Establishment Of The Nomogram For Predicting The Risk Of Intracranial Infection After Craniocerebral Surgery

Objective:This study retrospectively analyzed the Independent Risk Factors of intracranial infection after craniocerebral surgery and constructed a Nomogram.To clinically guide neurosurgical healthcare professionals to objectively and rapidly screen patients at high risk of intracranial infection after craniocerebral surgery and provide scientific guidance for the prevention and treatment of infection.Methods:According to the inclusion and exclusion criteria,clinical data of patients who underwent craniosurgery at the Department of Neurosurgery and Cerebrovascular Disease of Qingyuan People’s Hospital from May 2017 to April 2022 were collected by me.The main components include:(1)age,sex,hypertension,diabetes,long-term glucocorticoid therapy,radiotherapy,GCS score,diagnosis,open cranial injury;(2)date of surgery,name of surgery,timing of surgery,grade of surgery,classification of surgical incision,site of surgery,duration of surgery,amount of intraoperative bleeding,intracranial implants,re-operation;(3)Indwelling drains,leaking cerebrospinal fluid,frequent retention of cerebrospinal fluid and injectable drugs through drains,prolonged bed rest;(4)Intracranial infection,pathogenic examination,and regression of disease.The cases were divided into infected and non-infected groups according to whether or not a post-cranial intracranial infection was diagnosed.Statistical analysis of the data was carried out using R4.2.0 software.Independent risk factors for intracranial infection after craniocerebral surgery were screened by Univariate analysis,Lasso regression analysis and Multivariate analysis.Construction of a Nomogram based on Independent Risk Factors for intracranial infections after cranial surgery.Discrimination of Nomogram was evaluated using Area Under Curve(AUC)and C-index.The Hosmer-Lemeshow fit test,calibration curve and decision curve analysis curves were applied to evaluate the Fit,Calibration and Clinical utility of the Nomogram respectively.The final Dynamic Nomogram was drawn and a web version of the Dynamic Nomogram was created with the help of shinyapps.io software.Results:A total of 419 patients were studied and 53 patients underwent craniosurgery with a co-infection rate of 12.6% after intracranial infection.After Univariate analysis,Lasso regression analysis and Multivariate analysis,five Independent Risk Factors were derived.Includes: Duration of surgery(OR=1.004),intraoperative bleeding(OR=1.002),cerebrospinal fluid leakage(OR=5.004),reoperation(OR=2.452),frequent retention of cerebrospinal fluid through drains and injection of drugs(OR=3.941),(p<0.05).Based on the 5 Independent Risk Factors,I have successfully constructed the Nomogram.The C-index and AUC for Nomogram were both 0.826,indicating that Nomogram has a high Discrimination.The results of the Hosmer-Lemeshow good of fit test showed P > 0.05,indicating that the prediction was similar to the actual.The results of the calibration curve show that the Nomogram predictions agree with the actual results.The results of Nomogram’s Decision Curve Analysis curve show that Nomogram has a high Clinical utility.Based on Nomogram,I finally succeeded in creating a web version of Dynamic Nomogram.Conclusion:(1)Duration of surgery,intraoperative bleeding,cerebrospinal fluid leakage,reoperation,frequent retention of cerebrospinal fluid through drains and injection of drugs are Independent Risk Factors for intracranial infection after cranial surgery.Of these,cerebrospinal fluid leakage has the greatest impact on intracranial infection,and intraoperative bleeding has the least impact on intracranial infection.(2)Surgical site and prolonged bed rest are only Potential Risk Factors,not Independent Risk Factors.(3)The web-based version of the Dynamic Nomogram for this study has good Discrimination,Fit,Calibration,and clinical utility.The web version of Dynamic Nomogram is easy to use and its calculations are accurate and can be applied in clinical practice.