A risk scoring model for predicting adverse maternal outcomes in pregnancy with pre-eclampsia.

OBJECTIVE

Pre-eclampsia can lead to mortality and morbidity. The aim of this study was to establish a multivariate model to predict the adverse maternal events associated with pre-eclampsia in the Thai population and to provide the composite measure of adverse perinatal events.

METHODS

A retrospective study was conducted at the Songklanagarind Hospital between January 2011 and December 2020. Multivariable logistic regression analysis was used to identify clinical risk factors for adverse maternal outcomes in pregnancies complicated with pre-eclampsia and to establish a predictive model. A nomogram was generated to determine the probability of adverse event in pre-eclampsia patient.

RESULTS

Of the 1094 cases of pre-eclampsia identified, 202 (18.46%) developed adverse outcomes. The majority of adverse consequences were cardiorespiratory complications (110 cases, 54.46%), intensive care unit admissions (66 cases, 32.67%), and hematologic disorders (50 cases, 24.75%). Maternal death occurred in four cases (0.37%). Significant risk factors included a history of drug abuse, dyspnea, visual disturbance, systolic blood pressure ≥180 mmHg, hemoglobin <10 g/dL, platelet count <150 000/μL, total bilirubin >0.8 mg/dL, aspartate aminotransferase >35 IU/L, lactate dehydrogenase >500 IU/L, INR >1, creatinine >1 mg/dL, urine protein/creatinine ratio >1 mg/mg, and uric acid >7.5 mg/dL. The regression model for prediction of adverse maternal outcomes in pre-eclampsia achieved a receiver operating characteristic area of 0.86 (95% confidence interval: 0.83-0.89). Pre-eclampsia patients with a total score of ≥3 had an increased risk of developing adverse maternal outcomes (sensitivity: 80%, specificity: 78%).

CONCLUSION

A new simplified scoring model and nomogram based on clinical characteristics and laboratory tests were developed to facilitate risk stratification of adverse maternal outcomes associated with pre-eclampsia in routine clinical practice.