A proposed method to predict preterm birth using clinical data, standard maternal serum screening, and cholesterol

Area of Science:

• Reproductive Health
• Biomarkers
• Predictive Modeling

Background:

• Preterm birth (PTB) remains a leading cause of neonatal morbidity and mortality.
• Accurate prediction of PTB is crucial for timely intervention and improved outcomes.

Purpose of the Study:

• To develop a predictive model for preterm birth (PTB) using readily available clinical data and serum analytes.
• To assess the efficacy of serum screening markers in predicting PTB, independent of maternal characteristics.

Main Methods:

• A cohort of 2699 Iowa women had serum samples analyzed in the first and second trimesters.
• Clinical data, routine screening results, cholesterol levels, and health information were linked to birth certificate data.
• Stepwise logistic regression was employed to identify the optimal PTB predictive model.

Main Results:

• Serum screening markers were significant predictors of PTB, even after adjusting for maternal factors.
• The best predictive model incorporated maternal characteristics, first-trimester total cholesterol, inter-trimester cholesterol change, and second-trimester alpha-fetoprotein and inhibin A.
• This model demonstrated superior discriminatory ability compared to PTB history alone and performed consistently in women without prior PTB.

Conclusions:

• A potentially useful PTB predictor was constructed using clinical and serum screening data.
• Further validation and replication in diverse populations are recommended.
• Incorporating additional risk factors, such as cervical length, could enhance predictive capabilities and identify more women for intervention.