Impact of FDA advisories on smoking cessation for Veterans with chronic obstructive pulmonary disease

Many persons with chronic obstructive pulmonary disease (COPD) require medication to quit smoking, yet few receive highly effective therapy with varenicline. One likely barrier is historical Food and Drug Administration safety advisories for psychiatric risk, which were associated with reduced varenicline prescribing nationwide. It is unknown whether prescribers avoided varenicline even among persons with high smoking related morbidity or low mental health risk. To inform future prescribing directives, we examined the association of advisories and varenicline dispensation among Veterans hospitalized with COPD between 2007 and 2011 who were currently smoking tobacco despite recent cessation medications. Using electronic health record and pharmacy data, we compared rates of post-hospital varenicline dispensation between persons who were hospitalized before versus after the first FDA advisory in February 2008. We estimated rate ratios using Poisson regression, stratified by the presence of any mental health disorder and adjusted for likely confounders of demographics and smoking related morbidity. Among 344 eligible Veterans, advisories were associated with lower varenicline dispensation (adjusted RR 0.47, 95% CI 0.27 – 0.82). This association was not modified by the presence of mental health disorders. In secondary analysis, veterans who received varenicline versus other therapy were more likely to quit smoking at 12 months (adjusted RR 1.73, 95 % CI 1.07 – 2.78). These findings suggest that FDA safety advisories were associated with a nonspecific decrease in varenicline prescribing, representing a lost opportunity for smoking cessation among persons with tobacco related morbidity. Better understanding of prescribers’ response to safety advisories might allow more targeted prescribing directives and mitigate unintended effects.