Low birth weight among neonates: Investigating incidence,risk factors,and AI-enabled predictive modeling for risk estimation

Abstract

Background

Low birth weight serves as a vital measure of maternal health and the efficacy of prenatal care globally. The study was aimed to assess the incidence and risk factors of low birth weight among neonates. Further to develop a predictive model that identifies the risk factors contributing to low birth weight using artificial intelligence.

Methods

The study employed a dual research design, incorporating both descriptive and case-control methodologies. The data was analyzed using descriptive and inferential statistics. Further a predictive model was developed using logistic regression through artificial intelligence.

Results

The incidence rate of low-birth-weight babies was approximately 304.7 (30.47%) per 1000 live births. Logistic regression analysis identified significant risk factors for low birth weight (LBW), with notably high adjusted odds ratios (AOR). Key factors included inadequate weight gain during pregnancy <9 kg (AOR = 11.89, 95% CI: 6.03–23.44), gestational age <37 weeks (AOR = 12.81, 95% CI: 6.55–25.02), fetal complications reported during pregnancy (AOR = 13.25, 95% CI: 6.81–25.77), and multiple gestation (AOR = 26.88, 95% CI: 3.31–217.99). The developed AI-enabled predictive model demonstrates a high overall accuracy of 90%.

Conclusion

Most identified risk factors are modifiable, and early prenatal care can greatly reduce LBW incidence and improve neonatal outcomes. The predictive model demonstrated strong accuracy in classifying newborns by birth weight. Integrating the model into healthcare systems can aid early risk detection, reducing low birth weight and improving neonatal outcomes.