regression for repeated outcomes to evaluate lapse and recovery dynamics between the groups. Sixty percent of HIV+ smokers refrained from smoking at the TQD while only 33% of HIV+ smokers did (odds ratio [OR] = 0.32, 95% CI [0.18, 0.56], p <.001). The point-prevalence abstinence rates at Week 12 were 31% (HIV-) and 28% (HIV+ OR = 0.7, 95% CI [0.42, 1.16], p =.16) and the point prevalence abstinence rates at Week 24 were 22% (HIV-) and 15% (HIV+ OR = 0.87, 95% CI [0.49, 1.57], p =.65). Although there was no interaction between HIV status and lapse risk, χ²(3) < 1, there was a significant interaction for the recovery model, (χ²(3) = 20.4, p < 0.001): as the number of events increased, the time to the next recovery became longer among smokers with HIV, compared to smokers without HIV.