Daily Sepsis Research Analysis

OBJECTIVE: To develop and validate a machine learning model for postoperative sepsis in critically ill traumatic spinal injury (TSI) patients, a frequent and severe complication without dedicated predictive tools. METHODS: Model development used the MIMIC-IV 3.1 database, with external validation in the eICU-CRD 2.0 database and a Chinese TSI cohort. Variables documented within 24 h of postoperative ICU admission were screened using univariable testing and refined through Boruta and Group-Lasso regression to identify the final predictors. Thirteen base learners were trained and combined in a stacking ensemble optimized by fivefold cross-validation and hyperparameter tuning. Performance was assessed using receiver operating characteristic (ROC-AUC), average precision from precision-recall (PR-AP), calibration, decision, and lift curves, along with accuracy, sensitivity, specificity, precision, and F1 scores. Interpretability was evaluated through SHAP analysis. RESULTS: The development cohort comprised 808 patients, with 461 (57.1 %) sepsis cases, and the external validation cohort consisted of 358 patients, with 86 (24.0 %) events. Twelve predictors entered modeling, with the stacking model achieving an ROC-AUC of 0.918 and PR-AP of 0.938 in training and 0.889 and 0.936 in validation, maintaining close calibration, superior clinical utility confirmed by decision and lift curves, and balanced classification metrics, while most first-level models deteriorated markedly. External validation confirmed consistent performance and effective high-risk stratification. SHAP analysis underscored surgical burden, severity, hemodynamic, renal, and coagulation domains as key contributors, ensuring interpretability at cohort and individual levels. CONCLUSION: This first validated model for postoperative sepsis in critically ill TSI patients shows relatively robust performance and interpretability, enabling early risk stratification and supporting clinical decision-making.

BACKGROUND: Post-intensive care syndrome (PICS) encompasses new or worsening impairments in physical, cognitive, or mental health that manifest and persist after critical illness, yet the burden of PICS in this population remains under-characterized. We sought to quantify the incidence of PICS in sepsis survivors and to identify its independent predictors. METHODS: We conducted a prospective observational study of adult sepsis patients discharged from the medical ICU of Jinling Hospital, Nanjing University School of Medicine, between October 2024 and February 2025. Comprehensive baseline data-including demographics, clinical characteristics, and ICU interventions-were collected. Standardized telephone and in-person assessments were performed at 1, 3, and 6 months post-discharge to diagnose PICS across physical, cognitive, and mental health domains. Multivariate logistic regression was used to determine risk factors associated with incident PICS. RESULTS: Of 238 eligible patients, 150 met all inclusion criteria and completed follow-up assessment. The incidence of PICS was 72% at 1 month, 68% at 3 months, and 44% at 6 months. Multi-domain impairment (≥ 2 domains) was present in 38%, 34%, and 20% of survivors at the same timepoints, respectively; concurrent impairment in all three domains declined from 10.7% to 0.7%. After adjustment, higher APACHE II scores (OR = 1.57, 95% CI 1.27-1.94, P < 0.001) and elevated peak blood glucose (OR = 2.23, 95% CI 1.54-3.23, P < 0.001) independently predicted PICS, whereas higher education level (OR = 0.66, 95% CI 0.53-0.82, P < 0.001) conferred protection. CONCLUSIONS: PICS is common among sepsis survivors, with incidence declining but remaining substantial up to 6 months after ICU discharge. APACHE II scores and glycaemic control are robust, modifiable predictors, while educational attainment appears protective. TRIAL REGISTRATION: Registered in the Chinese Clinical Trial Registry ChiCTR2400094977 (Date: 31/12/2024).

Diabetes increases susceptibility to acute lung injury (ALI), yet the mechanisms linking hyperglycemia to pulmonary damage remain incompletely understood. Here, we demonstrate advanced glycation end products (AGEs)-metabolic byproducts elevated in diabetes-as promoters of ferroptosis that contribute to ALI pathogenesis. Clinical analysis of 170 patients with sepsis-related ALI showed that diabetic individuals had heightened inflammation and reduced PaO₂/FiO₂ ratios. Bioinformatic analysis revealed overlapping ferroptosis-related gene signatures between DM and ALI. In lipopolysaccharide (LPS)-induced ALI mice with diabetes mellitus (DM), reducing AGEs levels attenuated inflammatory cell infiltration and pro-inflammatory cytokine production, decreased Fe²⁺ accumulation and malondialdehyde (MDA) levels, and increased the expression of glutathione peroxidase 4 (GPX4) and solute carrier family 7 member 11 (SLC7A11). In vitro experiments suggested that AGEs exacerbate ferroptotic injury in LPS-treated bronchial epithelial (BEAS-2B) cells partly by suppressing AMP-activated protein kinase (AMPK)/acetyl-CoA carboxylase (ACC) signaling, an effect mitigated by pharmacological AMPK activation. These findings support a potential mechanistic link between DM and ALI through AGEs-driven ferroptosis and raise the possibility that targeting the AMPK/ACC pathway could offer therapeutic benefit.