8 Augmenting or Replacing Obstetric Comorbidity Index with Labor & Delivery Features Improves Prediction of Non-Transfusion Severe Maternal Morbidity

The obstetric comorbidity index (OCI), a weighted aggregation of maternal comorbidities defined by ICD-10 code, has been validated in predicting non-transfusion severe morbidity (ntSMM). We sought to improve OCI using clinical features that are automatically extractable during labor and delivery. Retrospective analysis performed all women delivering four-hospital system from January 2016 through 2020. Patients were excluded if pre-delivery Hematocrit (Hct) was unknown. Demographic & data extracted an automated query the hospital electronic warehouse. score calculated per patient weights reported Leonard et al, 2020, denoting likelihood ntSMM. Logistic regression used predict ntSMM combinations features. An ‘Augmented OCI’ also which informatic including Hct < 31.5, Gestation Weeks 37, BMI>40 swapped for indicators: Anemia, Obesity, Preterm. Model accuracy assessed AUC. Of 60,034 deliveries, 4,754 (7.9%) due missing records. Characteristics 55,280 remaining deliveries shown Table 1, segmented whether patients had Two sets defined: L&D-OCI redundant with indicators, L&D-additional, additional not might models run these feature or Augmented (Table 2). Models plus extra (AUC=0.88) outperformed alone (AUC=0.85). model (AUC=0.85) similarly OCI. A only extractible (AUC=0.8) well. found augmenting L&D surpasses ntSMM, predictive suggesting decision support tool real-time risk prediction is attainable.View Large Image Figure ViewerDownload Hi-res image Download (PPT)