In Reply: Dr Weinrauch and Takeuchi question whether the mortality differences seen between stroke centers and nondesignated hospitals were “nonspecific” because stroke centers also had lower observed mortality for patients admitted with acute myocardial infarction or gastrointestinal hemorrhage. Although there were nonsignificantly lower mortality rates at stroke centers for both conditions, patients admitted with acute myocardial infarction or gastrointestinal hemorrhage at stroke centers were relatively younger. They also tended to have less comorbidity than patients admitted to nondesignated hospitals. After adjustment for these factors using either a multivariable logistic regression model or instrumental variable analysis, we found no differences in mortality rates between patients with acute myocardial infarction and gastrointestinal hemorrhage who were treated at the stroke vs nondesignated centers. Thus, our finding regarding better mortality at stroke centers appears specific for stroke and not for other common causes for admission.
Xian Y, Peterson ED. Stroke Center Designation and Mortality—Reply. JAMA. 2011;305(16):1656. doi:10.1001/jama.2011.528