Henry Ford HEART Score Randomized Trial: Rapid Discharge of Patients Evaluated for Possible Myocardial Infarction.

Background: Hospital evaluation of patients with chest pain is common and costly. The HEART score risk stratification tool that merges troponin testing into a clinical risk model for evaluation emergency department patients with possible acute myocardial infarction (AMI) has been shown to effectively identify a substantial low-risk subset of patients possibly safe for early discharge without stress testing, a strategy that could have tremendous healthcare savings implications.

Method And Results: A total of 105 patients evaluated for AMI in the emergency departments of 2 teaching hospitals in the Henry Ford Health System (Detroit and West Bloomfield, MI), between February 2014 and May 2015, with a modified HEART score ≤3 (which includes cardiac troponin I <0.04 ng/mL at 0 and 3 hours) were randomized to immediate discharge (n=53) versus management in an observation unit with stress testing (n=52). The primary end points were 30-day total charges and length of stay. Secondary end points were all-cause death, nonfatal AMI, rehospitalization for evaluation of possible AMI, and coronary revascularization at 30 days. Patients randomized to early discharge, compared with those who were admitted for observation and cardiac testing, spent less time in the hospital (median 6.3 hours versus 25.9 hours; <0.001) with an associated reduction in median total charges of care ($2953 versus $9616; <0.001). There were no deaths, AMIs, or coronary revascularizations in either group. One patient in each group was lost to follow-up.

Conclusions: Among patients evaluated for possible AMI in the emergency department with a modified HEART score ≤3, early discharge without stress testing as compared with transfer to an observation unit for stress testing was associated with significant reductions in length of stay and total charges, a finding that has tremendous potential national healthcare expenditure implications.

Clinical Trial Registration: URL: http://www.clinicaltrials.gov. Unique identifier: NCT03058120.