6374 papers
This study developed and externally validated a 72-hour landmark XGBoost model integrating static and dynamic clinical features that outperformed conventional severity scores in predicting in-hospital mortality for ICU patients with sepsis and pre-existing hypertension.
In a prospective cohort study of 7,137 Chinese adults aged 45 and older, a higher non-HDL to HDL cholesterol ratio was found to be independently associated with an increased risk of new-onset type 2 diabetes, an association that was not significantly mediated by systemic inflammation as measured by hs-CRP.
This study utilized unsupervised clustering of ten oral biomarkers in 200 community-dwelling older adults to identify a reproducible dysphagia-risk-enriched phenotype characterized by lower orofacial function and greater dryness, which showed significant associations with elevated dysphagia screening scores and aspiration risk.
This retrospective study developed and validated a high-accuracy nomogram model incorporating history of dysmenorrhea, infertility, fibrinogen levels, and lymphocyte count to effectively predict the risk of ovarian endometrioma in patients with endometriosis, thereby facilitating early identification and stratified clinical management.
This retrospective study reveals significant disparities in mental health service access between publicly and privately insured children, demonstrating that while pediatricians provide initial interventions, strengthening embedded collaborative care models within primary care is essential to bridge these inequality gaps.
This case report highlights fibrotic abdominal tuberculosis as a rare but critical cause of small bowel obstruction in a "virgin abdomen," emphasizing the importance of considering this diagnosis in endemic regions despite misleading radiological findings and the necessity of surgical intervention combined with anti-tubercular therapy for favorable outcomes.
This qualitative study of twelve regular electronic gaming machine users in New Zealand reveals that gambling-related harm is sustained by a complex interplay of structural, environmental, and psychological factors, underscoring the need for population-level public health strategies that address commercial and social determinants rather than relying solely on individual-focused harm minimization.
This multicenter validation study developed and validated a highly accurate, SHAP-interpretable Random Forest machine learning model using early clinical data to predict periventricular leukomalacia in preterm infants under 32 weeks gestation, achieving an AUC of 0.930 to facilitate early screening and targeted interventions.
This nationwide Thai cohort study of over 126,000 EMS-transported stroke patients identifies pre-hospital level of consciousness and blood pressure as the strongest predictors of emergency department outcomes, while highlighting significant regional disparities and higher mortality among uninsured and advanced life support-dispatched patients to underscore the need for equity-oriented EMS reforms.
This 2023–2025 retrospective study of a tertiary NICU in Western Ukraine reveals that neonatal sepsis is predominantly caused by multidrug-resistant organisms, with a notable prevalence of coagulase-negative staphylococci and an increasing burden of resistant Gram-negative pathogens that significantly complicates treatment and contributes to mortality.