Quantitative blood loss (QBL) has been proposed to be more accurate than estimated blood loss (EBL) for management of obstetric hemorrhage. We sought to evaluate the effectiveness of QBL in identifying women with severe obstetric hemorrhage.METHODS:
We conducted a retrospective cohort study including all women who delivered at a tertiary center during 2/1/2013-1/31/2014. Medical charts were reviewed for relevant clinical data. Two methods were compared: quantitative blood loss and estimated blood loss. The data was analyzed using Spearman correlation and ROC curves where appropriate. P < 0.05 was considered significant.RESULTS:
During the study period there were 9,890 deliveries and 2.9% were complicated by obstetric hemorrhage (N=287). Among women with postpartum hemorrhage, 47.5% had a cesarean delivery and 22.8% had a cesarean after labor. Overall, 26.8% of the women required a blood transfusion. Women who required a blood transfusion compared to those who did not need one, had a lower admission hemoglobin and took longer to stabilize. QBL (r= 0.33, p<.001) but not EBL (r= 0.2, p=.08) was correlated with a drop in hemoglobin. QBL estimation (sensitivity 87%, specificity 67%) showed a small improvement over estimated blood loss in the prediction (66% sensitivity, 85% specificity) of women who would require a blood transfusion [AUC=.84 (95%CI 0.810 to 0.908) vs. AUC=.77 (95% CI 0.698 to 0.882)], but no difference in blood transfusion (28 vs 22 %, p=.34).CONCLUSION:
QBL has higher sensitivity in predicting severe obstetric hemorrhage but it was not associated with more frequent blood transfusions.