Accurate estimation of blood loss during cesarean deliveries: A secondary analysis of a randomized controlled trial comparing visual, quantitative and calculated approaches

Abstract

Introduction: Effective measurement of blood loss during delivery is key in timely diagnosis of hemorrhage and prevention of postpartum hemorrhage (PPH). Blood loss estimation in cesarean deliveries is challenging, with the risk of contamination of measured blood with amniotic and irrigation fluid. The objective of this study is to assess the level of agreement of visually estimated blood loss (vEBL), quantitative blood loss (QBL), and calculated estimated blood loss (cEBL) in cesarean deliveries.

Material and methods: This is a secondary analysis of a double-blinded, randomized controlled trial in the largest maternity unit in Singapore. Medical records from 200 patients enrolled in the prior study were analyzed, and their blood loss data reviewed for comparison. Blood loss estimation was assessed by vEBL (by the anesthetic and surgical teams), QBL (weighing of soiled gauzes and measuring fluid volume) and cEBL (formula-based calculation using pre- and postdelivery hemoglobin). Mean estimated blood losses (EBLs) obtained from all three methods were compared.

Results: The use of vEBL yielded the lowest mean blood loss, lowest proportion of women with EBL ≥500 and ≥1000 mL, while cEBL was the highest for all three outcomes. Intraclass correlation ranged from 0.29 (low <0.5) between vEBL and cEBL to 0.68 (moderate: 0.5-0.75) between vEBL and QBL. On average, vEBL was 249.7 mL (95% CI: -822.7-323.3) less than QBL, and 287.9 mL (95% CI: -1143.9-568.0) less than cEBL. Although vEBL tends to underestimate blood loss compared with QBL and cEBL on average, the wide confidence intervals suggest that these differences are not statistically significant. As blood loss increased, vEBL was more likely to underestimate blood loss. Women with body mass index (BMI) ≥30 kg/m² were more likely to have EBL ≥500 mL by cEBL (OR 1.13, 95% CI: 1.05-1.21, p < 0.01). Women with longer operative duration have higher odds of having EBL ≥500 mL by vEBL or QBL.

Conclusions: vEBL appears to grossly underestimate actual blood loss when compared with QBL and cEBL methods. Although the observed differences were not statistically significant, the wide confidence intervals suggest potential for substantial underestimation. Given this limitation, reliance solely on vEBL may lead to under-recognition and delayed management of PPH. Therefore, it is recommended that QBL and cEBL be incorporated into routine practice, particularly in high-risk cases. Clinicians should also account for factors that can influence EBL accuracy, such as operative duration and maternal BMI, when assessing blood loss and initiating interventions.

Keywords: cesarean delivery; estimated blood loss; obstetric labor complications; postpartum hemorrhage; puerperal disorders.

FIGURE 1

Bland–Altman plot of difference between vEBL and QBL against their mean.

FIGURE 2

Bland–Altman plot of difference between vEBL and cEBL against their mean.

FIGURE 3

Bland–Altman plot of difference between QBL and cEBL against their mean.