Even though hospitalization for delivery and the availability of blood for transfusion have dramatically reduced maternal mortality during the 20th century, obstetric hemorrhage still remains a leading cause of maternal death. For example, in 2005 in the United States, hemorrhage was the third leading cause of maternal death due to obstetric factors.1 The common causes of maternal death from hemorrhage include placental abruption, lacerations of the birth canal, uterine rupture, uterine atony, and placental implantation disorders such as previa or accreta.2 Such pregnancy complications have made it imperative that obstetric care include the capability of prompt administration of blood products.
The history of the use of blood products in the management of hemorrhage of any cause, to include obstetrics, is largely a result of U.S. military experience during World War II.3 When the United States entered the war, the military embraced freeze-dried human plasma as the primary transfusion product. However, casualties resuscitated primarily with plasma had poorer outcomes than expected, prompting the Army to use whole blood as the agent of choice in the resuscitation of battle casualties. Civilian blood banking was born and expanded rapidly during this time to fill the military’s need. After World War II, the development of whole blood fractionation techniques prompted the concept that blood could be used more effectively if the donated blood were separated into packed red blood cells, platelet concentrates, fresh frozen plasma, and cryoprecipitate. Separate components were felt to maximize the potential for use of each donated unit. As a result, the use of whole blood fell out of favor in the civilian community. For example, whole blood is not even mentioned in the most recent clinical management guidelines on obstetric hemorrhage disseminated by the American College of Obstetricians and Gynecologists.4
The development of the Parkland Hospital Obstetrics Service during the past 50 years has been considerably influenced by research in the management of obstetric hemorrhage. Jack A. Pritchard, MD,5–8 the first chief of service, published extensively on the hematologic problems associated childbirth and especially placental abruption. Central to those studies was the use of whole blood to maintain renal perfusion in the presence of severe maternal hemorrhage. He developed transfusion guidelines for management of women with hypovolemia to maintain urine flow at 30 mL/h and the hematocrit at 25–30% using transfusion of whole blood. This guidance has persisted until the present, and whole blood is requested any time a blood transfusion is ordered. However, whole blood has increasingly become unavailable, prompting us to prospectively study use of blood products in the Parkland Obstetrics Service. The objective of this study was to study the use of blood products, including whole blood, for the management of obstetric hemorrhage requiring transfusion.
MATERIALS AND METHODS
This was a population-based, observational study of transfusion practices in pregnant women delivered at Parkland Hospital. This study was approved by the Institutional Review Board of the University of Texas Southwestern Medical Center. Parkland Hospital is a tax-supported institution serving Dallas County that has a Level III neonatal intensive care unit adjacent to the labor and delivery unit. The obstetric service is staffed by house officers and faculty members of the Department of Obstetrics and Gynecology at the University of Texas Southwestern Medical School.
The transfusion medicine service at Parkland is staffed by house officers and faculty in the Department of Pathology from the University of Texas Southwestern Medical Center. Prestorage of leuko-reduced blood products were provided to Parkland Hospital by Carter Blood Care or the American Red Cross, both nonprofit distributors serving the Dallas-Fort Worth region.
Selected obstetric and neonatal outcomes for all women who give birth at Parkland Hospital are entered into a computerized database. Nurses attending each delivery complete an obstetric data sheet, and database nurse specialists assess the data for consistency and completeness before they are stored electronically. A subfile for women given blood products was added to the database and included prespecified information fields to include the reasons for transfusion, estimate of blood loss, criteria for hypovolemia met, evidence of organ dysfunction, transfusion reactions, and care levels required for each woman. This subfile was completed by one specific database nurse specialist who was notified by the hospital transfusion service whenever an obstetric patient received blood products.
All women presenting to the labor and delivery unit undergo ABO (Rh) blood typing and indirect Coombs antibody testing upon admission. If the antibody screen is positive, cross-matched packed red blood cells are prepared. Otherwise, if transfusion is to be performed, type-specific blood product is released from the blood bank, followed by confirmation of recipient and donor. If blood typing has not been done, O negative packed red blood cells are transfused on an emergency basis. A standardized blood order form for women with a negative antibody screen specifies whole blood is to be sent if available. However, this was typically limited to blood type O (Rh+) and A (Rh+) due to limited availability of other types. Transfusion for obstetric hemorrhage was limited to those women evidencing hypovolemia as described below. Lactated Ringers solution was always infused before transfusion of blood products.
