Analysis of risk factors for neonatal preterm birth and construction of nomogram prediction model
- Author:
XIANG Mei
;
LI Chuan-feng
;
ZHA NG Hong
;
YU Wei-hong
- Publication Type:Journal Article
- Keywords:
Neonatal preterm birth;
risk factors;
nomogram model
- From:
China Tropical Medicine
2023;23(6):563-
- CountryChina
- Language:Chinese
-
Abstract:
Abstract: Objective To analyze the risk factors for neonatal preterm birth in 12 hospitals in Yunnan Province from 2016 to 2017, and to establish a nomogram prediction model for neonatal preterm birth, providing scientific evidence for the prevention of preterm birth. Methods A total of 20 445 pregnant women who gave birth in 12 hospitals in Yunnan Province from 2016 to 2017 were collected and grouped into a preterm group (n=1 186) and a full-term group (n=19 259) according to whether they had a premature delivery. The general information questionnaire of pregnant women designed by the research team was applied to understand the basic conditions and pregnancy information of the two groups, and the risk factors of preterm birth were determined by logistic regression analysis, R software was applied to draw a nomogram prediction model of neonatal preterm birth, and its predictive performance was tested. Results There were significant differences in the proportions of twins and above (9.11% vs 7.10%), pregnancy-induced hypertension (21.67% vs 18.57%), gestational diabetes mellitus (18.21% vs 15.90%), anemia (24.28% vs 20.70%), premature rupture of membranes (11.64% vs 9.76%), and abnormal placenta (7.08% vs 5.51%) between the preterm group and the full-term group (χ2=6.731, 7.055, 4.441, 8.691, 4.437, 5.232, all P<0.05); the logistic regression analysis showed that the risk factors for neonatal preterm birth were twins and above (OR=2.378), pregnancy-induced hypertension (OR=2.039), gestational diabetes mellitus (OR=1.824), anemia (OR=1.825), and premature rupture of membranes (OR=2.313) (all P<0.05); the discrimination (area under the curve was 0.794, 95%CI=0.738-0.850) and precision (goodness of fit HL test, χ2=8.864, P=0.312) of the nomogram model constructed to predict the occurrence of neonatal preterm birth were both good. Conclusions The nomogram model for preterm birth constructed based on 5 factors including number of fetuses, pregnancy-induced hypertension, gestational diabetes mellitus, anemia and premature rupture of membranes can predict the occurrence of neonatal preterm birth well, thus providing reference for the prevention of neonatal preterm birth.
- Full text:1.Analysis of risk factors for neonatal preterm birth and construction of nomogram prediction model.pdf
