ABSTRACT
BACKGROUND: Previous studies using cardiac troponin levels to investigate the relationship between myocardial injury and mortality in sepsis patients have been conflicting. Our aim was to investigate the relationship between plasma high-sensitive cardiac troponin T (hs-cTnT) level and 30-day and 1-year mortality in sepsis patients and 30- to 365-day mortality in sepsis survivors. METHODS: Sepsis patients requiring vasopressor support and admitted to our institution between 2012 and 2021 (n = 586) were included in this retrospective cohort study. Elevated hs-cTnT values (≥15 ng/L) were divided into quartiles (Q): Q1 15-35 ng/L; Q2 36-61 ng/L; Q3 62-125 ng/L; Q4 126-8630 ng/L. Stratified Kaplan-Meier curves and multivariable Cox regression were used for survival analyses. RESULTS: First sampled hs-cTnT was elevated in 529 (90%) patients. One-year mortality was 45% (n = 264). Increasing level of hs-cTnT was independently associated with higher adjusted hazard ratios (HR) for 1-year mortality compared with normal levels: Q1 HR 2.9 (95% confidence interval [CI], 1.03-8.1); Q2 HR 3.5 (95% CI, 1.2-9.8); Q3 HR 4.8 (95% CI, 1.7-13.4); Q4 HR 5.7 (95% CI, 2.1-16). In acute phase survivors, first sampled hs-cTnT was an independent predictor of 30- to 365-day mortality (HR 1.3; 95% CI, 1.1-1.6 per loge hs-cTnT). CONCLUSIONS: First sampled plasma hs-cTnT in critically ill sepsis patients was independently associated with 30-day and 1-year mortality. Importantly, first sampled hs-cTnT was associated with mortality during the convalescence phase (30- to 365-day) and could be a feasible marker to identify acute phase survivors at high risk of death.
