Background: Suboptimal maternal health already from preconception onwards is strongly linked to an increased risk of birth complications.
Objective: To enable identification of women at risk of birth complications, we aimed to develop a prediction model for birth complications using maternal preconception socio-demographic, lifestyle, medical history and early-pregnancy basic clinical and biomarker characteristics in a general population.
Design/Methods: In a population-based prospective cohort study among 8,340 women, we obtained information on 35 maternal characteristics in the preconception period and first half of pregnancy (<21 weeks gestation). Preterm birth was defined as <37 weeks gestation. Small-for-gestational-age (SGA) and large-for-gestational-age (LGA) at birth were defined as gestational-age-adjusted birthweight in the lowest or highest decile, respectively. Because of their co-occurrence, preterm birth and SGA were combined into a composite outcome.
Results: The basic preconception model consisted of easy obtainable maternal preconception characteristics including age, ethnicity, parity, body mass index and smoking. This model had an area under the receiver operating characteristics curve (AUC) of 0.63 (95% confidence interval (CI) 0.61; 0.65) and 0.64 (95% CI 0.62; 0.66) for preterm birth/SGA and LGA, respectively. Extension to complex models by adding maternal socio-demographic, lifestyle, medical history and early-pregnancy basic clinical and biomarker characteristics led to a small, though significant, improvement of models. The full model for prediction of preterm birth/SGA had an AUC 0.66 (95% CI 0.64; 0.67) with a sensitivity of 22% at 90% specificity. The full model for prediction of LGA had an AUC of 0.67 (95% CI 0.65; 0.69) with a sensitivity of 28% at 90% specificity. The developed prediction models had a reasonable level of calibration in different socio-economic subsets of our population. In additional analyses, we observed that only for LGA paternal characteristics slightly improved model performance. The models were good at predicting the risk of fetal distress, caesarian section and low Apgar score. Conclusion(s): In preconception, maternal characteristics easy obtainable in clinical practice are well able to predict birth complications. When validated, the proposed prediction model could be used as a screening tool to identify women at risk of birth complications on a population level already from preconception onwards.
Authors/Institutions: Rama J. Wahab, Erasmus MC, Rotterdam, , Netherlands