Women’s Health
Applying WHO Prognostic Blood Loss Clinic Markers in a US-based Pregnancy Cohort Nelson Barba* Nelson Barba Barba University of Arizona
Background: Postpartum Hemorrhage (PPH) is the number one cause of mortality in childbirth globally and contributes significantly to pregnancy-related morbidity. In 2025, the WHO published new prognostic markers using vital sign data for predicting severe morbidity due to blood loss in pregnancy. WHO guidelines did not use data from US-based participants, and studies were primarily vaginal deliveries.
Methods: Using a cohort of 669 deliveries with data abstracted from medical records, we applied WHO guidelines regarding abnormal levels of blood loss, systolic & diastolic blood pressure, shock index, and heart rate. Diagnostic outcomes were pulled from abstracted data. Sensitivity (sens) and specificity (spec) were calculated for this cohort using the three diagnostic rules from WHO. Data was analyzed for the whole cohort and then by delivery type.
Results: In our study population (median age: 29 years (IQR: 25-33), 53.6% Hispanic, 74.3% vaginal delivery), the WHO rule 3 — 300 mL of blood loss with 1 abnormal clinical marker or a blood loss of 500 mL— showed a 91.4% (sens) & 43.3% (spec) for a diagnosis with PPH and it showed a 85.7% (sens) and 44.7% (spec) for symptoms of hypovolemia which were slightly short of the WHO goal of 80/50 for sens/spec. In vaginal deliveries, rule 3 achieved 87.5% 9 (sens) and 57.7% (spec), and for c-sections, it was 100% (sens) and 1.2% (spec). When the blood loss guideline levels were raised in the c-sections by 500 mL to 800/1000 mL, the specificity increased to 34.7%
Conclusion: This research suggests that the guidelines from the WHO may be practical to populations within the United States for vaginal births but not for c-sections. For c-section deliveries, 97.8% reached the higher blood loss threshold (500 mL), which adversely impacted the specificity. Data suggests that deliveries via c-section will require higher blood loss cutoffs, and additionally, they may require adjusted cutoffs for abnormal vital sign data.
