Abstract 13592: Addition of the Heart Score to a hsTnT Accelerated Diagnostic Protocol: Effect on Negative Predictive Value at Different hsTnT Thresholds

Abstract

Background: The majority of pts with CP and low to modest elevations of high sensitivity troponin (hsTn) are safe for discharge from the ED; thus improvement of the negative predictive value (NPV) of the CP accelerated diagnostic protocol (ADP) can have significant impact on ED throughput. Prior studies have examined the role of adding bioclinical risk scores such as the HEART score (HS) to a hsTn ADP. In general these studies suggest the HS does not add to prediction of 30d MACE for pts with low peak troponin levels, i.e. less that the 99 th percentile. Little is known about the effect of the HS on hsTn ADP for higher values of troponin. Research question: How would the addition of the HS affect the NPV of a hsTnT ADP across a range of modest hsTnT elevations? Methods: Retrospective analysis of pts presenting with CP and assessed with a validated hsTnT ADP as well as an EHR-embedded HS decision tool from 5/2022-3/2023. The hsTnT ADP was derived from the ESC guidelines and pts were excluded if peak hsTnT was > 52ng/l (the cutoff for “rule-in”). For calculation of NPV for MACE at 30 days post-discharge pts were grouped by peak hsTnT as <6ng/l, <12ng/l, and <52ng/l. HS was grouped as < 3 (low risk) or >3 (intermediate-high risk). Comparison of NPV results was performed using chi square analysis. Results: Summarized in the Table. The addition of the HS did not significantly change the NPV for 30d MACE for CP pts with peak hsTnT <6ng/l or <12ng/l. In contrast, NPV in pts with peak hsTnT <52ng/l was improved significantly by addition of the HS. Conclusion: The additive value of the HS to a hsTnT ADP for identifying pts safe for ED discharge depends on the range of troponin measurement- HS did not change NPV at very low levels (ie from < limit of quantitation – approx. 99 th percentile) but did improve the NPV at modest elevations between the 99 th percentile and the cutoff for “rule-in” by our ADP. This findings should be considered when designing hsTn CP evaluation protocols that incorporate HS into the algorithm.