Background: Primary percutaneous coronary intervention (PCI) is an important treatment option for patients with acute myocardial infarction (MI). Although an inverse association between a hospital's PCI volume and in-hospital mortality has been observed in Western studies, previous Japanese investigations have not found any such relationship.
Methods and Results:A retrospective analysis of 8,391 cases of acute MI, obtained from administrative data from 2006. The primary outcome was in-hospital mortality. Hospitals were divided into quartiles based on the number of PCI procedures per half-year (6-13, 14-22, 23-38, 39-134) and mortality rates were compared across the groups. Crude-mortality in the lowest-volume quartile was 7.0%, compared with 4.9% in the highest-volume quartile. An inverse association was found between primary PCI procedure volume and crude in-hospital mortality (P=0.016). After case-mix adjustment, a significant decrease in mortality risk for patients treated at high-volume (3 rd and 4 th quartile) hospitals compared to the lowest-volume (1 st quartile) hospitals was found.
Conclusions:Based on this administrative data, there is an inverse association between a hospital's primary PCI volume and in-hospital mortality for patients with acute MI. Periodic outcomes research is necessary in conjunction with progress in PCI practice and technology to establish the recommended PCI volume and regionalization for improvements in care.
Methods
Data Source and Study PopulationThe data source for this study consisted of discharge claim records from Japanese hospitals that were either using or preparing to implement the Diagnosis Procedure Combination (DPC) code as a payment scheme in 2006. The data were voluntarily offered to the DPC study group by the hospitals that agreed to be used for research by the DPC study group. The data were anonymous and could not be linked with any other information to identify patients when they were collected by the research group. The study was given prior approval by the Ethics Committee of Tokyo Medical and Dental University. The data contained patient information on the most resource-consuming diseases, comorbidities, complications, demographics, procedures, medications, and materials. We used discharge data from July through December 2006. For the purposes of this study, we selected patients with MI if their DPC classification at discharge was MI-related (DPC code: 050030) and the most resource-consuming disease during their hospitalization was identified as acute MI (I21.0-21.9) by International Classification of Diseases (ICD-10) coding. We included patients who underwent a PCI procedure, and we excluded patients who received thrombolytic agents during their reperfusion therapy in order to limit the sample to primary PCI patients. Additionally, if patients were hospitalized more than once, only the data from the first hospitalization were used in order to maintain the independence of observations. Finally, patients who underwent PCI followed by coronary artery bypass ...