Objective:To construct a prediction model for gestational diabetes mellitus (GDM) complicated with reproductive tract infections based on the decision tree algorithm, and explore prevention and control strategies for GDM complicated with reproductive tract infections.Methods:This cross-sectional study used convenience sampling to select 256 GDM patients treated at the Obstetrics and Gynecology Hospital, School of Medicine, Zhejiang University from January 2023 to June 2024 as the study subjects. Based on the occurrence of reproductive tract infections, the patients were divided into a non-infection group and an infection group. Multivariate Logistic regression analysis and the decision tree method were employed to create a prediction model for GDM complicated with reproductive tract infections, and the risk prediction performance of the model was evaluated.Results:Among the 256 GDM patients, the age ranged from 20 to 43 (33.42 ± 4.25) years. The incidence of reproductive tract infections was 18.75% (48/256). Multivariate Logistic regression analysis revealed that age, glycosylated hemoglobin>6%, history of reproductive tract infections, uncontrolled blood glucose, and frequency of sexual activity ≥2 times during mid-to-late pregnancy were independent risk factors for reproductive tract infections in GDM patients ( χ2 values were 2.08-4.24, all P<0.05). The decision tree model for GDM complicated with reproductive tract infections consisted of 4 layers and 17 nodes, with the nodes selecting five indicators: age, frequency of sexual activity during mid-to-late pregnancy, blood glucose control status, sharing of vulvar cleaning tools, and history of reproductive tract infections. Among these, age was the most significant predictor of reproductive tract infections in GDM patients. The area under the curve (AUC) value of the decision tree model for GDM complicated with reproductive tract infections was 0.948, while the AUC value of the multivariate logistic regression model was 0.927. There was no statistically significant difference between the AUC values of the decision tree model and the multivariate logistic regression model ( P>0.05). Conclusions:The constructed decision tree model for GDM patients developing genital tract infections has high application value and can serve as a basis for clinical screening of potential GDM patients at risk of genital tract infections.

Objective:To construct a prediction model for gestational diabetes mellitus (GDM) complicated with reproductive tract infections based on the decision tree algorithm, and explore prevention and control strategies for GDM complicated with reproductive tract infections.Methods:This cross-sectional study used convenience sampling to select 256 GDM patients treated at the Obstetrics and Gynecology Hospital, School of Medicine, Zhejiang University from January 2023 to June 2024 as the study subjects. Based on the occurrence of reproductive tract infections, the patients were divided into a non-infection group and an infection group. Multivariate Logistic regression analysis and the decision tree method were employed to create a prediction model for GDM complicated with reproductive tract infections, and the risk prediction performance of the model was evaluated.Results:Among the 256 GDM patients, the age ranged from 20 to 43 (33.42 ± 4.25) years. The incidence of reproductive tract infections was 18.75% (48/256). Multivariate Logistic regression analysis revealed that age, glycosylated hemoglobin>6%, history of reproductive tract infections, uncontrolled blood glucose, and frequency of sexual activity ≥2 times during mid-to-late pregnancy were independent risk factors for reproductive tract infections in GDM patients ( χ2 values were 2.08-4.24, all P<0.05). The decision tree model for GDM complicated with reproductive tract infections consisted of 4 layers and 17 nodes, with the nodes selecting five indicators: age, frequency of sexual activity during mid-to-late pregnancy, blood glucose control status, sharing of vulvar cleaning tools, and history of reproductive tract infections. Among these, age was the most significant predictor of reproductive tract infections in GDM patients. The area under the curve (AUC) value of the decision tree model for GDM complicated with reproductive tract infections was 0.948, while the AUC value of the multivariate logistic regression model was 0.927. There was no statistically significant difference between the AUC values of the decision tree model and the multivariate logistic regression model ( P>0.05). Conclusions:The constructed decision tree model for GDM patients developing genital tract infections has high application value and can serve as a basis for clinical screening of potential GDM patients at risk of genital tract infections.