Classification performance of clinical risk scoring in suspected acute coronary syndrome beyond a rule-out troponin profile

Abstract

Aims High-sensitivity cardiac troponin strategies can provide risk stratification in patients with suspected acute coronary syndrome (ACS) in the emergency department (ED). This study evaluated whether clinical risk scoring improves the classification performance of a rule-out profile in suspected ACS. Methods and results Patients presenting to ED with suspected ACS as part of the RAPID-TnT trial randomized to the intervention arm were included. Results ≥5 ng/L were available for all participants in this analysis. We evaluated the Thrombolysis In Myocardial Infarction (TIMI) risk score, History ECG Age Risk factors Troponin (HEART) score, and Emergency Department Assessment of Chest pain Score (EDACS) in addition to a rule-out profile based on the 0/1-h high-sensitivity cardiac troponin T protocol (<5 ng/L or ≤12 ng/L and a change of <3 ng/L at 1-h) using test performance parameters focusing on low-risk groups to identify the primary endpoint (TIMI ≤ 1, HEART ≤ 3, EDACS < 16). Primary endpoint was a composite of type 1/2 myocardial infarction (MI) at index presentation and all-cause mortality or type 1/2 MI at 30 days. A total of 3378 participants were enrolled between August 2015 and April 2019 of which 108 were ineligible/withdrew consent (intervention arm: n = 1638). Sensitivity, specificity, negative predictive value (NPV), and area under the curve (AUC) of the rule-out profile was 94.4%, 76.8%, 99.6%, and 0.86, respectively with 72.9% identified as ‘low-risk’. Adding the clinical risk scores did not improve the sensitivity, NPV, or AUC with significantly lower specificity and ‘low-risk’ classified participants. Conclusions Addition of clinical risk scores to rule-out profile did not demonstrate improved classification performance for identifying the composite of type 1/2 MI at index presentation and all-cause mortality or type 1/2 MI at 30 days. Clinical trials registration URL: https://www.anzctr.org.au. Reg. No. ACTRN12615001379505.

Association of Chest Pain Protocol–Discordant Discharge With Outcomes Among Emergency Department Patients With Modest Elevations of High-Sensitivity Troponin

“…A recent analysis of the the Rapid Assessment of Possible Acute Coronary Syndrome in the Emergency Department with High-Sensitivity Troponin T Study (RAPID-TnT) study found no improvement in classification performance in adding the HEART score to ruled-in patients in terms of identifying 30-day death or MI. 24 The present study differs from RAPID-TnT in that the HEART score was applied to a cohort of patients with chest pain who ruled-in but whom clinicians felt were candidates for discharge based on factors beyond the ADP result, and this difference may explain the additive value of the HEART score in this study.…”

Development and validation of a comprehensive early risk prediction model for patients with undifferentiated acute chest pain

“…While such models are useful in some circumstances, they may be less useful at a patient level, where exclusion of a single condition does not provide a diagnosis or estimation of overall patient risk. The HEART, EDACS and TIMI risk scores stratify chest pain patients according to risk of ACS or subsequent MI, and the discriminatory metrics of these scores may relate to differentiating ACS from other serious conditions that require treatment [5] , [9] , [10] , [11] . Moreover, these scores were developed in the conventional troponin era and while each have been validated with high-sensitivity troponin assays, they may offer more limited benefits in the setting of newer rapid rule-out pathways [5] , [6] .…”

“…The HEART, EDACS and TIMI risk scores stratify chest pain patients according to risk of ACS or subsequent MI, and the discriminatory metrics of these scores may relate to differentiating ACS from other serious conditions that require treatment [5] , [9] , [10] , [11] . Moreover, these scores were developed in the conventional troponin era and while each have been validated with high-sensitivity troponin assays, they may offer more limited benefits in the setting of newer rapid rule-out pathways [5] , [6] . Existing scores and rapid troponin pathways do not provide information regarding the risk of a non-coronary diagnosis being present, which partially limits their utility in facilitating early discharge.…”

“…Existing risk score tools for undifferentiated chest pain used in pre-hospital and emergency department settings are focused on determining a patient’s risk of acute coronary syndromes (ACS), which account for approximately 10% of presentations [1] , [2] , [3] . These scores were largely developed prior to the availability of high sensitivity troponin assays, and with the development of rapid high-sensitivity troponin rule-out pathways, their incremental value in improving classification performance has lessened [5] , [6] , [7] , [8] . Importantly, existing risk tools provide no information on the presence of serious non-coronary conditions (e.g.…”