Establishment Of A Scoring System For Clinical Diagnosis Of Cerebral Venous Sinus Thrombosis

[Objective] To establish a scoring system for clinical diagnosis and prediction of cerebral venous sinus thrombosis(CVST),and to evaluate its diagnostic efficacy in order to identify CVST patients in early stage.In addition,to analyze the characteristics of SWI imaging.[Materials and Methods] We retrospectively collected patients' clinical data in the First Affiliated Hospital of Fujian Medical University.The case group are CVST patients from January 2012 to October 2016 in the Department of Neurology,all of them were confirmed by CTV,MRV or DSA.The control group are the cases of Intracranial Infection except Cryptococcus Meningoencephalitis from September 2015 to January 2016 in the Department of Neurology in our hospital,excluding those with serious chronic heart,brain,lung and other serious diseases and clinical data is not complete.Chose all cases from the case group and randomly selected 2/3 from the control group as the derivation cohort,and then randomly selected 40 cases from the case group and the rest 1/3 from the control group as the validation cohort.A prediction rule was developed from a Logistic multiple regression model by using a regression coefficient-based scoring method,and then was validated.The diagnostic efficiency of the scoring system was valued by its calibration,discrimination and accuracy.[Results] The case group included 85 cases of qualified research object,the control group included 148 subjects qualified,Chose all cases from the case group and randomly selected 2/3 from the control group as the derivation cohort,and then randomly selected 40 cases from the case group and the rest 1/3 from the control group as the validation cohort.Univariate analysis showed that 13 factors are the potential risk predictors in patients with CVST(P<0.05),including age,neck resistance,D-dimer,intracranial pressure,headache,having diseases which could lead to a state of hypercoagulable,the recent history of traumatic brain injury(within 1 months),with congenital disease,CT direct signs,CT indirect signs,MRI T1 WI showed increased signal of sinus,MRI T2 WI showed disappearing blood flow void of sinus,MRI SWI sequence of thrombus signal.Multivariate regression analysis showed that the scoring system is comprised 4 variables : headache,having diseases which could lead to a state of hypercoagulable,CT direct signs,CT indirect signs,MRI T2 WI showed disappearing blood flow void of sinus.The 5 variables of the beta based on value,the scores were 2 points,4 points,2 points,3 points and 4 points.ROC curve area of the scoring system in the derivation cohort and verification cohort were 0.805 and 0.826,Hosmer-Lemeshow goodness-of-fit test results show that the P values of the scoring system in the derivation cohort and verification cohort were 0.586 and 0.883.Score 3 was used as the predictive cut-off value.With the predictive cut-off value,the sensitivity,specificity,accuracy,positive predictive value,negative predictive value,positive and negative likelihood ratios critical value of CVST scoring system in the derivation cohort were 62.4%,83.8%,63%,76.8%,72.2%,3.85,0.45,in the verification cohort were 62.5%,89.7%,77.5%,83.3%,74.6%,6.07,0.42.[Conclusion] The clinical diagnostic scoring system has good accuracy and division,It can provide a basis for the identification of suspected CVST patients.