Identifying ED Patients With Acute Chest Pain for Early Discharge

Study Questions:

Is the implementation of the HEART Pathway, an electronic health record (EHR)-based clinical decision support tool, associated with improved safety and decreased utilization in patients presenting to the emergency department (ED) with acute chest pain?

Methods:

The HEART Pathway is triggered by patients presenting with chest pain with at least one troponin ordered. The tool prompts providers with a series of questions to risk-stratify eligible patients in real time, classifying patients as either low-risk for acute coronary syndrome or non–low-risk based on the HEART score incorporating troponin levels. Enrollment was comprised of 3,713 patients presenting with acute chest pain, but no ST-segment elevation myocardial infarction (MI) on electrocardiography at three hospitals in North Carolina over 12 months pre-implementation of HEART, and 4,761 patients over 12 months post-implementation. The primary effectiveness outcome of hospitalization rate at 30 days and the primary safety outcome was death/MI at 30 days.

Results:

Overall the cohort consisted of 54% women and 29% African Americans, with a median age of 54 years, and an overall death/MI rate of 6.5%. Differences between patients in the pre- and post-implementation phases were minor. The HEART Pathway identified 31% as low-risk and 53% as non–low-risk. Among those classified as low-risk, six patients (0.4%) met the outcome of death/MI. More MIs were detected in the post-implementation period (6.6% vs. 5.7%), but there were no differences in the primary safety outcome (1.3% vs. 1.1%) or death (0.2% vs. 0.3%) between the pre- and post-implementation periods. Overall, healthcare utilization was notably decreased with HEART: There was a 6% decrease in hospitalizations, more frequent early discharge (43% vs. 37%), less frequent stress testing or angiography (31% vs. 35%), and shorter length of stay (16 hours vs. 18 hours). Last, nonadherence to the HEART pathway guidance occurred in 16% of low-risk and 1.2% of non–low-risk patients.

Conclusions:

In this multicenter implementation of an EHR-based clinical decision support tool, healthcare utilization significantly decreased, all the while minimizing adverse events at 30 days.

Perspective:

This study is a good example of how EHR-based interventions can lead to improved safety outcomes while reducing utilization and costs. While small in magnitude, large-scale implementation could lead to substantial savings in healthcare resources. Yet we are only scratching the surface. The HEART Pathway uses a basic scoring system, interrupts workflow requiring physician input, and likely contributes to EHR alert fatigue. With the creation of large, well-characterized registries (i.e., big data), the use of machine learning in healthcare, and the adoption of human-centered design, it is conceivable that better algorithms and processes requiring minimal human input will lead to even better risk stratification, personalized care, and an overall more cost-effective healthcare system.

Clinical Topics: Acute Coronary Syndromes, Invasive Cardiovascular Angiography and Intervention, Prevention, Interventions and ACS, Interventions and Imaging, Angiography, Nuclear Imaging

Keywords: Acute Coronary Syndrome, Angiography, Chest Pain, Decision Support Systems, Clinical, Electrocardiography, Electronic Health Records, Emergency Service, Hospital, Exercise Test, Myocardial Infarction, Patient Discharge, Risk Assessment, Secondary Prevention, Troponin


< Back to Listings