Daily Sepsis Research Analysis

IMPORTANCE: Timely identification of acute brain injury (ABI) in children receiving extracorporeal membrane oxygenation (ECMO) support is critical for early neuroprotective interventions. OBJECTIVES: To determine if elevations in plasma glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), and tau levels in children receiving ECMO precede new ABI confirmed by neuroimaging, and if they are associated with mortality and functional outcomes. DESIGN, SETTING, AND PARTICIPANTS: This was a prospective observational cohort study conducted from 2019 to 2023, with 18-month follow-up completed in 2025. Children aged 2 days to younger than 18 years at ECMO cannulation were recruited from 11 US children's hospitals. Study data were analyzed from May to August 2025. EXPOSURES: GFAP, NfL, and tau measured in plasma samples collected serially during the ECMO course. MAIN OUTCOMES AND MEASURES: Unfavorable short-term outcome was a composite of in-hospital mortality or discharge Pediatric Cerebral Performance Category score of 3 or greater with decline of at least 1 point from baseline. Unfavorable long-term outcome was a composite of mortality or Vineland Adaptive Behavior Scales, third edition, composite score less than 85 at 18 months after ECMO. RESULTS: This study included 219 participants (224 ECMO courses; 1089 serial blood samples). Median age was 11 months (IQR, 30 days-9 years), and 121 (54%) were male. Among 60 ECMO courses with new ABI during the ECMO course, GFAP and NfL levels increased significantly, by 6.4% (95% CI, 1.4%-11.6%) and 16.1% (95% CI, 10.5%-22.0%), respectively, for each 24 hours preceding neuroimaging diagnosis of new ABI. Geometric means for GFAP, NfL, and tau were all significantly higher in those with unfavorable vs favorable outcome at hospital discharge for both the first sample receiving ECMO and peak levels during ECMO support. A 2-fold increase in GFAP and NfL levels from first sample receiving ECMO was significantly associated with unfavorable outcome after adjusting for baseline GFAP and NfL levels, age, and ECMO indication (GFAP adjusted hazard ratio [aHR], 1.48; 95% CI, 1.22-1.79; NfL aHR, 1.43; 95% CI, 1.14-1.79). Similar models for tau showed no significant association with outcomes. CONCLUSIONS AND RELEVANCE: Results suggest that GFAP and NfL may be promising candidates for real-time neurologic monitoring in children receiving ECMO and may aid in diagnosis, association with outcomes, and potentially guiding neuroprotective strategies.

Bacteremia is a life-threatening complication and a leading cause of sepsis and septic shock in patients. Conventional diagnostic methods, such as blood culture, remain the clinical gold standard but require 24-72 h, often necessitating the empirical use of broad-spectrum antibiotics, which significantly contribute to the development and spread of antimicrobial resistance (AMR). We hypothesized that the host immune system mounts a specific, detectable systemic metabolic response to bloodstream infection, biochemically distinct from that elicited by focal bacterial infection (FBI) or viral etiologies. This study presents a rapid (<1 h), objective, and culture-independent diagnostic method for bacteremia based on host-response profiling using Fourier-transform infrared (FTIR) spectroscopy of white blood cells (WBCs). Blood samples from 410 pediatric oncology patients were clinically categorized into 71 bacteremia cases, 75 FBI, 157 viral infections, and 107 afebrile controls. WBCs were analyzed using FTIR spectroscopy to capture immune-metabolic fingerprints. Spectral profiles were classified using Logistic Regression with Principal Component Analysis (PCA) feature vectors and Log-Likelihood Ratio decision logic to differentiate bacteremia. The FTIR + Machine Learning (ML) platform successfully resolved the subtle biochemical differences, achieving 94.5% accuracy, 96.5% sensitivity, and 87.8% specificity in diagnosing bacteremia from all other categories combined (FBI, viral, and control). Importantly, the platform maintained high diagnostic performance, achieving 94.6% accuracy in distinguishing bacteremia from the FBI. This approach provides early, targeted diagnostic information that can support clinical decision-making, offering a powerful analytical tool to guide antibiotic stewardship and combat the global threat of AMR in this vulnerable population.

INTRODUCTION: Sepsis-associated acute kidney injury (S-AKI) is a severe complication in critically ill patients, linked to increased short-term mortality and chronic kidney disease. Existing prognostic tools like SOFA and SAPS II lack full representation of intricate clinical variable interactions. Machine learning (ML) models have potential in intensive care, but few validated and interpretable models focus on the in-hospital mortality rate of patients with S-AKI. This study aims to create and validate ML models for forecasting in-hospital mortality in S-AKI patients, identifying the most effective predictive model. METHODS: We conducted a retrospective analysis of data from the MIMIC-IV 3.0 database to identify adult ICU patients who met theSepsis-3.0(Sepsis-3 was defined as suspected infection with an acute increase in SOFA score ≥2) and KDIGO criteria for S-AKI. Additionally, a prospective cohort study from the General Hospital of Ningxia Medical University spanning 2023 to 2025 was included. Predictors recorded within 24 h of ICU admission included demographic information, comorbidities, vital signs, laboratory results, treatments, and severity scores. Variables with more than 20% missing data were excluded, and the remaining data were processed using interpolation. Feature selection was performed using the Boruta algorithm, and five machine learning models were trained (XGBoost, Random Forest, LightGBM, Decision tree, logistic regression). Model performance evaluation was based on metrics such as AUC, accuracy, sensitivity, specificity, F1 score, and clinical efficacy assessed through decision curve analysis. To enhance model interpretability, the SHapley Additive exPlanations (SHAP) method was employed. RESULTS: Among 16,800 patients with S-AKI, non-survivors(in-hospital mortality) exhibited older age, higher disease severity scores, more pronounced fluid overload, poorer renal function, metabolic acidosis, coagulation disorders, and heightened inflammatory responses. The XGBoost model demonstrated superior discriminative power (AUC 0.8799 ) in internal validation, surpassing other ML models, with exceptional sensitivity, accuracy, and F1 score. Decision curve analysis revealed that LightGBM offered the most significant net clinical benefit across various threshold probabilities. SHAP analysis consistently identified SAPS II score, AKI stage, oxygenation index, and key biochemical markers (e.g., serum sodium and blood urea nitrogen) as primary contributors to mortality risk, while the added value of basic demographic variables was limited. External validation confirmed that the XGBoost model has potential discrimination and robustness, highlighting the robustness and wide applicability of the machine learning-based prognostic framework. CONCLUSION: This study established an externally validated and interpretable ML model for riskstratification in S-AKI, enabling early identification of high-risk patients, personalized management strategies, and enhanced clinical outcomes in sepsis care.