199 Implementation of a Modified HEART Score Pathway in an Urban, Safety Net Hospital: A Before-After Study

Abstract

We sought to decrease practice variation in our emergency department (ED) for patients presenting with acute, non-traumatic chest pain through implementation of a chest pain risk stratification guideline using the HEART score. Additionally, we sought to validate the HEART score in a large, urban, safety net population. We performed a before-after study using standardized medical record abstraction at a single, urban, safety net institution. Consecutive adult patients were identified by ED ICD-10 diagnosis codes. Inclusion criteria were: age ≥ 19 years, a troponin order during ED evaluation, and an ED diagnosis of chest pain or acute coronary syndrome. Exclusion criteria were: ST elevation myocardial infarction or a non-cardiac diagnosis for chest pain (eg, pneumonia or pulmonary embolism). Prior to the development of an institutional chest pain risk stratification guideline, 3 months of consecutive eligible patient medical records underwent standardized medical record abstraction by emergency physicians to understand our current practice and retrospectively evaluate the HEART score in our patient population. Using this data, a multidisciplinary chest pain risk stratification guideline similar to the HEART score pathway was developed and subsequently implemented in our institution. The impact of this guideline on ED length of stay, hospital admissions, and utilization of non-invasive stress testing and cardiac catheterization were evaluated through an additional 3 months of standardized medical record abstraction by emergency physicians. Patients with either MACE identified during the initial encounter or a health care encounter in the medical record 6 weeks after the index ED encounter were used to validate the HEART score for predicting major adverse cardiac events (MACE) in our population. Descriptive and bi-variate statistics as well as estimates of effect size were performed. 1173 patients met inclusion in our study including: 521 in the before group and 652 in the after group. The HEART score identified 57% of our patients as low risk (≤ 3), 38% moderate risk (4-6), and 5% high risk (≥7). Among the 769 patients in the HEART score validation cohort, the prevalence of MACE was 0% (95% CI 0-1%) in the low risk group, 10% (95% CI 8-14%) in the moderate risk group, and 55% (95% CI 41-68%) in the high risk group. Implementation of an institution specific HEART score pathway increased admission for the moderate risk group by 38% (95% CI 29-47%), decreased the median ED length of stay by 37 minutes (95% CI 17-58 min), and increased objective cardiac testing among moderate and high risk patients by 10% (95% CI 0-19%). Adherence to our institutional HEART score pathway was 84% (95% CI 81-86%) after implementation. The HEART score is a valid risk stratification tool in an urban, safety net hospital population. In contrast to prior studies, the implementation of a modified HEART score pathway resulted in increased admissions at our institution as well as an increased objective cardiac testing and a decrease in ED LOS.