Risk factor analysis and nomogram predictive model construction for neonatal hypospadias: a large-scale multicenter...

To conduct a retrospective multicenter study using objective data from a large sample size, which in order to identify reliable risk factors for hypospadias and establish a practical predictive model.

We retrospectively studied the medical records of the mothers who gave birth at Women's Hospital of Zhejiang University, School of Medicine, which was the obstetric treatment center in East China, and Xinchang Maternal and Child Health Hospital, and Quzhou Maternal and Child Health Hospital, from January 2019 to December 2021.

42809 male babies were born in the three hospitals, including 139 of which were diagnosed with hypospadias. The incidence of hypospadias was 0.325%. 33591 mothers and 35517 newborns were finally included to the statistical analysis. Among them, there were 121 newborns diagnosed with hypospadias and 121 mothers who gave birth to the hypospadias newborns. Premature birth, multiple pregnancies, gestational hypertension, fetal distress and abnormal placental morphology were set as the predictor to construct the nomogram model. The internal verification results displayed the area under the curve (AUC) was 0.834 (95% CI = 0.769-0.900), with a specificity of 85.6% and a sensitivity of 68.9%. The temporal verification results displayed the AUC was 0.857 (95% CI = 0.795-0.919), with a specificity of 74.2% and a sensitivity of 85.0%. The slope of the calibration curves were both close to 1, indicating high fitting degree of the model.

Our findings suggested that premature birth, multiple pregnancies, gestational hypertension, fetal distress and abnormal placental morphology were independent risk factors for hypospadias in newborns. We successfully established a nomogram predictive model for hypospadias, which can effectively, conveniently, clearly and visually predict the risk of hypospadias occurrence.