BACKGROUND
Studies of hospital quality and national performance measures for acute myocardial infarction (AMI) frequently exclude transfer patients. Little is known about the clinical characteristics and outcomes of patients with AMI transferred for revascularization.
HYPOTHESIS
Transfer patients have greater clinical comorbidity and worse hospital survival than non-transfer patients, and negatively impact hospital quality measures.
METHODS
A retrospective-cohort study was performed using all patients with ST-elevation myocardial infarction who underwent coronary intervention or coronary artery bypass grafting (CABG) in Washington State from 2002 – 2005. Data on clinical and procedural characteristics, medications, and complications were obtained from the Clinical Outcomes Assessment Program. Hospitals were compared by rates of death and discharge with aspirin, beta-blockers, lipid lowering agents, and ACE inhibitors. Logistic regression was used for adjusted analysis.
RESULTS
Of patients undergoing revascularization for AMI, 7080 were directly admitted and 2910 were transferred. Diabetes (23.4 v. 19.7%, p<0.01), hypertension (61.3 v. 55.7%, p<0.01), and thrombolysis (32.3 v. 3.4%, p <0.01) were greater among transfers. Transfers presented with a higher rate of left main and three-vessel disease, intra-aortic balloon pump use (6.4 v. 3.6%, p<0.01) and underwent CABG more frequently (15.4 v. 5.5%, p <0.01). Transfer patients had a lower risk of death (3.9 v. 4.9%, p=0.03), but no difference in discharge medication prescription. Adjusting for major risk factors, procedure, and hospital type, transfers had a similar risk for in-hospital death compared to non-transfers (OR 0.9, CI 0.5 – 1.6). Hospitals with a high percentage of transfers treated higher-risk patients, but had similar outcomes to those with few transfers. Excluding transfers from the hospital-level analysis did not appreciably change these results.
CONCLUSION
Transfers were higher-risk, but had similar in-hospital mortality and were equally likely to receive appropriate medication at discharge compared to directly admitted patients. Inclusion of transfers did not affect hospital-level inpatient mortality or measurements of adherence to quality guidelines.