Readmission after inpatient percutaneous coronary intervention in the Medicare population from 2000 to 2012

Abstract

Since year 2000, reducing hospital readmissions has become a public health priority. In addition, there have been major changes in percutaneous coronary intervention (PCI) during this period. The cohort consisted of 3,250,194 patients admitted for PCI from January 2000 through November 2012. Overall, 30-day readmission was 15.8%. Readmission rates declined from 16.1% in 2000 to 15.4% in 2012 (adjusted odds ratio for readmission 1.33 in 2000 compared with 2012). Of all readmissions after PCI, the majority were for cardiovascular-related conditions (>60%); however, only a small percentage (<8%) of total readmissions were for acute myocardial infarction, unstable angina, or cardiac arrest/cardiogenic shock. A much larger percentage of patients were readmitted with chest pain/angina (7.9%), chronic ischemic heart disease (26.6%), and heart failure (12%). A small proportion was due to procedural complications and gastrointestinal (GI) bleeding. The use of PCI with stenting during readmissions was variable, increasing from 14.2% in 2000 to 23.7% in 2006 and then declining to 12.1% in 2012. Hospital mortality during readmission was 2.5% overall and varied over time (2.8% in 2000, decreasing to 2.2% in 2006 and then rising again to 3.1% in 2012). Patients who were readmitted had >4× higher 30-day mortality than those who were not. Among Medicare beneficiaries, readmission after PCI declined over time despite patients having more comorbidities. This translated into a 33% lower likelihood of readmission in 2012 compared with 2000. A small proportion of readmissions were for acute coronary syndromes.