Study Of Clinical Features And Diagnostic Optimization Of Neuropsychiatric Systemic Lupus Erythematosus

BackgroundNeuropsychiatric lupus erythematosus(NPSLE)is one of the most serious complications of systemic lupus erythematosus(SLE).The clinical manifestations of NPSLE are complex,and lack of characteristic markers.The specificity of the current classification definition standard is poor and the misdiagnosis rate is high.To study the clinical characteristics ofNPSLE,optimize the classification and diagnostic criteria of NPSLE is of great significance to improve the efficacy of diagnosis and reduce the clinical misdiagnosis and mistreatment.ObjectiveThe study aims to explore the clinical characteristics and risk factors of NPSLE and establish a diagnostic prediction model for NPSLE.MethodsPatients from January 2009 to April 2022 were selected if they had met the 1997 revised American College of Rheumatology(ACR)classification criteria for SLE.The clinical data of the SLE-related NP events group and the non-SLE-related NP events group were collected retrospectively.The clinical characteristics were compared between the two groups.All NP events were assessed by the research team to provide an attributable diagnosis.All the NP events were subjected to the Italian model which includes four items with scores from 0 to 10.The efficacy was evaluated by the area under the curve(AUC)of the receiver operating characteristic curve(ROC).The risk factors were included in the logistic regression and the regression equation for the predictive model was estabished.The predictive ability of the model was assessed by AUC under the ROC curve.SLE patients with NP events from May 2022 to October 2022 were selected as the validation group,and the predictive ability of this model was assessed using AUC under the ROC curve.ResultsA total of 2498 hospitalized SLE patients were retrospectively collected,including 455 patients with NP events,in which are 280 patients with SLE-related NP events and 175 patients with nonSLE-related NP events.There statistically significant differences in age,duration of hospital stay,SLEDAI score,initial diagnosis,course of SLE,NP events during hospitalization,Combined with blood system damage,hemoglobin count,platelet count,lymphocyte count,complement C3,rheumatoid factor,ANA titer,SSA/Ro60 antibody positive,anticardiolipin antibody IgM positive,anti-β 2-glycoprotein1 IgM titer,Paneth globulin positive,cerebrospinal fluid IgG,IgM,and IgA between the two groups(P<0.05).When the optimal cut-off value was≥6 points,the Italian model had a sensitivity of 70.40%and a specificity of 55.40%for the diagnosis of SLE-related NP events,with an AUC=0.67 under the ROC curve(P<0.001).Indicators with statistical differences in univariate analysis were included and entered into logistic regression model by stepwise regreesion method.The results showed that a higher Italian model score,younger age,higher SLEDAI score,and abnormal head MRM were independent risk factors of SLErelated NP events.When the cut-off value was 0.65,the sensitivity was 76.11%,the specificity was 78.38%,and the AUC under the ROC curve was 0.82(P<0.001).In validation,the logistic model showed a sensitivity of 63.41%and a specificity of 86.36%when the cut-off value was 0.74,and the AUC under the ROC curve of this model was 0.75(P<0.01).ConclusionsIt is suggested that higher score in The Italian algorithm,younger age,higher SLEDAI score,and abnormal head MRI are independent predictors for the diagnosis of SLE-related NP events in the logistic model,Valuable diagnostic predictive ability for NPSLE of the establised prediction model was confirmed in our clinical center.