(389) Antenatal Risk Stratification Using the Obstetric Co-Morbidity Index (OBCMI) during Prenatal Care

The obstetric comorbidity index (OBCMI) predicts risk for severe maternal morbidity (SMM) based on chronic and pregnancy-specific conditions and was devised from comorbidities present at the time of delivery admission. This study assesses the ability of an OBCMI calculated during antenatal care to predict risk of SMM.

Study Design:

This is an observational study of all deliveries > 20 weeks at 3 delivery hospitals within a large health system from 1/1/2022-6/30/2024. Patient demographic and clinical information was extracted from the electronic health record (EHR) and OBCMI scores were calculated at 3 time points based on the clinical conditions identified from the EHR during that encounter: (1) initial prenatal visit, (2) 32-week visit, and (3) delivery admission. Patient characteristics were evaluated in univariable analysis based on the presence of an OBCMI >3 at any of the 3 time points. Receiver-operator characteristic (ROC) curves were constructed and compared for OBCMIs calculated at the initial and 32-week visits versus the delivery admission OBCMI in terms of SMM prediction.

Results:

28,812 deliveries were included and 9,625 patients (33.4%) had an OBCMI > 3 at any point in pregnancy, with significant differences in all examined patient characteristics based on OBCMI score [Table]. The incidence of SMM was 1.8% (N=525). The prediction of SMM by OBCMI scores calculated during antenatal care at both time points was significantly inferior, with area-under-the-curve (AUC) of 0.59 (95% CI 0.57 – 0.61) for initial prenatal visit and 0.58 (95% CI 0.55 – 0.610) for the 32-week visit, compared to AUC of 0.81 (95% CI 0.79 – 0.83) for delivery admission OBCMI (p < 0.01 for all comparisons) [Figure].

Conclusion:

While the OBCMI reliably predicts SMM when calculated based on conditions documented at delivery, it performs poorly as a tool for antenatal risk stratification. This is likely multifactorial due to both EHR under-coding at antenatal visits and the development of clinically significant risk factors late in pregnancy. Better tools for maternal risk prediction prior to delivery are essential.