Prediction of hemorrhage at cesarean delivery.

OBJECTIVE: To identify and quantitate the risk factors that might be predictive of hemorrhage during abdominal delivery. METHODS: Over a 2-year period, 1610 women underwent cesarean delivery and 127 (7.9%) had hemorrhage, defined as a decrease in hematocrit of 10% or greater, estimated blood loss greater than 1500 mL, or packed red blood cell administration. These women were compared through a case-control study design with the next abdominal birth without hemorrhage that could be matched for age, parity, indication for cesarean delivery, type of anesthesia, type of skin incision, and antepartum hematocrit. RESULTS: Preeclampsia (odds ratio 3.6, 95% confidence interval [CI] 1.8-7.4), disorders of active labor (odds ratio 4.4, 95% CI 1.4-13.7), Native American ethnicity (odds ratio 6.4, 95% CI 1.8-22.4), previous postpartum hemorrhage (odds ratio 8.4, 95% CI 1.9-37.4), and obesity of greater than 250 lb (odds ratio 13.1, 95% CI 1.7-102.7) were all statistically associated with significant bleeding during abdominal delivery. Combinations of two or more of these factors were associated with a markedly increased risk for hemorrhage, with odds ratios of 18.4 or greater. CONCLUSIONS: Patients undergoing cesarean delivery who have factors exposing them to increased risk of hemorrhage can be identified prospectively. These women will benefit greatly from extended preoperative counseling when possible, effective utilization of blood bank technology through type and cross-match requests, and preventive measures during abdominal delivery to minimize blood loss.

The effect of provider education on blood utilization practices.

OBJECTIVE: Our aim was to determine prospectively if a process of provider education and subsequent audit of transfusion criteria can reduce inappropriate blood product infusion. STUDY DESIGN: Beginning in December 1990, a year-long process of provider education and quality assurance audit, with the use of guidelines based on National Institutes of Health blood product consensus conferences, was instituted. After this orientation and education period, the medical records were reviewed for patients on the obstetrics and gynecology service from Dec. 1, 1990, through Sept. 30, 1991, who received blood products (packed red blood cells, fresh-frozen plasma, cryoprecipitate). This group was compared with patients cared for by our service over a similar period before the institution of these guidelines. The incidence of cesarean birth and operative intervention for gynecologic malignancies, as well as the number of major operative gynecologic procedures, was found to be unchanged during the total study period (Dec. 1, 1988, through Sept. 30, 1991). RESULTS: In 1989 there were 1236 units of blood products transfused; in contrast, in 1991 428 units of blood were transfused. In 1989, 107.9 +/- 59.9 units of packed cells per month were used in 33.6 +/- 16.8 patients (3.2 units per patient). In contrast, in 1991, 14.4 +/- 5.8 patients per month received transfusions (2.82 units per patient) with 40.7 +/- 17.2 units of packed cells (p < 0.0001). This represents a 75% decrease in the total number of packed cells and a 60% decrease in the number of patients undergoing transfusion per month. Similar reductions in the usage of cryoprecipitate and fresh-frozen plasma were noted (p = 0.024 and 0.002, respectively). Acute operative blood loss was the most common indication for transfusion. Abdominal hysterectomy was the most common procedure followed by exploratory laparotomy and cesarean section. No patients in whom blood was not used had untoward effects. CONCLUSION: Education as to appropriate blood utilization and concurrent quality assurance audit techniques can safely reduce blood usage on a busy obstetrics and gynecology service in a tertiary care center.