Medication performance measures and mortality following acute coronary syndromes.
PURPOSE: To identify patient and health care factors which are related to the use of medical treatments that comprise quality measures and to assess the relation of these measures with mortality. METHODS: The study sample consisted of 20 140 patients with acute coronary syndromes from the international GRACE registry. Multivariable logistic regression modeling was used to determine predictors of quality performance. Quality indicators were use of aspirin and beta-blockers within 24 hours and at hospital discharge, use of angiotensin-converting enzyme (ACE) inhibitors at discharge, and in-hospital mortality. RESULTS: Use of medications in eligible patients at discharge ranged from 73% for ACE inhibitors to 93% for aspirin. High-risk features (eg, heart failure, older age) were related to failure to use aspirin and beta-blockers. Being at a teaching hospital and care by a cardiologist were associated with better use of aspirin and beta-blockers. Coronary artery bypass surgery was associated with failure to use ACE inhibitors and aspirin. When hospitals were divided into quartiles of quality performance, adjusted in-hospital mortality was 4.1% in the top versus 5.6% in the bottom quartile, representing a 27% (95% confidence interval: 11% to 42%) lower relative mortality. CONCLUSION: Identification of factors associated with failure to use proven treatments, including high-risk groups that would derive particular benefit from effective therapies, provides an opportunity to focus quality improvement interventions. The association of lower hospital mortality with better use of selected medical treatments supports their measurement to improve quality of care.
Granger, CB; Steg, PG; Peterson, E; López-Sendón, J; Van de Werf, F; Kline-Rogers, E; Allegrone, J; Dabbous, OH; Klein, W; Fox, KAA; Eagle, KA; GRACE Investigators,
Volume / Issue
Start / End Page
Pubmed Central ID
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)