老年重症肺炎患者并发急性胃肠功能损伤风险预测模型的构建OA

Objective To identify risk factors for acute gastrointestinal injury(AGI)in elderly patients with se-vere pneumonia and to develop and validate a nomogram prediction model.Methods A total of 300 elderly patients with severe pneumonia admitted between March 2023 and March 2025 were retrospectively enrolled and randomly divided into a training cohort(n=150)and a validation cohort(n=150).In the training cohort,patients were further classified into AGI(n=97)and non-AGI groups(n=53)based on the occurrence of acute gastrointestinal injury.Baseline clinical data were analyzed.Univariate analysis was used to identify potential risk factors,followed by least absolute shrinkage and selection operator(LASSO)regression for variable selection.Multivariable logistic regression was then performed to deter-mine independent predictors.A nomogram model was constructed using R software,and its performance was evaluated by discrimination,calibration,and clinical utility.Results The incidence of AGI in the training cohort was 64.67％(97/150).Significant differences were observed between the two groups in stool characteristics,white blood cell(WBC)count,intestinal dysbiosis,APACHE Ⅱ score,SOFA score,use of vasoactive drugs,and mechanical ventilation(P＜0.05).Multivariable analysis identified elevated WBC count,intestinal dysbiosis,higher APACHE Ⅱ score,higher SO-FA score,and mechanical ventilation as independent risk factors for AGI(P＜0.05).The nomogram demonstrated excel-lent predictive performance,with a concordance index(C-index)of 0.935(95％CI:0.895-0.975).Calibration curves showed good agreement between predicted and observed outcomes.Decision curve analysis indicated that the model provided a net clinical benefit when the threshold probability was greater than 0.15.The Hosmer-Lemeshow test(x2=12.581,P=0.127)confirmed good model calibration.Conclusion Elevated WBC count,intestinal dysbiosis,higher APACHE Ⅱ and SOFA scores,and mechanical ventilation are independent risk factors for AGI in elderly patients with se-vere pneumonia.The developed nomogram exhibits strong discrimination and calibration,suggesting its potential utility in clinical risk stratification.