Myocardial Injury and Predictors of In-Hospital Mortality in Severe or Critical COVID-19: A Single-Center Observational Cohort Study

Background and Objectives: In-hospital mortality remains high among patients with severe or critical COVID-19. This study aimed to identify independent predictors of in-hospital mortality and assess the prognostic value of myocardial injury in this high-risk population. Materials and Methods: This single-center observational cohort study included 172 adults hospitalized with severe or critical COVID-19 across multiple pandemic waves. Mortality-associated factors were assessed using logistic regression and receiver operating characteristic curve analysis. Myocardial injury was defined as hs-cTnT > 14 ng/L, measured on days 3–5 after admission. Results: Of 172 patients, 60 died during hospitalization (34.9%). In the primary multivariable model, based on 126 complete cases and 37 deaths, age (aOR = 2.55 per 10-year increase; 95% CI, 1.55–4.22), pre-existing cardiovascular disease (aOR = 3.06; 95% CI, 1.03–9.07), LDH (aOR = 6.31 per doubling; 95% CI, 2.70–14.78), and CT pulmonary involvement (aOR = 1.34 per 10% increase; 95% CI, 1.01–1.80), were independently associated with in-hospital mortality. In a separate analysis of 121 patients with 35 deaths, myocardial injury remained independently associated with mortality after adjustment for age, cardiovascular disease, and pulmonary involvement (aOR = 5.19; 95% CI, 1.82–14.77). Among 131 patients with complete phenotype data, mortality was 6.7% (2/30) in patients with neither component, 19.6% (11/56) with isolated respiratory involvement, 50.0% (3/6) with isolated myocardial injury, and 56.4% (22/39) with both components. The primary multivariable model had an apparent AUC of 0.898 (95% CI, 0.828–0.968). Conclusions: Age, LDH, and myocardial injury were independently associated with in-hospital mortality, while the associations of cardiovascular disease and CT involvement were attenuated in the multiple-imputation sensitivity analysis. Combined assessment of clinical, pulmonary, and cardiac factors may improve risk stratification, but these findings require validation in independent cohorts.