Construction Of Predictive Model Of Nomogram In Patients With Idiopathic Membranous Nephropathy

Objective:To establish a predictive model for predicting the prognosis of patients with idiopathic membranous nephropathy(IMN)by Nomogram diagram,so as to find the poor prognosis of IMN high-risk groups to give early intervention,to guide patient treatment,Thus improving the prognosis of IMN patients.Methods:We selected the newly diagnosed patients from The Affiliated Hospital of Chengde Medical College who was diagnosed IMN by renal biopsy from January 2018 to December 2020 and followed up for24 months.the patients were divided into two groups according to whether the patients had renal endpoint events at the end of follow-up.Patients who did not have terminal events at the end of follow-up were enrolled in the group that did not reach the renal endpoint.The influencing factors with p<0.2 in univariate Logistic analysis were analyzed by multivariate Logistic regression analysis,and the optimal Logistic regression model was selected according to Akaike information criterion(AIC)to construct the predictive model of poor prognosis of IMN patients.We used the area under the receiver operating characteristic(AUROC)to verify and evaluated the discrimination ability of the model,and the calibration curve and decision curve analysis(DCA)were drawn to evaluate the calibration,clinical net income and practicability of the model.Results:1.Comparison of general data between the two groups of IMN patients who met the renal endpoint and those who did notWe found gender,age,disease duration before renal puncture and MAP were statistically significant differences in between the two groups(P<0.05),By comparing the general information of the two groups of patients,We found Body mass index(BMI),smoking or alcohol consumption were no significant differencebetween the two groups(P>0.05).2.The laboratory results of IMN patients who met the renal endpoint group and those who did not were ComparisonThe laboratory results of the two groups of patients were analyzed,and it was found that the two groups of patients in 24hour-urinary protein(24h-UP)、 protein(TP)、 Albumin(Alb)、 estimated Glomerular Filtration Rate(e GFR)、 Blood urea nitrogen(BUN)、 Serum creatinine(Scr)differences were statistically significant(P <0.05),reaching the renal end point group of patients in 24h-UP、BUN、Scr was higher than those who did not,Total TP、Alb、e GFR was lower than those who did not.There were no significant differences in anti-phospholipase A2 receptor(anti-PLA2R)、 hemoglobin(Hb)、 total cholesterol(TC)、triglyceride(TG)and uric acid(UA)between two groups(P > 0.05).3.Risk factors associated with poor renal prognosis in IMN patientsUnivariate analysis showed that male,elderly(age > 60 years),long duration of disease(> 3 months),abnormal renal function(e GFR< 30ml·min-1·(1.73m2)-1),hyperalbuminuria(24h urinary protein volume ≥10g),low total protein,hypoproteinemia,high urea,and high total protein content.High serum creatinine was a risk factor for poor renal outcomes in patients with membranous nephropathy(P <0.05).Single factor analysis will be carried out.4.Construction of prediction modelMultivariate Logistic regression analysis was performed for the factors with P < 0.2 in the univariate analysis.The results showed that advanced age,mean arterial pressure,long disease course before renal puncture,albumin and serum creatinine were the influencing factors for poor prognosis of IMN patients.Logistic regression analysis was conducted with the above variables to form a regression model,and Nomgram histogram was obtained from the model to predict the probability of IMN patients reaching the renal function endpoint group within 2 years.5.Verification of prediction modelsROC curve analysis of the prediction model: the area under the curve of the prediction model was 0.729(95% : 0.641-0.818).The calibration ability of the prediction model: the quasi curve fitted the ideal curve well,and the Hosmer-Lemeshow goodness-fit testχ2=1.4364,P=0.4876,indicating that the model had better calibration ability.Decision analysis of the prediction model: This model is clinically applicable when the prediction probability of IMN is between 0.17 and0.44 in the DCA curve.Conclusion:In this study,a predictive model for predicting the poor prognosis of IMN patients was constructed.The predictive model has good predictive ability,calibration ability,and clinical net benefit.It can be used to predict the prognosis of IMN patients.