Objective:To establish the prediction model for postoperative delirium in elderly patients with mild stroke undergoing non-cardiac and non-intracranial surgery.Methods:This was a nested case-control study. Seven hundred and fifty elderly patients of either sex with mild stroke, aged ≥65 yr, undergoing elective surgical procedures under general anesthesia in the Department of Gastrointestinal Surgery, Orthopedics and Urology at the Tangshan Workers Hospital from May to December 2023, were selected. The perioperative clinical data were collected. The incidence of postoperative delirium was assessed using the Confusion Assessment Scale 1-7 days after surgery or 1 day before discharge. The patients were assigned to the training set and the validation set in a ratio of 7∶3 using a simple random sampling method. Multivariate logistic regression was used to identify the risk factors for postoperative delirium, a postoperative delirium risk prediction model was established based on the risk factors, the nomogram was developed, and the receiver operating characteristic (ROC) curve, calibration curve and decision curve were plotted to assess the accuracy of the prediction model. The prediction model was verified using the validation set, and the calibration curve and ROC curve were plotted to assess the predictive performance of the model.Results:A total of 721 patients were finally included, and 108 patients developed postoperative delirium. Older age, high American Society of Anesthesiologists Physical Status classification, history of preoperative hypertension, short years of education, high preoperative Pittsburgh sleep quality index score, high preoperative National Institutes of Health Stroke Scale score, high intraoperative hypothermia, intraoperative hypotension and high postoperative numerical rating scale score were independent risk factors for postoperative delirium ( P<0.05). The area under the ROC curve of the training set prediction model was 0.996, with a sensitivity of 1.000, and specificity of 0.945. The slope of the calibration curve was close to 1, and the predicted risk of postoperative delirium was in good agreement with the actual risk. When the threshold probability of the decision curve was 0-0.9, the net return rate was higher than the null line. Validation set: In the calibration curve of the prediction model, the cohort and calibration curves were close to the ideal line, with an area under the ROC curve of 0.997, sensitivity of 1.000, and specificity of 0.962. Conclusions:Based on age, American Society of Anesthesiologists Physical Status classification, history of preoperative hypertension, years of education, preoperative Pittsburgh sleep quality index score, National Institutes of Health Stroke Scale score, intraoperative hypothermia and hypotension and postoperative numerical rating scale score, the prediction model for postoperative delirium is developed and has a good predictive performance in elderly patients with mild stroke undergoing non-cardiac and non-intracranial surgery.

Background/Aims: The utility of Baveno-VII criteria of clinically significant portal hypertension (CSPH) to predict decompensation in compensated advanced chronic liver disease (cACLD) patient needs validation. We aim to validate the performance of CSPH criteria to predict the risk of decompensation in an international real-world cohort of cACLD patients. Methods: cACLD patients were stratified into three categories (CSPH excluded, grey zone, and CSPH). The risks of decompensation across different CSPH categories were estimated using competing risk regression for clustered data, with death and hepatocellular carcinoma as competing events. The performance of “treating definite CSPH” strategy to prevent decompensation using non-selective beta-blocker (NSBB) was compared against other strategies in decision curve analysis. Results: One thousand one hundred fifty-nine cACLD patients (36.8% had CSPH) were included; 7.2% experienced decompensation over a median follow-up of 40 months. Non-invasive assessment of CSPH predicts a 5-fold higher risk of liver decompensation in cACLD patients (subdistribution hazard ratio, 5.5; 95% confidence interval, 4.0–7.4). “Probable CSPH” is suboptimal to predict decompensation risk in cACLD patients. CSPH exclusion criteria reliably exclude cACLD patients at risk of decompensation, regardless of etiology. Among the grey zone, the decompensation risk was negligible among viral-related cACLD, but was substantially higher among the non-viral cACLD group. Decision curve analysis showed that “treating definite CSPH” strategy is superior to “treating all varices” or “treating probable CSPH” strategy to prevent decompensation using NSBB. Conclusions Non-invasive assessment of CSPH may stratify decompensation risk and the need for NSBB in cACLD patients.