Background/aim: Acute gastrointestinal (GI) tract bleeding is a common and potentially life-threatening condition among patients presenting to emergency departments. In this study, we calculated AIMS65, Rockall, Glasgow-Blatchford Score (GBS), and Canada-United Kingdom-Australia (CANUKA) scores in patients with GI bleeding admitted to the emergency department and compared the sensitivity of these scoring systems in predicting the rates of admission to intensive care units and mortality. It is aimed to contribute to clinical practice and help determine an effective risk assessment tool in the management of patients with GI bleeding. Materials and methods: The study was conducted with patients who were diagnosed with GI bleeding. The study was conducted retrospectively between 1 January 2020 and 31 December 2023. The data of the patients were obtained from the hospital automation system. Patients with missing data were excluded from the study. AIMS65, Rockall, GBS, and CANUKA scores of the patients were calculated and recorded separately. Results: A total of 916 patients were included in our study. The median age was 70 years, and 62.3% of the patients were male. A total of 22.2% of the patients were hospitalized in the intensive care unit (ICU), and the in-hospital mortality rate was 0.9% (n = 8). According to the results of receiver operator characteristic (ROC) analysis of continuous measurements in terms of ICU hospitalization, the ability of the 4 scores to predict ICU hospitalization was statistically significant (p < 0.001). The CANUKA Score had the highest and best discriminative ability to predict ICU admission (area under the ROC curve [AUC] = 0.734). According to the results of ROC analysis of continuous measures in terms of mortality, the ability of AIMS65, CANUKA, and Rockall scores to predict mortality was statistically significant (p-values <0.001, <0.001, and 0.001, respectively). Conclusion: The CANUKA Score had the best discriminative ability in predicting intensive care unit admission and the best discriminative ability in predicting mortality after the AIMS65 Score.
1 October 2026 in Journal of Clinical and Analytical Medicine
Aim The Mayo Endoscopic Subscore (MES) in ulcerative colitis (UC) is limited by interobserver variability, whereas central reading is resource-intensive. We developed and temporally validated a video-based deep learning system for automated MES assessment. Methods This retrospective diagnostic accuracy study included 1,000 adults with UC who underwent white-light colonoscopy between November 2020 and November 2025. The model was developed in 750 patients using five-fold patient-level cross-validation and evaluated after model freezing in a temporally separated cohort of 250 patients. Two expert gastroenterologists assigned MES 0 to 3 from full-length videos, with third-reader adjudication of discordant cases. An EfficientNet-B0 model with attention-based multiple instance learning generated predictions. The primary outcome was four-class MES classification; dichotomous endpoints were MES 0 versus MES ≥1 and MES ≤1 versus MES ≥2. Results Pre-adjudication interobserver agreement was high (quadratic weighted κ, 0.84; 95% CI, 0.81–0.87). In temporal validation, macro-averaged AUROC was 0.95 (95% CI, 0.93–0.97), overall accuracy was 83.6%, macro-F1 score was 0.84, and quadratic weighted κ was 0.79 (95% CI, 0.68–0.87). AUROCs were 0.97 (95% CI, 0.95–0.98) for MES 0 versus MES ≥1 and 0.96 (95% CI, 0.94–0.97) for MES ≤1 versus MES ≥2. MES 1–2 errors accounted for 58.5% of classification errors. Brier scores were 0.07 and 0.06, respectively, and decision-curve analysis demonstrated a positive net benefit. Conclusions The model demonstrated high discrimination, good agreement with the adjudicated reference standard, good calibration, and interpretable errors. Multicenter external validation and prospective studies are required before implementation.
20 August 2026 in Akademik Gastroenteroloji Dergisi
Background and Aims: Acute lower gastrointestinal bleeding carries substantial risks of recurrent bleeding and mortality and remains a frequent reason for hospitalization. Most existing risk scores are designed to predict a single clinical outcome, limiting their usefulness for comprehensive early assessment. We aimed to develop and internally validate a pre-endoscopic risk score that concurrently predicts thirty-day recurrent bleeding and all-cause mortality in adults presenting with acute lower gastrointestinal bleeding. Materials and Methods: We conducted a retrospective cohort study of 3,000 adults randomly selected from 3,240 eligible patients managed at a tertiary-care center. Independent predictors available at initial presentation were identified using multivariable logistic regression analyses in a derivation cohort and used to construct a practical point-based clinical score. Model performance was then assessed through internal validation in a separate patient cohort. Results: The derived score comprises seven admission variables reflecting bleeding severity and patient vulnerability, including hemodynamic instability, advanced age, higher comorbidity burden, exposure to antithrombotic or anticoagulant therapy, lower hemoglobin, absence of abdominal pain, and lower albumin. In the validation cohort, the model demonstrated good discrimination for short-term mortality and moderate discrimination for recurrent bleeding. Patients classified as high risk experienced substantially higher rates of both outcomes than those categorized as low risk, supporting the clinical relevance of this stratification. Conclusion: This simple pre-endoscopic risk score enables early stratification of adults with acute lower gastrointestinal bleeding according to their simultaneous risks of recurrent bleeding and mortality. External validation is warranted to confirm generalizability and clinical utility.
19 May 2026 in Journal of Health Sciences and Medicine
Aims: Autoimmune pancreatitis (AIP) and immunoglobulin G4-related sclerosing cholangitis (IgG4-SC) can mimic pancreatobiliary malignancies, leading to diagnostic delay and malignancy misclassification. We evaluated phenotype-specific diagnostic delay and associated clinical consequences in a real-world cohort.Methods: In this retrospective cohort study (2019-2024), 68 adults with type 1 AIP and/or IgG4-SC were classified as combined AIP with IgG4-SC, isolated AIP, or isolated IgG4-SC. Diagnostic evaluation included cross-sectional imaging, cholangiography, endoscopic ultrasonography, histopathology, and corticosteroid response. Diagnostic delay, malignancy misclassification, and three harm domains-major iatrogenic harm, healthcare utilization burden, and documented psychosocial harm-were assessed.Results: Among 68 patients, 45.6% had combined AIP with IgG4-SC, 41.2% had isolated AIP, and 13.2% had isolated IgG4-SC. Diagnostic delay was longest in isolated IgG4-SC (median 210 days) compared with combined disease (85 days) and isolated AIP (70 days) (p
Background and Aim: Elevated liver enzymes are a common reason for hepatology referral, but etiological overlap complicates diagnosis. Most artificial intelligence applications address binary tasks, whereas routine practice requires multiclass discrimination. We developed and internally validated an explainable multiclass model to differentiate five major causes of persistent liver enzyme elevation. Materials and Methods: This prospective proof-of-concept cohort included 280 consecutive adults with unexplained persistent aminotransferase elevation. Final diagnoses—metabolic dysfunction-associated steatotic liver disease, viral hepatitis, alcohol-related liver disease, drug-induced liver injury, or autoimmune hepatitis—were independently adjudicated by three hepatologists. Data were split into training (70%) and independent test (30%) sets, and four supervised algorithms were evaluated. To assess incorporation bias, a basic model using routine clinical and biochemical variables was compared with an extended model including serological markers. Interpretability was assessed using SHapley Additive exPlanations. Results: eXtreme gradient boosting achieved the best test-set performance, with an overall accuracy of 0.86 (95% CI 0.78–0.93) and balanced accuracy of 0.87 (95% CI 0.82–0.92). Class-specific AUROC values ranged from 0.92 to 0.95. Serological markers improved discrimination for autoimmune hepatitis (AUROC increase, 0.15; p