Résumé
<B>Objective:</B> Pharmaceutical manufacturers conduct drug use-results surveys (DUSs) for drug reexamination applications in Japan. However, most DUSs are single-cohort studies to confirm drug safety; therefore little information is obtained from a DUS for safety comparison with other drugs. Integrating DUS data is a novel method to compare drug safety profiles among drugs. This study examined the usefulness of integrating DUS data.<br><b>Design:</b> Active surveillance<br><b>Methods:</b> We integrated DUS data obtained from two DUSs of antihypertensive drugs-a calcium channel blocker (CCB), azelnidipine (CalblockR), and an angiotensin-receptor blocker (ARB), olmesartan medoxomil (OlmetecR)-with similar study protocols. Both studies were conducted by the same pharmaceuticalcompany. Matching patients using propensity scores, we examined whether the DUS results could be applied to pharmacoepidemiology data resources for hypothesis strengthening. Safety outcomes included vasodilation-related events, which are typical adverse drug reactions (ADRs) to CCB. The incidence of safety outcomes was compared by conditional logistic regression models. Two definitions for safetuoutcomes were employed: definition 1, physician-reported adverse events; and definition 2, physicianreported ADRs.<br><b>Results:</b> In a total of 7196 patients included in the analysis, the propensity-matched patients were well balanced. Most safety outcomes were detected in the CCB group including a significant increase in vasodilation-related events (odds ratio for definition 1 versus ARB group, 1.75; 95% confidence interval,1.01-3.03).<br><b>Conclusion:</b> Integrated DUS data collated by pharmaceutical manufacturers following clear criteria are potentially useful to commence pharmacoepidemiology studies designed to strengthen hypotheses.