44 Emergency Physician Adherence to a Chest Pain Risk Stratification Decision Aid

Abstract

Clinical decision aids, such as the HEART Pathway, have been developed to safely identify patients with acute chest pain for early discharge without stress testing. When followed, they may assist emergency care providers by reducing over and under testing. However, the frequency of non-adherence, which may threaten decision aid effectiveness, has not been well described in the literature. Therefore, the objective of this study is to test provider adherence to a chest pain risk stratification decision aid and determine rates of over and under testing. From September 2012 to February 2014 the HEART Pathway was implemented as part of a randomized controlled trial. Adult emergency department (ED) patients (>21 years old), with acute chest pain concerning for acute coronary syndrome were enrolled and randomized to the HEART Pathway or usual care. In the HEART Pathway arm, a HEART Score was calculated by the attending emergency physician and serial troponins (at 0 and 3 hours) were used for real-time decision support regarding the need for further testing and disposition. Hospitalization and objective cardiac testing (stress testing or coronary angiography) was recommended for high-risk patients, with a HEART score >3 or a positive troponin. Early discharge from the ED without objective testing was recommended for low-risk patients (HEART score <3 and negative troponins). In the usual care arm, providers were encouraged to follow ACC/AHA guidelines. Participants were followed for 30 days to determine index visit disposition, non-adherence (over and under testing), and major adverse cardiac events (MACE). MACE was defined as death, myocardial infarction, and revascularization at 30 days. Non-adherence to the HEART Pathway was defined as discharging a high-risk patient from the ED without objective testing (under testing) or admitting/obtaining objective cardiac testing on a low-risk patient (over testing). Non-adherence in the usual care arm was defined by guideline non-compliance. Adherence to the HEART Pathway was compared to usual care (guideline) adherence using chi-squared testing. Logistic regression was used to determine patient attributes (age, race, sex, cardiac risk factors, etc) associated with non-adherence. Over the study period, 282 patients were enrolled with 141 patients randomized to each arm (HEART Pathway and usual care). Non-adherence to the HEART Pathway occurred in 27/141(19.1%, 95% CI:12.7-25.6%) compared to 26/141(18.4%, 95% CI:12.0-24.8%) in the usual care group (P=0.89). In the HEART Pathway arm over-testing occurred in 19/141(13.5%, 95% CI:7.8-19.1%) and under-testing occurred in 8/141(5.7%, 95% CI:1.8-9.5%). No patient attributes were significant predictors of non-adherence. None of the 19 patients with over-testing or 8 with under-testing in the HEART Pathway arm suffered a MACE during the index visit or 30-day follow-up period. Real-time use of the HEART Pathway, a chest pain risk stratification decision aid, had a non-adherence rate of nearly 20%. Most non-adherence was due to over-testing, in patients who ultimately did not have MACE within 30 days. Non-adherence rates did not differ between the HEART Pathway and ACC/AHA guidelines. Further study is required to determine and ameliorate the causes of non-adherence, which may threaten decision aid effectiveness.