Hypovolemia was diagnosed in women experiencing obstetric hemorrhage sufficient enough to provoke hemodynamic instability, as indicated by the presence of one or more of the following: 1) systolic blood pressure less then 100 mm Hg not due to regional analgesia or anesthesia; 2) pulse 100 beats per minute or more; 3) a positive “tilt” test (20 beats per minute increase in pulse or decrease in systolic blood pressure of 20 mm Hg) or of the static symptoms (to include dizziness, fainting, nausea, or vomiting upon sitting up); and 4) urine flow less than 30 mL/h. Anyone with a hematocrit less than 20% secondary to hemorrhage or who had a hematocrit between 20% and 30% in the face of ongoing hemorrhage and evidence of hemodynamic instability per the above criteria received blood. A blood use review committee systematically audited on an ongoing basis all use and wastage of blood products for Parkland hospital. This review includes compliance with prespecified transfusion criteria for hypovolemia.
The subfile of women transfused was electronically linked to the other perinatal elements in the database, and an anonymous data set was created. The outcomes of interest included consequences of severe hypovolemia or transfusion or both, to include acute tubular necrosis (diagnosed when the serum creatinine was 1.6 mg/dL or more in women with prior normal renal function), admission to intensive care, adult respiratory distress syndrome, pulmonary edema, and transfusion reactions. Statistical analysis was performed using χ2, Wilcoxon rank sum test, and Kruskal-Wallis test for analysis of variance. Pair-wise comparisons were performed after overall statistical significance for higher order χ2 tests and for the Kruskal-Wallis test. The significance of pair-wise tests were adjusted by the method of Bonferroni. Values of P<.05 were considered significant. We used SAS 9 (SAS Institute Inc., Cary, NC).
Between March 24, 2002, and June 12, 2006, 1,540 (2.3%) women received transfusions for obstetric hemorrhage among 66,369 women delivered at Parkland Hospital. As shown in Figure 1, 659 (43%) women were transfused with whole blood only, 593 (39%) with packed red blood cells only, and the remaining 288 (19%) received combinations of blood products. The mean units transfused were 2.2 and 2.3 units in the whole blood and packed red blood cell transfusion groups, respectively. Women given combinations of blood products received a significantly larger number of units (mean 5.5 units, P<.001 compared with the whole blood and packed red blood cell–only groups, Wilcoxon rank sum test). Approximately 72% of these women received combinations of whole blood and packed red blood cells, with the remaining 28% receiving other component therapy.
Maternal demographic characteristics for women transfused are summarized in Table 1. There were no significant differences except in the American Society of Anesthesiologists class. Specifically, the incidence of Class III was significantly increased in women given combinations of blood products compared with the other two transfusion groups. The criteria used to diagnose hypovolemia was similar in groups A and B; however, a systolic BP less than 100 mm Hg, a pulse rate of more than 100 beats per minute, and oliguria were all greater in the combination of products group, P<.01. The median (interquartile range) hematocrit at the time of transfusion in the packed red blood cell–only group was 24.2 (21.6–27.5), the whole blood group was 24.1 (21.3–27.2), and the combination group was 24.3 (20.9–27.1). The mean±standard deviation crystalloid use was greater in the combination groups, 13.1±11 L compared with 9.1±4.7 L in the packed red blood cell–only group and 8.9±4.3 in the whole blood group, P<.001.
Obstetric complications in women transfused are shown in Table 2. There were no significant differences except for placental previa or abruption, hysterectomy, and perineal trauma, which were all increased in the combination blood products group.
Morbidities, to include acute tubular necrosis, adult respiratory distress syndrome, pulmonary edema, hypofibrinogenemia, and intensive care unit admission, were all significantly related to the different blood transfusion groups. Acute tubular necrosis was significantly decreased in women given whole blood only compared with packed red blood cells only or combination therapy, 0.3% compared with 2% compared with 4%, respectively, P<.001 (Table 3).
The hospital courses of the 14 women developing acute tubular necrosis in the packed red blood cell and whole blood groups were reviewed. Two of the 14 were in the whole blood only group and were discharged from the hospital with an elevated serum creatinine that went on to normalize without the need for dialysis. The remaining 12 women were in the packed red blood cell only group. Five of the women had a return of the creatinine to normal before discharge. Four were discharged from the hospital with elevated serum creatinine values that returned to normal in the following weeks. One woman had chronic renal insufficiency, with a creatinine of 1.5 mg/dL. The remaining two women given packed red blood cell transfusions required dialysis; one died of multiorgan failure, and the other woman continued on long-term dialysis due to complete renal failure.
