Construction and Evaluation of a Predictive Model for Liver Injury Associated with Heatstroke: A Multicenter Retrospective Study

Abstract

Background. Liver injury related to heatstroke plays a pivotal role in secondary multiorgan damage and is a direct cause of mortality in affected patients of heatstroke. This study was designed to identify independent risk factors associated with liver injury in heatstroke and to construct a clinically applicable predictive model. Methods. We conducted a retrospective analysis of 188 patients diagnosed with heatstroke, treated in the emergency departments of eight medical institutions from July 1, 2022, to September 30, 2023. Patients were categorized into a liver injury group (n = 80) and a nonliver injury group (n = 108), based on liver function indices recorded during hospitalization. Lasso regression was employed for variable refinement, while multifactorial logistic regression was utilized to identify independent risk factors for liver injury in heatstroke and to construct a nomogram model. The model’s efficacy was evaluated using the C-index, calibration curves, and decision curve analysis, examining its discriminative ability, calibration, and clinical utility. Results. The nomogram included predictive factors such as the Glasgow score, absolute lymphocyte count, lactate dehydrogenase levels, and creatine kinase isoenzyme. The model showed high accuracy and discriminative capability. The C-index was 0.852 (95% CI 0.80–0.905) with a calibration index of 0.843. Decision curve analysis revealed significant clinical applicability for this nomogram. Conclusion. The study identified four key independent risk factors for liver injury in heatstroke patients. The constructed nomogram, based on the four clinical indicators, demonstrated robust predictive accuracy, discriminative power, and clinical relevance.