Daily Sepsis Research Analysis

BACKGROUND: Metabolic syndrome (MetS) is characterized by chronic low-grade inflammation and immune dysregulation, which may increase susceptibility to sepsis. However, epidemiologic evidence remains limited. This study aimed to evaluate the association of MetS with the risk of sepsis and sepsis-related mortality. METHODS: This study included 359,633 participants from the UK Biobank and 152,317 participants from the Kailuan Study. MetS was defined as the presence of ≥3 metabolic abnormalities. Multivariable Cox proportional hazards models were used to estimate hazard ratios ( RESULTS: During a median follow-up of 13.7 years, 11,040 sepsis cases were identified in the UK Biobank, whereas 5672 cases were documented in the Kailuan Study during a median follow-up of 16.4 years. After multivariable adjustment, MetS was associated with higher risks of sepsis ( CONCLUSIONS: This study demonstrated that MetS was associated with an increased risk of sepsis and sepsis-related mortality. These associations were partially mediated through inflammatory responses. The findings highlight the importance of maintaining metabolic health as well as promoting healthy lifestyles as strategies to reduce its burden.

JUSTIFICATION: Delirium is a frequent complication in critically ill and septic patients and has been linked to endothelial dysfunction, microvascular injury and blood-brain barrier disruption. Circulating endothelial cells may reflect endothelial phenotypic alterations beyond soluble markers. We investigated the association between endothelial subsets and postoperative sepsis-related delirium in ICU patients. OBJECTIVE: To investigate the role of circulating endothelial subsets in the development of delirium in post-surgical sepsis patients and their relationship with hypoperfusion and clinical outcomes, to identify potential prognostic biomarkers and mechanistic insights. METHODS: In this prospective cohort study, 214 postoperative ICU patients were enrolled at the time of surgery or sepsis diagnosis and classified as non-septic ICU (n=77), sepsis (n=61) or septic shock (n=76) according to Sepsis-3 criteria. Blood samples were obtained within 24 hours of critical illness onset. Circulating endothelial subsets were characterized using high-dimensional flow cytometry with unsupervised clustering. Delirium was assessed daily using the CAM-ICU. Cox regression, ROC analysis and causal mediation models were applied to evaluate associations with 28-day delirium and organ-dysfunction related clinical events occurring after sampling. RESULTS: Among 13 endothelial subpopulations identified, CD32b⁺ subset were independently associated with 28-day delirium (HR 2.41, 95% CI 1.32-4.40; p=0.004). CD32b⁺ subset demonstrated discriminative performance for delirium (AUC 0.79, 95% CI 0.60-0.98), which improved after adjustment for age and sex (AUC 0.89, 95% CI 0.82-0.98). Models based solely on organ-dysfunction related clinical events showed lower performance (AUC 0.69, 95% CI 0.52-0.86). Mediation analysis indicated that approximately 20% of the total effect was mediated through organ-dysfunction related events, suggesting partial mediation, while the remaining 80% may involve alternative endothelial and microvascular mechanisms not captured by conventional measures. CONCLUSIONS: Elevated CD32b⁺ subset are associated with postoperative delirium and organ-dysfunction related clinical events in critically ill patients, supporting an association between endothelial phenotypic alterations and vulnerability to brain dysfunction.

BackgroundAccurate prediction of short-term mortality in sepsis patients is critical for timely clinical decision-making. However, existing deep learning models often focus on static physiological parameters while neglecting the dynamic response to medical interventions, leading to risk underestimation due to the "masking effect" of therapeutic measures.MethodsWe propose a lightweight hybrid deep learning framework that integrates dynamic intervention responses to predict 24-h all-cause mortality. Utilizing the MIMIC-IV v3.1 database, we included 13,788 adult sepsis patients. The model employs a dual-branch architecture: a Bidirectional LSTM to capture local temporal trends and a Transformer Encoder to extract global long-range dependencies. Crucially, we constructed a high-resolution feature set that includes vasopressor infusion rates and hourly urine output to quantify physiological feedback to resuscitation.ResultsThe proposed model achieved an Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.8139, significantly outperforming seven mainstream baselines, including LightGBM (0.8015), Bi-LSTM (0.7870), and pure Transformer models (0.7704). Feature importance analysis revealed that indicators of treatment response, specifically urine output and norepinephrine dosage, were among the top predictive features, validating the clinical hypothesis that drug dependency and renal perfusion are sensitive markers of prognosis. Furthermore, external validation on the independent multi-center eICU Collaborative Research Database demonstrated robust generalizability: a zero-shot transfer yielded an AUROC of 0.6620, which improved to 0.7347 after lightweight domain adaptation fine-tuning, with a Negative Predictive Value (NPV) of 90.04%, confirming the model's cross-institutional applicability as a reliable rule-out tool.ConclusionOur LSTM-Transformer Fusion architecture effectively captures the complex "drug-physiology" interactions with low computational cost. By explicitly modeling the dynamic response to treatment and demonstrating cross-institutional generalizability through external validation on the eICU database, this lightweight model offers a robust and interpretable tool for early warning systems in resource-constrained intensive care environments.