1.67 (95% CI, 1.20-2.40), and 6.50 (95% CI, 4.39-9.61), for the low-risk, medium-risk, and high-risk groups, respectively. In all 3 groups, the women who required RBC transfusion received a median of 2 U (range, 1-22 U). In decreasing order of predictive value the CMQCC risk factors associated with a significantly increased RR of having a significant hemorrhage were known coagulopathy, placenta previa, chorioamnionitis, multiple gestation, parity >4, uterine myomas, and previous uterine surgery. The 2 criteria that did not convey increased risk were fetal macrosomia and morbid obesity. Among the additional factors associated with an increased risk, but not included in the original CMQCC criteria (again in decreasing order of predictive value), were uterine rupture, a prediction of accreta or percreta, cesarean delivery, antenatal hypertension, and preterm delivery (P < 0.001 for these parameters). When patients were reclassified as high risk with any of these 5 factors, the new high-risk group included 85% of all women who had a significant peripartum hemorrhage; 2.6% of women in the new high-risk category had a significant peripartum hemorrhage.The clinical significance of the results remains uncertain because most women, even those in the high-risk group, did not require peripartum RBC transfusions. Unnecessary type and screens increase hospital charges and use valuable nursing and transfusion service resources. Pretransfusion testing is not required in women in the low-risk and medium-risk CMQCC groups but those in the high-risk category should undergo testing when they are admitted for delivery. The physician's clinical assessment is another critical factor in the determination of the need for testing.

II.