Inflammatory burden index as a predictor of mortality in septic patients: a retrospective study using the MIMIC-IV database

BMC Infect Dis. 2025 Apr 17;25(1):552. doi: 10.1186/s12879-025-10936-0.

ABSTRACT

PURPOSE: Previous studies have identified the Inflammatory Burden Index (IBI) as a potential predictor of mortality risk in inflammatory diseases. However, its relationship with mortality rates specifically in septic patients has not been thoroughly investigated. This study aimed to explore the association between IBI and mortality risk in patients with sepsis.

PATIENTS AND METHODS: We sourced clinical records of 1,828 septic patients from the Multiparameter Intelligent Monitoring in Intensive Care IV (MIMIC-IV 3.0) dataset, covering the period from 2008 to 2022. The primary endpoint was mortality within 28 days, with secondary endpoints including mortality during intensive care unit (ICU) stays and throughout hospitalization. Patients were categorized into quartiles based on their log-transformed Inflammatory Burden Index (LnIBI) levels. Binary logistic regression was utilized to examine the independent influence of IBI on mortality outcomes, adjusting for confounders. Additionally, the association between IBI and these outcomes was explored using restricted cubic splines and Kaplan-Meier analysis. Further comparison of receiver operating characteristic (ROC) curves was conducted to investigate the predictive performance.

RESULTS: The study involved 1,828 septic patients, including 1,047 males. The all-cause mortality rates were 17.78% (325/1828) within 28 days, 17.34% (317/1828) during ICU stays, and 18.22% (333/1828) over the course of hospitalization. In the adjusted model, a positive correlation was found between LnIBI and mortality at 28 days (OR 1.093[1.014, 1.179], P = 0.021), during ICU stay (OR 1.106[1.025, 1.195], P = 0.01), and throughout hospitalization (OR 1.1[1.022, 1.187], P = 0.012). The analysis using restricted cubic splines showed a linear correlation between LnIBI and mortality risks. The areas under the curve (AUC) of LnIBI was larger than that of CRP (P < 0.05), and there were no significant differences between LnIBI and Neutrophil counts or Lymphocyte counts (P > 0.05). Kaplan-Meier plots revealed significantly lower survival rates for patients in the highest quartile of LnIBI (P < 0.001).

CONCLUSION: Elevated IBI values are significantly linked with higher mortality risks within 28 days, during ICU, and throughout the hospitalization period in septic patients.

PMID:40247163 | PMC:PMC12007314 | DOI:10.1186/s12879-025-10936-0