Factors associated with failure of trial of labor in primiparae with preeclampsia and establishment of risk prediction model / 中华围产医学杂志

ABSTRACT

Objective:To investigate the risk factors associated with failure of trial of labor in primiparous women with preeclampsia (PE) and to establish a risk prediction model.Methods:Primiparae with PE who underwent trial of labor in the Department of Obstetrics of Suzhou Ninth People's Hospital from February 2018 to July 2020 were retrospectively selected as the modeling set, and divided into two groups the success group and the failure group. Various parameters were compared between the two groups and those data with statistically significant difference were analyzed with multivariate logistic regression analysis. Those factors related to vaginal delivery failure in primiparous women with PE were identified. Based on the results, a risk prediction model was established using R language. Its performance was assessed with receiver operating characteristic (ROC) curve and goodness-of-fit test. This study also retrospectively enrolled primiparae with PE who underwent trial of labor in the same hospital from August 2020 to December 2021 as the validation set. Bootstrap method was used for verification and a calibration chart was created.Results:A total of 312 PE patients were selected as the modeling set with 89 in the failure group and 223 in the success group. Another 146 primiparae with PE were selected as the validation set. Logistic regression analysis showed that older age ( OR=1.609, 95% CI 1.251-2.483), higher body fat rate in early pregnancy ( OR=1.456, 95% CI 1.209-2.159) and higher ratio of umbilical artery systolic to diastolic flow velocity within a week before delivery ( OR=1.799, 95% CI 1.372-2.794) were risk factors for vaginal delivery failure in primiparae with PE, while more maternal education during pregnancy ( OR=0.233, 95% CI 0.054-0.672) and higher Bishop score ( OR=0.395, 95% CI 0.258-0.756) were protective factors. A nomogram model to predict the risks of vaginal delivery failure was constructed based on the above five factors. The area under the ROC curve (AUC) of the modeling set was 0.921 (95% CI 0.847-0.963) with the cut-off value of 0.213, and the corresponding sensitivity and specificity were 0.871 and 0.852, respectively;goodness-of-fit test showed that the observed values matched with those expected ( χ2=7.69, P=0.464); and the calibration curve indicated that the consistency of the prediction model was good. The AUC of the validation set was 0.903 (95% CI 0.835-0.942) with the sensitivity and specificity of 0.892 and 0.796, respectively; the discrepancy between the observed values and those expected was not significant as indicated by goodness-of-fit test ( χ2=6.82, P=0.512); calibration curve of the validation set showed that the predicted values of the model was consistent with the actual values. Conclusions:The failure of vaginal delivery in primiparae with PE is associated with maternal age, prenatal body fat percentage, ratio of fetal umbilical artery systolic to diastolic flow velocity within a week before delivery, maternal education during pregnancy and Bishop score. The nomogram model based on these five risk factors for prediction of vaginal delivery failure performs well.