Accuracy of Methods Used to Diagnose Postpartum Hemorrhage in the First 24 Hours After Vaginal Birth

How sensitive and specific are tests used to diagnose postpartum hemorrhage in the first 24 hours after vaginal birth?

Visual estimation of blood loss is not very sensitive (22%–48%) but is very specific (97%–99%) when used to diagnose postpartum hemorrhage in patients who deliver vaginally in a hospital setting. In contrast, using a calibrated drape plus observations (eg, heart rate, blood pressure, uterine tone, blood flow) to measure blood loss has good sensitivity (93%) and specificity (95%) for detecting postpartum hemorrhage, which is defined as blood loss of 500 mL or more for primary postpartum hemorrhage.¹ (Strength of Recommendation: A, consistent, good-quality, patient-oriented evidence.)

Postpartum hemorrhage is defined by the American College of Obstetricians and Gynecologists as a cumulative blood loss of 1,000 mL or more or blood loss accompanied by signs or symptoms of hypovolemia within 24 hours after birth.²

According to the World Health Organization (WHO), maternal death is defined as the annual number of deaths from any cause (excluding accidental or incidental causes) of female patients during pregnancy or within 42 days of being pregnant that is related to or aggravated by pregnancy or the management of pregnancy.³ Postpartum hemorrhage remains the leading preventable cause of maternal illness and death worldwide.

According to the WHO, approximately 14 million women experience postpartum hemorrhage annually, resulting in approximately 70,000 maternal deaths globally.^4,5 In the United States, 11% of all maternal deaths are linked to post-partum hemorrhage.⁶ Quantifying blood loss in the postpartum period is critical for early recognition and treatment of postpartum hemorrhage. The authors of this Cochrane review aimed to assess the accuracy of methods used to diagnose post-partum hemorrhage.¹

The authors searched for studies that compared various diagnostic approaches for postpartum hemorrhage, including weighted blood loss measurements and other objective methods of quantifying blood loss, in patients delivering vaginally in any setting.¹ Primary postpartum hemorrhage was defined as blood loss of 500 mL or more within the first 24 hours of birth, and severe postpartum hemorrhage was defined as blood loss of 1,000 mL or more within the first 24 hours after birth. The authors identified 18 studies (N = 291,040), including 14 studies that evaluated primary postpartum hemorrhage and seven studies that evaluated severe postpartum hemorrhage. Most studies (16 of 18) were conducted in a hospital setting. Women of all ages were included.

Index test methods included the use of visual estimation, calibrated blood collection devices, the Signaling a Postpartum Hemorrhage Emergency (SAPHE) Mat, uterine assessment, clinical variables including heart rate and blood pressure, shock index, early warning charts, hemoglobin levels, and predelivery fibrinogen levels.¹ Reference standards included objective methods such as the gravimetric methods (weight-based blood loss measurement, including weighing collected blood, blood-soaked gauze, and sheets), calibrated devices to measure blood loss volume, the alkaline hematin method of blood loss estimation, and blood obtained using machine extraction and measured spectrometrically as oxyhemoglobin.

For detecting primary postpartum hemorrhage (ie, blood loss of 500 mL or more)¹:

• Visual estimation compared with gravimetric measurement of blood loss had a sensitivity of 48% (95% CI, 44%–53%; GRADE [Grading of Recommendations, Assessment, Development, and Evaluation] = moderate-certainty evidence) and a specificity of 97% (95% CI, 95% to 99%; GRADE = high-certainty evidence) in four studies with 196,305 patients.
• Visual estimation compared with volumetric measurement of blood loss had a sensitivity of 22% (95% CI, 12%–37%; GRADE = moderate-certainty evidence) and a specificity of 99% (95% CI, 97%–100%; GRADE = moderate-certainty evidence) in two studies with 514 patients.
• Calibrated drape plus observations (ie, physical examination findings) compared with gravimetric blood loss measurements had a sensitivity of 93% (95% CI, 92%–94%; GRADE = high-certainty evidence) and a specificity of 95% (95% CI, 95%–96%; GRADE = high-certainty evidence) in two studies with 53,762 patients.
• A hemoglobin level of less than 10 g/dL (100 g/L) compared with gravimetric blood loss measurement had a sensitivity of 37% (95% CI, 30%–44%; GRADE = high-certainty evidence) and specificity of 79% (95% CI, 76%–82%; GRADE = high-certainty evidence) in one study with 1,058 patients.