Written by Jacob Altholz
Both the HEART- and HET-score (using history, ECG changes, and high-sensitivity troponin) performed similarly when risk-stratifying ACS rule-out patients, but the HET-score had higher overall diagnostic accuracy.
Should we move this from our HEART to our HET?
This study was broken into two phases in which adults with chest pain suggestive of ACS were initially stratified by 1) standardized guidelines from the European Society of Cardiology and ACC/AHA guidelines, 2) then in the second phase prospectively using the HEART-score. Both phases were combined, and performance of the HEART- and HET-score was evaluated in all 1,181 patients across 6 centers in two Swedish cities. The primary endpoint was a composite of MI at the index visit OR readmission to the hospital due to a new MI or cardiovascular death within 30 days, validated by two cardiologists. These scores were considered low, intermediate, and high: HEART 0–2, 3–5, and 6–10; HET 0, 1–2, and 3–6, respectively. The Roche hs-cTnT assay was used at five centers and Abbott hs-cTnI assay at one.
In total, 132 patients met the primary endpoint. All variables within the HEART-score correlated with higher risk; however, taken together in multivariable logistic regression, only history, ECG, and troponin were significant. Troponin levels were most strongly associated with the outcome (OR 6.21), and risk factors showed a non-significant, lower risk, OR 0.72 (95%CI 0.52–1.00, p=0.051). Overall, the scores performed similarly in discrimination, though with different distributions.
Comparing the HEART-score against the HET-score in ROC-analysis, the HET-score slightly outperformed, with an AUC of 0.887 vs. 0.853 (p<0.001). While the study was relatively small and was performed in a homogenous demographic, it suggests a HET-score may be safe.
How will this change my practice?
This study helps confirm my suspicion that cardiac risk factors and age blur the utility of the HEART-score. In those with no ECG changes, undetectable or stable troponin, and a history that doesn’t suggest ACS, regardless of age/risk factors, I feel more confident in having a lower suspicion for ACS.
Peer Reviewed by Dr. Ketan Patel
Editor’s note: Although there was a prospective component, this was actually a retrospective study. It used high-sensitivity troponin assays, which must be considered. I manually calculated diagnostic accuracy measures: HEART Score 0-2 had 98.4% sensitivity, 99.5% NPV, 36.1% specificity, 16% PPV; HET Score 0 had 97.7% sensitivity, 99.4% NPV, 43.7% specificity, 17.5% PPV. This is interesting and hypothesis generating, but I would not use HET in practice based on this study. ~Clay Smith
These were not the first authors to take letters away from HEART, there’s also a HEAR score.
HEART-score can be simplified without loss of discriminatory power in patients with chest pain – Introducing the HET-score. Am J Emerg Med. 2023 Oct 1;74:104-111. doi: 10.1016/j.ajem.2023.09.037. Epub ahead of print.