Predicting adverse pregnancy outcomes of pregnant mothers with syphilis based on a logistic regression model: a retrospective study

Abstract

Objective: Maternal syphilis could cause serious consequences. The aim of this study was to identify risk factors for maternal syphilis in order to predict an individual's risk of developing adverse pregnancy outcomes (APOs).

Methods: A retrospective study was conducted on 768 pregnant women with syphilis. A questionnaire was completed and data analyzed. The data was divided into a training set and a testing set. Using logistic regression to establish predictive models in the training set, and its predictive performance was evaluated in the testing set. The probability of APOs occurrence is presented through a nomogram.

Results: Compared with the APOs group, pregnant women in the non-APOs group participated in a longer treatment course. Course, time of the first antenatal care, gestation week at syphilis diagnosis, and gestation age at delivery in weeks were independent predictors of APOs, and they were used to establish the nomogram.

Conclusions: Our study investigated the impact of various characteristics of syphilis pregnant women on pregnancy outcomes and established a prediction model of APOs in Suzhou. The incidence of APOs can be reduced by controlling for these risk factors.

Keywords: APOs; adverse pregnancy outcomes; logistic (logit) regression; maternal syphilis; nomogram.

Figure 1

Nomogram prediction model for predicting the occurrence of APOs.

Figure 2

(A) Receiver operating characteristic curve for APOs prediction models in training set; (B) Calibration curve for the prediction model in the training set; (C) The decision curve of the model in the training set; (D) Receiver operating characteristic curve for APOs prediction models in testing set; (E) Calibration curve for the prediction model in the testing set; (F) The decision curve of the model in the testing set.