Combined multivariate prediction on survival of infants with congenital diaphragmatic hernia
裴锦丹,花晓琳,杨洁,万盛,伍岳琳,王俊,刘明,杨梦楠,王伟鹏,王磊
DOI: https://doi.org/10.3760/cma.j.cn113903-20210219-00148
2021-01-01
Abstract:Objective:To investigate the influencing factors of the survival in infants with congenital diaphragmatic hernia (CDH) and to develop a prediction model for CDH.Methods:Clinical data of 252 infants with a prenatal diagnosis of CDH in Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine from January 2010 to December 2018 were retrospectively analyzed. Adverse outcomes were defined as neonatal death on discharge. Chi-square and t test were used for univariate analysis of 16 perinatal and five postnatal risk factors. Regression analysis was used to determine the independent predictors of survival. The receiver operating characteristics (ROC) curves of the risk factors for predicting the survival of CDH were drawn. A prediction model based on the combination of risk factors for predicting adverse outcomes of CDH was established using the cut-off value. ROC curves of the model were drawn and the area under curve (AUC), sensitivity, and specificity were calculated. Results:Out of the 252 patients, 173(68.7%) survived on discharge. Univariate analysis showed that lung-to-head ratio (LHR), polyhydramnios, right diaphragmatic hernia, liver herniation, intrathoracic stomach, premature birth, low birth weight, 5 min Apgar score <7, and amniotic fluid index (AFI) were significantly associated with the survival of CDH. Regression analysis showed that polyhydramnios ( OR=11.19,95% CI:2.83-45.33), liver herniation ( OR=2.81,95% CI:1.32-11.92), intrathoracic stomach ( OR=5.02, 95% CI:1.29-17.13), low birth weight ( OR=8.58,95% CI:1.59-45.01) and AFI ( OR=3.68, 95% CI:1.37-14.72) were independent risk factors for survival at discharge in children with CDH, while LHR ( OR=0.36, 95% CI:0.01-0.69) were protective factors. The predictive cut-off values of LHR, polyhydramnios, liver herniation, intrathoracic stomach, low birth weight, and AFI were 1.6, 1.0, 1.0, 1.0, 1.0, and 232.5 mm, respectively. The model based on the combination of the above indicators for predicting CDH adverse outcomes was shown with an AUC value of 0.904, predictive sensitivity of 0.747, and specificity of 0.896. Conclusions:In this study, LHR, liver herniation, intrathoracic stomach, polyhydramnios, low birth weight, and AFI are independent risk factors for CDH survival. The combination of prenatal and postnatal indicators is noted for a higher accuracy for predicting CDH survival.