Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: Drs Ting and Long raise important points regarding the challenge of observational analyses. We concur that the handling of transfer patients in analyses such as ours may be problematic due to censoring. Transferred patients in CRUSADE were slightly younger than nontransferred patients (mean age, 66 vs 67 years, respectively) and had fewer prior strokes (9% vs 11%) or signs of coronary heart failure at presentation (20% vs 22%) (P<.001 for all comparisons). Acute treatment adherence patterns (prior to transfer) were also slightly higher in those patients transferred vs those not transferred: aspirin use (92% vs 91%), β-blocker (82% vs 79%), and glycoprotein IIb/IIIa inhibitors (37% vs 35%) (P<.01 for all comparisons).
Peterson ED. Hospital Performance and Acute Coronary Syndrome Outcomes. JAMA. 2006;296(11):1349-1350. doi:10.1001/jama.296.11.1349-b