Purpose: To describe utilization patterns for anti-diabetes medications among a cohort of diabetes patients in the Military Health System (MHS) before and after warnings about rosiglitazone issued in May 2007.
Methods: We used segmented regression analysis to compare changes in the level and trend of rosiglitazone utilization and use of other anti-diabetes therapies in the period prior to the drug warnings (between April 2006 and May 2007) and the period after the warnings were issued (between October 2007 and May 2008).
Results: The level and trend of rosiglitazone use changed after the highly publicized warnings. The number of prescriptions filled fell by almost 7000 after the warning (p < 0.001). The number of prescriptions filled for pioglitazone, sulfonylureas, and other diabetes drugs increased significantly after the warnings (p < 0.05 in all models). Overall, the level and trend of filled prescriptions per month for all anti-diabetic drugs did not significantly change after the warnings.
Conclusions: Utilization patterns changed in response to warnings about rosiglitazone. While overall utilization of anti-diabetic drugs did not change, further study is needed to determine the associated health outcomes.
(c) 2009 John Wiley & Sons, Ltd.