妊娠期高血压疾病风险预测模型的构建及价值分析OA

Objective To construct a risk prediction model for hypertensive disorders of pregnancy(HDP),and to analyse its value.Methods A retrospective analysis was conducted on the clinical data of 3 862 pregnant women with natural pregnan-cy and singleton pregnancy who underwent regular prenatal examination and delivery in Dongguan Kanghua Hospital from January to December 2021,and 268 cases were selected for inclusion in the study,which they were divided into control group(88 cases)and HDP group(180 cases)according to whether they had HDP.The differences in risk factors and serological indicators of HDP between the two groups were compared,statistical difference variables were screened,and multivariate logistic regression analysis was included to identify independent risk factors.Based on the regression results,the HDP prediction model was constructed,and the RO C curve was drawn to evaluate the efficiency.Results Univariate analysis showed that the age at due date,pre-pregnancy body mass index(BMI),uric acid multiple of median(MOM)value,alkaline phosphatase(ALP)MOM,cystatin C(Cys C)MOM,parity time,β-subunit of human chorionic gonadotropin(β-hCG)MOM,pregnancy associated plasma protein A(PAPP-A)MOM were related to the occurrence of HDP,the differ-ences were statistically significant(P＜0.05).Multivariate analysis using logistic regression model showed that th e age a t delivery(β=0.485,OR=1.624,95%CI:1.035-2.549)and pre-pregnancy BMI(β=0.258,OR=1.294,95%CI:1.194-1.403),uric acid MOM(β=0.380,OR=1.462,95%CI:1.149-1.861),ALP MOM(β=0.441,OR=1.554,95%CI:1.103-2.190),Cys C MOM(β=0.094,OR=1.099,95%CI:1.018-1.186)were independent risk factors for HDP,with yield(β=-0.868,OR=0.420,95%CI:0.235-0.748),PAPP-A MOM(β=-0.413,OR=0.662,95%CI:0.459-0.955),and β-hCG MOM(β=-0.436,OR=0.647,95%CI:0.441-0.948)were independent protective factors for its occurrence(P＜0.05).The prediction model was constructed Y=-7.654+age at expected delivery×0.485+Pregestational BMI×0.258+childbirth×(-0.868)+PAPP-A MOM×(-0.413)+β-hCG MOM×(-0.436)+uric acid MOM×0.380+ALP MOM×0.441 Cys C MOM×0.094.The Hosmer-Lemeshow test had good goodness of fit.The ROC results showed that the AUC of the prediction model for HDP occurrence was 0.908,the sensitivity was 0.840,the specificity was 0.880,and the approximate index was 0.720.Conclusion The HDP prediction model based on pre-pregnancy BMI,age,uric acid,ALP,Cys C,parity and serological indicators of PAPP-A and β-hCG has good discrimination and calibration.