Background: There is little national data on the characteristics of patients who receive high quality inpatient care defined as either the receipt of all appliable processes (all-or-none performance) or the proportion of applicable processes received during thier hospitalization.
Objectives: To assess the quality of care provided to patients hospitalized for acute myocardial infarction (AMI), heart failure or pneumonia, to describe variations in quality by patient and hospital characteristics, and the sensitivity of all-or-none performance to the number and type of processes.
Design and Subjects: Retrospective analysis of previously unavailable Hospital Quality Alliance patient-level data on 2.3 million individuals receiving care in non-federal US hospitals in 2005.
Measures: The proportion of patients who received all applicable care processes, and the mean proportion of applicable processes received by hospitalized patients.
Results: 82.8% of AMI patients, 57.3% of HF patients, and 41.7% of PN patients received all applicable care processes during their hospitalizations. Within each condition, all-or-none performance varied by patient age, race/ethnicity, and admission source, and characteristics of the hospital in which they received care. In addition, patients who were eligible for more processes were less likely to receive all of them, and certain individual processes had large impacts on all-or-none performance.
Conclusions: Large numbers of U.S. hospital patients fail to receive all recommended care. All-or-none composites are intriguing indicators of quality; however designers of such systems should address the sensitivity of these composites to both the number and type of individual processes included.