Incidence proportion and risk prediction model development for preterm birth among pregnant women who had antenatal care follow-up at University of Gondar Comprehensive specialized Hospital

Abstract

Introduction:-Globally 13.4 million neonates were born in preterm period in 2020. Preterm complications are the leading cause of death in children under the age of 5 year. Estimating the probability of a pregnant woman at risk of preterm delivery is challenging in a resource limited settings. Available prediction models used unaffordable and inaccessible predictors and also missed important predictors. Objectives: - To determine Incidence proportion and develop risk prediction model for preterm birth among pregnant women who had antenatal care follow-up at University of Gondar Comprehensive specialized Hospital 1 2021 to June ,1 2022 Methods: - An institutional based retrospective follow-up study was conducted with a total of 1039 pregnant women who were enrolled from June 1, 2021 to June 1, 2022 at University of Gondar Comprehensive Specialized Hospital. Computer generated simple random sampling was used to select samples and the data was collected using Kobo toolbox and then exported in to Stata version-17 for analysis .Important predictors were selected by Least Absolute Shrinkage and Selection Operator and were entered to multivariable logistic regression. Statistically and clinically significant predictors after model reduction were used for the Nomogram development. Model performance was assessed by Area under the Receiver Operating Curve (AUROC) and calibration plot. Decision curve analysis was performed to evaluate the clinical and public health impacts of the model. Internal validation was done through bootstrapping method. Result: - The incidence proportion of preterm birth among pregnant women was 14.15% (95% CI: 12.03, 16.27). Antepartum hemorrhage, preeclampsia, polyhydroamnions, anemia, human immune virus, mean arterial blood pressure, premature rupture of membrane, and diabetic mellitus were significant in multivariable logistic regression of reduced model and were used to develop the Nomogram. The Nomogram had discriminating power AUROC of 0.79 (95% CI; 0.74, 0.83) and the calibration plots of the nomogram exhibited optimal agreement between the predicted and observed values, the Hosmer-Lemeshow test yielded a P-value of 0.602 The optimal cutoff value for the predicted probability was 0.12(Sensitivity; 0.70, Specificity; 0.69) .The Nomogram was internally validated by bootstrapping method with AUROC of 0.78(95% CI; 0.73, 0.82). Moreover, the decision curve analysis revealed that the nomogram would add net clinical benefits at the threshold probabilities less than 0.8(80%) x Conclusion and recommendation: the incidence proportion of preterm birth was high. The developed nomogram had good level of discrimination and well calibration, thus, using this model could help to identify pregnant women at a higher risk of having a preterm birth and providing an intervention like corticosteroid administration, nutritional support, antibiotic treatment in the event of infection, and other services.