自发性脑出血患者术后感染风险模型的构建与经济负担评价OA

Objective To analyze the risk factors for postoperative infection in patients with spontaneous intracerebral hemorrhage,to construct a postoperative infection risk model and to evaluate the associated economic burden,thereby providing a reference for developing clinical prevention and control strategies. Methods Patients with spontaneous intracerebral hemorrhage who underwent surgery at the First Affiliated Hospital of Nanyang Medical College from January 2021 to December 2022 were retrospectively and consecutively enrolled.Postoperative infection(including lower respiratory tract infection,intracranial infection,urinary tract infection,gastrointestinal infection,bloodstream infection,and surgical site infection)was diagnosed based on clinical manifestations,laboratory findings,and imaging results.Patients were divided into an infection group and a non-infection group according to the presence of postoperative infection.Univariate analysis and least absolute shrinkage and selection operator(LASSO)regression were used to screen factors influencing postoperative infection.These factors were then included in multivariate logistic regression to identify independent risk factors for postoperative infection,and a nomogram model for predicting postoperative infection was constructed.Internal validation of the model was performed using a Bootstrap method(1000 resamples).The model's predictive performance was assessed by plotting the ROC curve and calibration curve.Propensity score matching was used to analyze the direct economic burden associated with postoperative infection. Results A total of 299 patients were included,including 116 in the infection group and 183 in the non-infection group.Univariate analysis showed that diabetes history,intraventricular extension of hematoma,hematoma volume,preoperative GCS score,nasogastric feeding,tracheostomy,surgical procedure,use of mechanical ventilation,central venous catheter placement,indwelling urinary catheter,first white blood cell count,and first albumin level differed significantly between the two groups.LASSO regression selected seven predictive variables.Multivariate logistic regression further identified the following as independent risk factors:high first white blood cell count(OR 1.071,95％CI 1.008-1.142,P=0.032),low preoperative GCS score(OR 1.713,95％CI 1.234-2.398,P=0.001),and indwelling urinary catheter(OR 2.694,95％CI 1.139-6.983,P=0.031).Higher first albumin level was an independent protective factor(OR 0.917,95％CI 0.866-0.969,P=0.002).The risk of postoperative infection after external ventricular drainage via burr hole was significantly lower than that after decompressive craniectomy(OR 0.279,95％CI 0.082-0.812,P=0.026).A nomogram model based on the regression equation was developed to predict postoperative infection.The ROC curve showed an AUC of 79.0％(95％CI 73.6％-84.3％,P<0.001),the C-index was 0.79,the mean absolute error between predicted and observed values was 0.044,and the Hosmer-Lemeshow test yielded χ2=9.857(P>0.05),indicating good discrimination and calibration.After propensity score matching,74 matched pairs were successfully obtained.Compared with the non-infection group,the infection group had an additional direct economic burden of 26 522.93 yuan and a 17.0-day longer hospital stay. Conclusions Postoperative infection increased patients'economic burden and prolonged hospital stay.The nomogram model constructed in this study can effectively estimate the probability of postoperative infection.