Construction of early warning model of surgical site infection in elderly pituitary tumor patients after transoasal butterfly approach
Objective To analyze the independent predictive factors of surgical site infection in elderly pituitary tumor patients after transnasal butterfly approach,and to construct a nomogram warning model to provide reference for preventing postoperative infection in elderly pituitary tumor patients.Methods Selected elderly patients undergoing endoscopic pituitary tumor resection from January 2020 to January 2023,Analysis of the clinical data,Infected(n=14)and non-infected(n=96).Univariate,multivariate logistic regression analysis was performed to determine independent risk factors for concurrent surgical site infection in elderly postoperative pituitary tumor patients.The clinical data of 47 patients with the same inclusion criteria in the same period were selected as the validation set(training set∶validation set=7∶3).R software was used to build the nomogram risk early warning model with parallel internal verification.The calibration and discrimination of the model were verified using the receiver operating characteristic(ROC)area under the curve(AUC),and the calibration map was used to verify the performance of the model.Results The probability of 110 elderly patients with complicated surgical site infection was 12.73％.The results of univariate studies showed surgical time,postoperative blood loss,intraoperative cerebrospinal fluid leakage,preoperative antimicrobial use,postoperative care type were the influencing factors of postoperative and concurrent surgical site infection in elderly pituitary tumor patients(P＜0.05).The results of the multivariate Logistic regression analysis showed that procedure time(＞240 min)(OR=11.534,95％CI=1.756～75.761,P=0.011),intraoperative blood loss(＞120 ml)(OR=9.648,95％CI=1.496～62.216,P=0.017),presence of intraoperative cerebrospinal fluid leakage(OR=6.75,95％CI=1.198～38.018,P=0.03),preoperative use of antimicrobial agents(OR=9.633,95％CI=1.547～59.982,P=0.015)was an independent risk factor affecting the occurrence of postoperative surgical site infection in elderly pituitary tumor patients(P＜0.05).Comprehensive postoperative care(OR=0.153,95％CI=0.026～0.893,P=0.037)was its protective factor.The nomogram model predicted the infection risk with the actual infection incidence,and internal validation showed that the model equation had a high degree of differentiation,ROC curve AUC of 0.870(95％CI=0.799～0.941,P＜0.01),sensitivity of 85.6％and specificity of 73.3％.Conclusion Elderly patients with postoperative surgical site infection by surgery(＞240 min),intraoperative bleeding(＞120 ml),intraoperative cerebrospinal fluid leakage and preoperative use of antimicrobial factors,postoperative comprehensive care can reduce the incidence of infection in patients,the risk warning model has good predictive value,can provide a basis for the prevention of infection.
Transnasal butterfly approachElderly pituitary tumor patientsPostoperative concurrent infectionEarly warning model
万方数据
维普
