Predictive models of immune microenvironment-related markers in patients with sepsis accompanied by myocardial dysfunction and their roles in diagnosis.

Objective: To evaluate immune microenvironment markers for predicting sepsis-induced myocardial dysfunction (SIMD) and establish three predictive models-nomogram, decision tree, and gradient boosting machine (GBM)-to compare their efficacy in assessing SIMD risk. Method: A retrospective analysis was conducted on the clinical data of 165 patients with sepsis who were admitted between January 2022 and February 2025. Patients were divided into SIMD and non-SIMD groups according to the occurrence of SIMD. Risk factors influencing the occurrence of SIMD in patients with sepsis were screened using univariate and multivariate logistic regression analyses. Nomogram, decision tree, and GBM models were constructed based on the results of the multivariate logistic regression analysis. The area under the receiver operating characteristic curve (AUC) was used to evaluate the discrimination of each model. The accuracy, sensitivity, specificity, and F1 scores of the three models were calculated. Result: : Among the 165 patients with sepsis included in the study, 75 were in the SIMD group, accounting for 45.45% (75/165). Univariate analysis showed significant differences between the two groups in APACHE II score, white blood cell count, N-terminal pro-brain natriuretic peptide (NT-proBNP), soluble triggering receptor expressed on myeloid cells-1 (sTREM-1), and high mobility group box 1 (HMGB1) levels ( Conclusion: The immune markers, sTREM-1 and HMGB1, were associated with SIMD. Elevated APACHE II score and NT-proBNP, sTREM-1, and HMGB1 levels are risk factors for SIMD in patients with sepsis. Predictive models based on these factors demonstrate strong performance and effectively identify high-risk individuals, aiding in early clinical intervention.