Supplementary figures for "A bedside risk score for 72-hour unplanned ICU readmission"

Résumé fourni par la source

# Supplementary figures — "A bedside risk score for 72-hour unplanned ICU readmission" Yıldız F, İnal V. Trakya University Faculty of Medicine, Department of Internal Medicine,Division of Intensive Care Medicine, Edirne, Turkey. Supplementary material for a single-centre retrospective cohort study (799 consecutivedischarge events from 573 adults, 1 Jan 2024 – 26 Dec 2025) that developed andinternally validated the Discharge Risk Score (DRS) for unplanned ICU readmissionwithin 72 hours. The main manuscript has been submitted for peer review; these figureswere produced from the same analytic dataset and are provided because the targetjournal does not publish supplementary material. ## Files | File | Content ||---|---|| Yildiz_FigS1_flow.jpg | Patient selection flow diagram (1,663 admissions → 799 analysed discharge events) || Yildiz_FigS2_forest.jpg | Forest plot of the six independent predictors (adjusted OR, 95% CI, log scale) || Yildiz_FigS3_scorecard.jpg | DRS bedside calculation card with point allocation and risk groups || Yildiz_FigS4_calibration.jpg | Calibration plot: observed vs predicted readmission by decile of predicted risk || Yildiz_FigS5_riskgroups.jpg | 72-h readmission, in-hospital and 30-day mortality by DRS risk group || Yildiz_FigS6_dca.jpg | Decision curve analysis: DRS vs APACHE-II, discharge SOFA, treat-all, treat-none | All figures: 300 dpi JPEG, greyscale-compatible. ## Model Logistic regression, six dichotomised predictors; intercept β0 = −3.613.Albumin < 2.5 g/dL (β 0.867, 2 points); lactate > 2.5 mmol/L (0.843, 2 points);discharge SOFA ≥ 4 (0.656, 1 point); age ≥ 75 y (0.628, 1 point);APACHE-II ≥ 20 (0.609, 1 point); after-hours discharge (0.601, 1 point). Score range 0–8.Apparent AUC 0.739 (95% CI 0.666–0.809); optimism-corrected AUC 0.701; calibrationslope 0.826 (1000 bootstrap resamples repeating the full model-building process). ## Ethics and data Approved by the Clinical Research Ethics Committee of Trakya University Faculty ofMedicine (decision no. 10/17); informed consent waived (retrospective design).Patient-level data are not deposited here; they are available from the correspondingauthor on reasonable request, subject to institutional approval. Corresponding author: Faruk Yıldız, MD — faruk__yildiz@hotmail.com

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.