Construction and validation of a universal delirium risk prediction model for elderly surgical patients
ZHANG Dandan
GUO Yuanjun
TANG Shunluan
ZHOU Yuhua
ZHANG Ping
Abstract：Objective:To construct a universal prediction of postoperative delirium in elderly surgical patients(UPRE-PODES)and verify its predictive effect.Methods:A total of 1 250 patients aged≥60 years who visited the Department of Orthopedics,Gastrointestinal Surgery,Cardiothoracic Surgery,and Oncology Surgery of Shantou Central Hospital from May 1 to October 31,2022 were retrospectively selected as the research subjects.They were divided into a training set(n=833)and a validation set(n=417).A risk prediction model was established by Logistic regression,and its predictive value was verified.The regression coefficients of each independent influencing factor were converted into a simple scoring table.Results:The optimal critical value of the simple scoring table generated by the constructed UPRE-PODES was 5 points.When the model score was≥5 points,the patient was considered to be at risk of postoperative delirium(POD).The area under the receiver operating characteristic curve(AUROC)of the training set and validation set were 0.925 and 0.918,respectively.The χ2 values of the Hosmer-Lemeshow goodness-of-fit test were 10.797 and 3.361,respectively,both P>0.05.The sensitivity was 84.4%and 83.5%,respectively.The specificity was 87.3%and 84.3%,respectively.The prediction accuracy was 86.8%and 84.2%,respectively.Conclusions:The constructed UPRE-PODES has a good prediction effect and can be used in different surgical patients,which is conducive to the homogeneous management of POD.
Keywords：universalitydeliriumprediction modelelderly patientssurgerynursinginfluencing factors
Publication Date：2025-05-10
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：9( 1768-1776 )