Pulmonary edema was significantly increased in women given whole blood only compared with packed red blood cells only (7% compared with 4%, respectively, P<.001). Women in the combination blood product group experienced significantly increased rates of adult respiratory distress syndrome, hypofibrinogenemia, and admission to intensive care. Careful chart review of the women with pulmonary insufficiency ruled out transfusion-related acute lung injury. All cases of pulmonary edema were attributed to transfusion-associated circulatory overload and were successfully managed with diuretics and oxygen supplementation. Adult respiratory distress syndrome received ventilation support in our intensive care unit.
There were three maternal deaths; two received a combination of blood products and one received only packed red blood cells. There were no deaths in the whole blood group. One maternal death occurred in a woman who underwent a primary cesarean for a failed induction of labor for severe preeclampsia. She decompensated 36 hours after surgery, experienced multiorgan failure, and died within 24 hours. Although the exact cause of death was unable to be determined, it was thought that pulmonary embolism triggered the initial event. She received a combination of blood products as part of the resuscitation effort. A second maternal death was in a woman with diabetes and chronic congestive heart failure. She underwent a cesarean delivery for prolonged labor and a nonreassuring fetal heart rate pattern. She experienced rapid deterioration of her cardiac status postoperatively and died of respiratory failure. She received 2 units of packed red blood cells for blood loss at cesarean delivery. The third maternal death was in a woman with severe preeclampsia who experienced placental abruption. Several hours after a cesarean delivery for fetal distress, she developed liver failure of unknown cause. She received a combination of blood products during her postoperative course, but ultimately developed multisystem organ failure and died of respiratory failure 2 weeks after delivery.
In this population-based study of 66,369 women giving birth during a 51-month period, the incidence of obstetric hemorrhage severe enough to cause hypovolemia was 2.3%. Based on the blood products transfused, women with obstetric hemorrhage at our hospital were separated into three groups with approximately equal proportions receiving only packed red blood cells or whole blood (39% compared with 43%, respectively). A third group received combination therapy to include whole blood and packed red blood cells, fresh frozen plasma, platelets, or cryoprecipitate. The latter group significantly more often experienced complications, such as acute tubular necrosis, adult respiratory distress syndrome, pulmonary edema, hypofibrinogenemia, and intensive care unit admission. This group of women received an average of 5.5 units of blood products, compared with an average of approximately 2 units per woman in the packed red blood cell and whole blood only groups, suggesting that the combination transfusion group suffered more serious obstetric hemorrhages. It is important to emphasize that women given whole blood transfusions on our obstetric service are those who have sustained hemorrhages sufficient enough to provoke hemodynamic instability, often with ongoing bleeding. Women with asymptomatic anemias due to lesser hemorrhage are not routinely transfused unless their hematocrits are less than 20%.
The great majority (81%) of women with obstetric hemorrhage received either packed red blood cells or whole blood exclusively and were very similar with regard to maternal demographics and obstetric complications associated with hemorrhage. These two groups, however, experienced significantly different rates of organ system morbidities. Specifically, women transfused with packed red blood cells more often developed acute tubular necrosis compared with women given whole blood, whereas the latter group more often developed pulmonary edema. Six percent of these latter cases were associated with acute respiratory distress syndrome, similar to the packed red blood cell group. These differences could be attributed to insufficient replacement of circulatory blood volume in women developing acute tubular necrosis when only packed red blood cells were used and overreplacement of volume in women given whole blood.
Our use of whole blood, when available, is undoubtedly seemingly an anachronism when placed into the contemporary context of almost universal recommendations that packed red blood cells be used for treatment of hypovolemia due to hemorrhage. The continued use of banked whole blood in the Parkland Obstetrics Service has its origins in the study of placental abruption. This obstetric complication produces rapid, large blood loss that averages by direct measurement 2,000 mL per woman at the outset, and which is associated with hypofibrinogenemia due to consumptive coagulopathy.5 The goals of transfusion in this setting are restoration of circulatory blood volume as well as replacement of fibrinogen. Whole blood accomplishes such restoration, whereas packed red blood cells plus thawed fresh frozen plasma are required when component therapy is used. This strategy also reduces the number of donor exposures for a given woman. Urine flow of 30 mL/h and hematocrit values of 25–30% also evolved as useful guidelines for managing ongoing hemorrhage due to abruption. These principles of transfusion for women with placental abruption ultimately evolved into the preferred approach for treatment of hypovolemia because of any obstetric cause on our service. However, whole blood has become increasingly difficult for us to obtain, and this prompted us to perform this observational study.
