Optimization of high-sensitivity cardiac troponin T cutoff values stratified by sex, age, and comorbidities for enhanced diagnostic accuracy of acute myocardial infarction: a diagnostic accuracy study

Acute myocardial infarction (AMI) diagnosis uses high-sensitivity cardiac troponin T (hs-cTnT) thresholds (99th percentile upper reference limit), but universal cutoffs may not account for sex, age, and comorbidity variations. This study aimed to derive and validate subgroup-specific hs-cTnT cutoffs to improve diagnostic accuracy. This retrospective study analyzed 13,525 hospitalized patients (October 2013-March 2021) divided into derivation (70%, n = 9,467) and validation (30%, n = 4,058) cohorts. Subgroup-specific cutoffs (sex, age, renal function) were derived via ROC analysis (Youden’s index) and compared to the conventional cutoff (14 ng/L). Subgroup-specific cutoffs improved performance: Sex-specific cutoffs: 32.5 ng/L (male: sensitivity 85.0% [82.7–87.0], specificity 90.2% [89.4–91.0]) vs. 22.5 ng/L (female: sensitivity 87.6% [82.6–91.3], specificity 92.3% [91.4–93.2]). Age-specific cutoffs ranged from 77.5 ng/L (20–39 years: sensitivity 83.3% [64.2–93.3], specificity 96.9% [94.8–98.1]) to 36.5 ng/L (≥ 80 years: sensitivity 85.1% [77.7–90.4], specificity 89.2% [87.0–91.0]). Renal function: 126.5 ng/L (eGFR < 60 mL/min/1.73 m²: sensitivity 83.5% [74.2–89.9], specificity 86.2% [82.8–89.0]). All subgroups achieved AUCs > 0.9 ( P < 0.001). Validation showed higher specificity, PPV, and Youden’s index with stratified cutoffs. Subgroup-specific hs-cTnT cutoffs significantly enhance AMI diagnostic accuracy, supporting individualized clinical decisions.