To the Editor: The meta-analysis by Dr Singh and colleagues1 appears to demonstrate an increased risk of cardiovascular death, myocardial infarction, or stroke events in patients using inhaled anticholinergics. I believe that this study incorporates errors in article identification and data extraction.
At least 1 article that meets the authors' search criteria for long-acting anticholinergics and recorded cardiovascular events was not included.2 The event data in Casaburi et al3 showed no difference between tiotropium and controls (7 deaths in each group and serious adverse event rates of 18% in the tiotropium group vs 21% in the control group). However, Singh et al reported 12 vs 3 events, respectively. The study by Wedzicha et al4 apparently showed 36 events in the tiotropium group and 26 events in the control group when nervous system, vascular, and cardiac disorders were combined, whereas Singh et al reported 23 vs 13 events, respectively. Reanalysis of the data in the study by Singh et al with these changes might shift the risk ratios toward 1, which would indicate no significant cardiovascular event difference between tiotropium and other agents. This likelihood is substantiated by the results of the UPLIFT trial.5