Globally, preterm birth is a major public health issue, but if it can be predicted in expectant mothers ahead of time, it could help facilitate prompt intervention and reduce the number of preterm births. The purpose of this research was to create an easily accessible and implementable clinical model for predicting preterm birth. The case-control data utilized in this investigation were taken from the National Vital Statistics System (NVSS) between 2018 and 2019. To identify risk variables for premature birth, researchers used univariate and multivariate logistic regression analyses. These results were expressed as an odds ratio (OR) and a 95% CI. Model effectiveness was measured using the area under the curve AUC, precision, recall, and separation. There were a total of 3,006,989 pregnant women used in the 2019 model development and 3,039,922 in the 2018 model development and external validation. Premature births affected 324,700 (10.8%) of the 3,006,989 pregnant women. Preterm birth risk was lower among women who had completed at least a bachelor’s degree [bachelor (OR = 0.82; 95% CI, 0.81-0.84); master’s or above (OR = 0.82; 95% CI, 0.81-0.83)], who were not overweight or obese prior to pregnancy (OR = 0.96; 95% CI, 0.95-0.98), who received prenatal care (OR = 0.48; 95% CI, 0.47-0.50), and who were not primiparous In the validation set, the prediction model achieved an AUC of 0.688 (95% CI, 0.686-0.689) and an accuracy of 0.762 (95% CI, 0.762-0.763). In addition, a nomogram was developed to estimate a woman’s chance of having a premature baby based on data from both her and her partner’s pregnancies. The nomogram’s performance was satisfactory for predicting the risk of preterm delivery in pregnant women, and the relevant predictors are readily available in clinical settings, suggesting that the nomogram could serve as a useful and accessible tool for predicting preterm birth.
Source: bmcpediatr.biomedcentral.com/articles/10.1186/s12887-022-03591-w
