(327) Widely Used, but Does it Deliver? A Closer Look at the AWHONN PPH Risk Score

Objective:
To evaluate the association between a widely used postpartum hemorrhage (PPH) risk assessment tool and clinically relevant hemorrhage outcomes, including blood loss, transfusion, and hemoglobin decline.

Study Design:
This is a retrospective cohort study of all deliveries at a tertiary care center between January 2024 and July 2025. Patients were stratified into low, moderate, or high-risk categories using our electronic medical record (Epic Systems) based on the Association of Women’s Health, Obstetric, and Neonatal Nursing (AWHONN) PPH risk score. We compared the following outcomes across risk categories: estimated blood loss (EBL) ≥ 1000 mL, quantified blood loss (QBL) ≥ 1000 mL, any transfusion, and transfusion > 1 unit. We also performed a subgroup analysis of hemoglobin decline in patients with a documented postpartum hemoglobin value who did not receive a transfusion. Discriminative performance was evaluated using the area under the receiver operating characteristic curve (AUC).

Results:
Among 5,540 deliveries, increasing PPH risk category was significantly associated with EBL and QBL ≥ 1000 mL, as well as increased rates of any transfusion (Table). However, discriminative performance for these outcomes was poor (AUCs: 0.60, 0.61, 0.63, and 0.69, respectively). In a subgroup of 1793 patients without transfusion and with documented postpartum hemoglobin, median hemoglobin drop was significantly greater in the low and high-risk groups compared to the moderate group (Figure). Although hemoglobin drop ≥ 2 and ≥ 3 g/dL were significantly associated with risk category, the score failed to discriminate these outcomes better than chance (AUC 0.47 for both).

Conclusion:
PPH risk assessment based on AWHONN scoring was statistically associated with blood loss, transfusion, and hemoglobin decline; however, the score’s ability to predict adverse outcomes is poor. Despite statistical significance, the pattern of hemoglobin decline across risk categories is not clinically meaningful. These findings suggest that outcome-based validation of hemorrhage risk tools is urgently needed.