Construction of a prediction model for preterm birth risk
10.19485/j.cnki.issn2096-5087.2024.08.005
- Author:
WANG Qiong
;
CHEN Danqing
;
WEI Yili
;
QIAN Fangfang
- Publication Type:Journal Article
- Keywords:
preterm birth;
prediction model;
demographic characteristics;
clinical characteristics
- From:
Journal of Preventive Medicine
2024;36(8):663-668
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct a prediction model for preterm birth risk among pregnant women, so as to provide the reference for screening high-risk population and preventing preterm birth.
Methods:Pregnant women who received antenatal examination and delivered at the Women's Hospital, School of Medicine, Zhejiang University from January 1 to December 31, 2019 were selected as the study subjects, among them, 80% were included in the modeling group, and 20% were included in the validation group. Demographic and clinical information were collected. A multivariable logistic regression model was used to analyze the predictive factors of preterm birth risk in the modeling group, and a preterm birth risk prediction model was established based on the OR values of predictive factors. The model was validated with the data from the validation group. The Youden index was used to determine the critical score for predicting preterm birth risk. The prediction performance of the model was evaluated using the receiver operating characteristic (ROC) curve.
Results:A total of 15 197 pregnant women were surveyed, including 12 131 pregnant women in the observation group and 3 066 pregnant women in the validation group. There was no statistically significant difference in age, education level and gravidity between the two groups of pregnant women (all P<0.05). Multivariable logistic regression analysis identified the number of pregnancies, education level, place of residence, hypertension, diabetes, history of preterm birth, twin-pregnancy, placenta praevia, and gestational hypertension as risk prediction factors for preterm birth risk among pregnant women. The risk score system for preterm birth was established as follows: >2 pregnancies (2 points), high school education or below (4 points), college degree or above (-4 points), rural residence (5 points), hypertension (7 points), diabetes (11 points), history of preterm birth (11 points), twin-pregnancy (28 points), placenta previa (19 points), and gestational hypertension (12 points). The total score of the preterm birth risk scoring system ranged from -4 to 99 points. When the critical score was 8 points, the Youden index was the highest at 0.480, with an area under the ROC curve of 0.749 (95%CI: 0.732-0.767), a sensitivity of 0.610, and a specificity of 0.886, indicating good prediction performance of the model.
Conclusion:The preterm birth risk prediction model established in this study based on demographic and clinical characteristics of pregnant women can effectively predict the risk of preterm birth among pregnant women.
- Full text:202409261635563955早产风险预测模型研究.pdf
