Comparison of logistic regression and machine learning methods for predicting postoperative delirium in elderly patients: A retrospective study.

AIMS: To compare the performance of logistic regression and machine learning methods in predicting postoperative delirium (POD) in elderly patients. METHOD: This was a retrospective study of perioperative medical data from patients undergoing non-cardiac and non-neurology surgery over 65 years old from January 2014 to August 2019. Forty-six perioperative variables were used to predict POD. A traditional logistic regression and five machine learning models (Random Forest, GBM, AdaBoost, XGBoost, and a stacking ensemble model) were compared by the area under the receiver operating characteristic curve (AUC-ROC), sensitivity, specificity, and precision. RESULTS: In total, 29,756 patients were enrolled, and the incidence of POD was 3.22% after variable screening. AUCs were 0.783 (0.765-0.8) for the logistic regression method, 0.78 for random forest, 0.76 for GBM, 0.74 for AdaBoost, 0.73 for XGBoost, and 0.77 for the stacking ensemble model. The respective sensitivities for the 6 aforementioned models were 74.2%, 72.2%, 76.8%, 63.6%, 71.6%, and 67.4%. The respective specificities for the 6 aforementioned models were 70.7%, 99.8%, 96.5%, 98.8%, 96.5%, and 96.1%. The respective precision values for the 6 aforementioned models were 7.8%, 52.3%, 55.6%, 57%, 54.5%, and 56.4%. CONCLUSIONS: The optimal application of the logistic regression model could provide quick and convenient POD risk identification to help improve the perioperative management of surgical patients because of its better sensitivity, fewer variables, and easier interpretability than the machine learning model.

Intraoperative allogeneic transfusion is associated with postoperative delirium in older patients after total knee and hip arthroplasty.

Objective: To determine whether intraoperative transfusion of allogeneic or autologous blood is associated with an increased incidence of postoperative delirium (POD) after total knee arthroplasty (TKA) and total hip arthroplasty (THA). Methods: The medical records of 1,143 older (≥65 years old) patients who received an intraoperative blood transfusion while undergoing total knee or hip arthroplasty at the First Medical Center of Chinese PLA General Hospital from 2014 to 2019 were reviewed; of these patients, 742 (64.92%) received allogeneic blood, while 401 (35.08%) received autologous blood. Patients who received autologous transfusion were paired with those received allogeneic transfusion using 1:1 propensity score matching method. The primary outcome was POD. The secondary outcomes were postoperative complications, including heart failure, deep vein thrombosis, myocardial infarction, stroke, and lung infection. Multivariable nominal logistic regression was used to identify any independent associations between intraoperative blood transfusions and POD, and secondary postoperative complications, respectively. Results: Postoperative delirium occurred in 6.6% (49/742) of patients who had received an allogeneic blood transfusion and in 2.0% (8/401) of patients who had received an autologous blood transfusion. It is noteworthy that the multivariable logistic regression demonstrated a significant association between intraoperative allogeneic blood transfusion and POD (odds ratio [OR]: 4.11; 95% confidence interval [CI]: 1.95-9.77; p < 0.001). After PSM, Allogeneic transfusion was also the strongest predictor for POD (OR: 4.43; 95% CI: 2.09-10.58; p < 0.001). Conclusions: In the patients who had received THA or TKA, intraoperative allogeneic blood transfusions were associated with an increased risk of POD.