Development and validation of a predictive model for in-hospital mortality in patients with sepsis-associated liver injury

Abstract

Background: Sepsis is often accompanied by organ dysfunction and acute organ failure, among which the liver is commonly involved. Sepsis patients suffering from liver injury have a greater risk of mortality than patients suffering from general sepsis. As of now, there are no tools that are specifically designed for assessing the prognosis of such patients. This study aimed to develop and validate a model to predict the risk of in-hospital mortality in patients with sepsis-associated liver injury (SALI).

Methods: Data were obtained from the Medical Information Mart for Intensive Care (MIMIC)-IV database. In the analysis, all patients with SALI who met the inclusion and exclusion criteria were included. A primary outcome was in-hospital mortality, and clinical data were extracted for these patients. In a ratio of 8:2, the data were divided into training and validation groups at random. Least absolute shrinkage and selection operator (LASSO) regression was used for data dimension reduction and feature selection, and independent factors related to prognosis were identified through multi-factor logistics analysis. A nomogram was developed to visualize the model, and the performance of the model was evaluated by the area under the curve (AUC) as well as calibration and decision curve analysis (DCA) through internal verification.

Results: A total of 616 and 154 patients with SALI were included in the training and validation cohorts, respectively. The LASSO regression and logistic multivariate analysis showed that nine factors were associated with in-hospital mortality in patients with SALI. Both the training and validation cohorts had higher AUCs than sequential organ failure assessment (SOFA) and simplified acute physiology score 2 (SAPS2): 0.753 (95% CI: 0.715-0.791) and 0.783 (95% CI: 0.749-0.817), respectively. Both the training and validation cohorts showed good calibration results for the prediction model. In terms of clinical practicability, DCA of the predictive model demonstrated greater net benefits than the SOFA and SAPS2 scores.

Conclusions: We developed a predictive model that can effectively predict the in-hospital mortality of SALI patients, with satisfactory performance and clinical practicability. This model can assist clinicians in the early identification of high-risk patients and provide a reference for clinical treatment strategies.

Keywords: Sepsis-associated liver injury (SALI); hypoxic hepatitis (HH); nomogram; sepsis.

Figure 1

Inclusion and exclusion process. ICU, intensive care unit.

Figure 2

LASSO regression for feature selection. (A) Trajectories of change in coefficients for each variable in the lasso regression. (B) Min mean square error (left dotted line) and min distance (right dotted line) for the lambda values and number of variables. LASSO, least absolute shrinkage and selection operator.

Figure 3

Nomogram for predicting the in-hospital mortality of SALI patients. PTT, partial thromboplastin time; RDW, red blood cell distribution width; SALI, sepsis-associated liver injury.

Figure 4

ROC curves of the predictive model, SOFA and SAPS2 scores. (A) Training cohort; (B) validation cohort. ROC, receiver operating characteristic curve; AUC, area under the curve; SOFA, sequential organ failure assessment score; SAPS2, simplified acute physiology score II.

Figure 5

Calibration of the predictive model. (A) Training cohort; (B) validation cohort.

Figure 6

DCA curves of the predictive model vs. the SOFA, and SAPS2 scoring methods. (A) Training cohort; (B) validation cohort. DCA, decision curve analysis; SOFA, sequential organ failure assessment score; SAPS2, simplified acute physiology score II.