Is whole blood for transfusion of serious hemorrhage to be relegated to the archives of transfusion medicine? Just a few years ago, the answer to this question would have been a quick and unequivocal yes, given that component transfusion medicine maximizes use of each donated unit of whole blood. However, recent reports from the war in Iraq have raised the possibility that whole blood transfusion has advantages over component therapy in the management of acquired coagulopathy of trauma, which is a major cause of battlefield mortality.3,9,10 The transfusion of large amounts of preserved packed red blood cells contributes to this coagulopathy, which is primarily the result of thrombocytopenia and poor platelet function.10–13 Fresh whole blood, however, has been shown to reverse this coagulopathy in a retrospective study of 87 patients managed at a combat hospital in Iraq. There, military physicians have observed distinct advantages of fresh warm whole blood over component therapy and consequently developed a novel whole blood–based massive transfusion protocol. Military physicians have also noticed advantages of replacing blood loss with a 1:1 ratio of packed red blood cells to thawed plasma that include decreased coagulopathy in those cases experiencing significant blood loss.14 This has improved survival and is an area of ongoing research. Laine and colleagues,15 in a controlled randomized study of 203 patients, compared packed red blood cells plus fresh frozen plasma with whole blood during liver transplant surgery and concluded that although whole blood was equally effective to component therapy, whole blood, compared with multiple components transfusion, greatly simplified the mechanics of handling blood products in the laboratory as well as the operating room and also significantly reduced the number of donor exposures. These reports and our experiences prompt us to believe that there is emerging evidence that whole blood transfusion may have advantages over component therapy when dealing with serious hemorrhage. Given the choice and our experiences now reported, we are of the view that transfusion of whole blood should be reconsidered in the management of serious obstetric hemorrhage.
1. Kung HC, Hoy DL, Xu JQ, Murphy SL. Deaths: final data for 2005. Vol. 56, No. 10. Hyattsville (MD): National Center for Health Statistics; 2008.
2. Chichakli LO, Atrash HK, Mackay AP, Musani AS, Berg CJ. Pregnancy-related mortality in the United States due to hemorrhage: 1979-1992. Obstet Gynecol 1999;94:721–5.
3. Repine TB, Perkins JG, Kauvar DS, Blackborne L. The use of fresh whole blood in massive transfusion. J Trauma 2006;60:S59–69.
4. American College of Obstetricians and Gynecologists. ACOG Practice Bulletin: Clinical Management Guidelines for Obstetrician–Gynecologists Number 76, October 2006: postpartum hemorrhage. Obstet Gynecol 2006;108:1039–47.
5. Pritchard JA, Brekken AL. Clinical and laboratory studies on severe abruptio placentae. Am J Obstet Gynecol 1967;97:681–700.
6. Pritchard JA, Wiggins KM, Dickey JC. Blood volume changes in pregnancy and the puerperium. I. Does sequestration of red blood cells accompany parturition? Am J Obstet Gynecol 1960;80:956–64.
7. Pritchard JA, Baldwin RM, Dickey JC, Wiggins KM. Blood volume changes in pregnancy and the puerperium. II. Red blood cell loss and changes in apparent blood volume during and following vaginal delivery, cesarean section, and cesarean section and total hysterectomy. Am J Obstet Gynecol 1962;84:1271–82.
8. Pritchard JA. Changes in the blood volume during pregnancy and delivery. Anesthesiology 1965;26:393–9.
9. Kauvar DS, Holcomb JB, Norris GC, Hess JR. Fresh whole blood transfusion: a controversial military practice. J Trauma 2006;61:181–4.
10. Spinella PC, Perkins JG, Grathwohl KW, Repine T, Beekley AC, Sebesta J, et al. Fresh whole blood transfusions in coalition military, foreign national, and enemy combatant patients during Operation Iraqi Freedom at a U.S. combat support hospital. World J Surg 2008;32:2–6.
11. Counts RB, Haisch C, Simon TL, Maxwell NG, HeimbachDM, Carrico CJ. Hemostasis in massively transfused trauma patients. Ann Surg 1979;190:91–9.
12. Lim RC Jr, Olcott C 4th, Robinson AJ, Blaisdell FW. Platelet response and coagulation changes following massive blood replacement. J Trauma 1973;13:577–82.
13. Miller RD, Robbins TO, Tong MJ, Barton SL. Coagulation defects associated with massive blood transfusions. Ann Surg 1971;174:794–801.
14. Holcomb JB, Jenkins D, Rhee P, Johannigman J, Mahoney P, Mehta S, et al. Damage control resuscitation: directly addressing the early coagulopathy of trauma. J Trauma 2007;62:307–10.
15. Laine E, Steadman R, Calhoun L, Blackall D, Levin P, Braunfeld M, et al. Comparison of RBCs and FFP with whole blood during liver transplant surgery. Transfusion 2003; 43:322–7.