An interpretable machine learning model for predicting in-hospital mortality in ICU patients with ventilator-associated pneumonia

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(1), P. e0316526 - e0316526

Published: Jan. 7, 2025

Background Ventilator-associated pneumonia (VAP) is a common nosocomial infection in ICU, significantly associated with poor outcomes. However, there currently lack of reliable and interpretable tools for assessing the risk in-hospital mortality VAP patients. This study aims to develop an machine learning (ML) prediction model enhance assessment Methods extracted patient data from versions 2.2 3.1 MIMIC-IV database, using version training validation, external testing. Feature selection was conducted Boruta algorithm, 14 ML models were constructed. The optimal identified based on area under receiver operating characteristic curve (AUROC), accuracy, sensitivity, specificity across both validation test cohorts. SHapley Additive exPlanations (SHAP) analysis applied global local interpretability. Results A total 1,894 patients included, 12 features ultimately selected construction: 24-hour urine output, blood urea nitrogen, age, diastolic pressure, platelet count, anion gap, body temperature, bicarbonate level, sodium mass index, whether combined congestive heart failure cerebrovascular disease. random forest (RF) showed best performance, achieving AUC 0.780 internal 0.724 testing, outperforming other clinical scoring systems. Conclusion RF demonstrated robust performance predicting developed online tool can assist clinicians efficiently risk, supporting decision-making.

Language: Английский

---

IL-27 aggravates acute hepatic injury by promoting macrophage M1 polarization to induce Caspase-11 mediated Pyroptosis in vitro and in vivo

Cytokine, Journal Year: 2025, Volume and Issue: 188, P. 156881 - 156881

Published: Feb. 5, 2025

Language: Английский

---

Aspirin is associated with improved 30-day mortality in patients with sepsis-associated liver injury: a retrospective cohort study based on MIMIC IV database

Frontiers in Pharmacology, Journal Year: 2025, Volume and Issue: 16

Published: March 4, 2025

Background Sepsis-associated liver injury (SALI) is a common complication in sepsis patients, significantly affecting their prognosis. Previous studies have shown that aspirin can improve the prognosis of septic patients. However, there currently lack clinical evidence supporting use treatment SALI. Therefore, we conducted this study to explore association between and patients with Methods The were obtained from Medical Information Mart for Intensive Care IV (MIMIC-IV) database, version 3.0. primary outcome was 30-day all-cause mortality. Baseline characteristics non-aspirin groups balanced using propensity score matching (PSM). Kaplan-Meier survival curve Cox regression analysis used investigate Results Of 657 SALI study, 447 (68%) had not during hospitalization, whereas 210 (32%) had. After PSM, mortality 33.1% group 21% group, indicating reduced risk (HR, 0.57; 95% CI, 0.37–0.90; P = 0.016). Similarly, results multivariable inverse probability weighting (IPW) showed that, compared lower (Multivariable analysis: HR, 0.69; 0.48–0.99; 0.047; IPW: 0.62; 0.43–0.89; 0.010). Conclusion Aspirin reduce regardless dose or timing administration. careful assessment based on individual differences essential ensure safety effectiveness use.

Language: Английский

---

METTL3-mediated m6A modification in sepsis: current evidence and future perspectives

Epigenomics, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 13

Published: April 19, 2025

Sepsis, a severe systemic inflammatory condition triggered by infection, is associated with high morbidity and mortality worldwide. While medical diagnosis treatment have advanced in recent years, specific therapy remains unavailable. Recently, significant progress has been made studying the epigenetic RNA modification N6-methyladenosine (m6A) its core methyltransferase METTL3. The role of m6A sepsis also increasingly elucidated. This review aims to explore pathological mechanisms relationship m6A, focusing on key writer, METTL3, sepsis.

Language: Английский

---

Immunologic role of macrophages in sepsis-induced acute liver injury

International Immunopharmacology, Journal Year: 2024, Volume and Issue: 143, P. 113492 - 113492

Published: Oct. 30, 2024

Sepsis-induced acute liver injury (SALI), a manifestation of sepsis multi-organ dysfunction syndrome, is associated with poor prognosis and high mortality. The diversity plasticity macrophage subpopulations explain their different functional responses in diseases. Kupffer macrophages, capsular monocyte-derived macrophages are involved pathogen recognition clearance the regulation inflammatory responses, exacerbating progression SALI through pathways pyroptosis, ferroptosis, autophagy. Concurrently, they play an important role maintaining hepatic homeostasis repair processes SALI. Other recruited to diseased tissues under pathological conditions polarized into various phenotypes (mainly M1 M2 types) influence signaling molecules, transcription factors, metabolic reprogramming, thereby exerting roles functions. This review provides overview immune discusses multiple provide reference for future studies.

Language: Английский

---

Association between red cell distribution width and 30-day mortality in patients with sepsis-associated liver injury: a retrospective cohort study

Frontiers in Medicine, Journal Year: 2024, Volume and Issue: 11

Published: Dec. 18, 2024

Sepsis-associated liver injury (SALI) is a critical component of sepsis-induced multiple organ dysfunction with high mortality. Identifying biomarkers for risk stratification essential. Red cell distribution width (RDW), indicating variation in red blood volume, has been linked to adverse outcomes various diseases. This study aimed evaluate the association between RDW and 30-day mortality SALI patients.

Language: Английский

---