Variation in chest pain emergency department admission rates and acute myocardial infarction and death within 30 days in the Medicare population.

OBJECTIVES: The objective was to assess the relationship between emergency department (ED) admission rates for Medicare beneficiaries with chest pain and outcomes, specifically 30-day rates of acute myocardial infarction (AMI) and mortality.

METHODS: Using a 20% random sample of Medicare beneficiaries in 2009, 158,295 beneficiaries with a primary diagnosis of chest pain at the conclusion of their ED visits were selected to assess outcomes based on the decision to hospitalize or discharge home. The proportions of these patients admitted to inpatient or observation status at 2,219 U.S. hospitals were calculated, adjusting for differences in patient and hospital characteristics. Both bivariate analysis and multivariable logistic regression were used to estimate the effect of the adjusted admission rates (designed to be a measure of care intensity) on patient outcomes. Other covariates in the multivariable model included patient demographics, medical conditions, and hospital utilization in the 30 days prior to the ED visits. Results from the bivariate and multivariable analyses were compared for consistency.

RESULTS: The adjusted Medicare admission rate for ED patients with chest pain averaged 63% for the middle quintile of the patient sample and ranged from 38% to 81% in the lowest and highest quintiles. The multivariable model yielded estimates of 3.6 fewer cases of AMI (95% confidence interval [CI] = 1.5 to 5.1 cases) and 2.8 fewer deaths (95% CI = 0.6 to 4.1 deaths) per 1,000 chest pain patients associated with an admission rate of 81% versus 38%. The estimates from the bivariate analysis were of similar magnitude.

CONCLUSIONS: Considerable variation exists across U.S. hospitals in ED admission rates for Medicare patients with chest pain. Hospitals that approach admissions more conservatively (i.e., higher admission rates) in this population have lower rates of AMI and mortality.