Objective
We examined predictors of massive blood loss for women with placenta accreta who had undergone hysterectomy.
Study Design
A retrospective review of women who underwent peripartum hysterectomy for pathologically confirmed placenta accreta was performed. Characteristics that are associated with massive blood loss (≥5000 mL) and large-volume transfusion (≥10 units packed red cells) were examined.
Results
A total of 77 patients were identified. The median blood loss was 3000 mL, with a median of 5 units of red cells transfused. There was no association among maternal age, gravidity, number of previous deliveries, number of previous cesarean deliveries, degree of placental invasion, or antenatal bleeding and massive blood loss or large-volume transfusion ( P > .05). Among women with a known diagnosis of placenta accreta, 41.7% had an estimated blood loss of ≥5000 mL, compared with 12.0% of those who did not receive the diagnosis antenatally with ultrasound scanning ( P = .01).
Conclusion
There are few reliable predictors of massive blood loss in women with placenta accreta.
Obstetric hemorrhage is a major cause of morbidity and death in young women. Annually, it is estimated that >600,000 women die during childbirth and the puerperium, with up to one-third of these deaths attributed to obstetric hemorrhage in some regions of the world. In the United States, 10% of maternal deaths are attributable directly to obstetric hemorrhage.
See Journal Club, page 85
Abnormal placentation, which includes placenta accreta and its variants, is one of the most common causes of major obstetric hemorrhage. Although there are a number of risk factors for placenta accreta, previous cesarean delivery is the most important. The risk of placenta accreta increases in direct proportion to the number of previous cesarean deliveries. Although conservative management techniques have been described, the treatment of placenta accreta often requires peripartum hysterectomy.
Peripartum hysterectomy for placenta accreta is associated with substantial morbidity. The procedure often is performed emergently in the setting of heavy bleeding and without adequate personnel. In addition to operative complications, women who undergo peripartum hysterectomy often require large quantities of blood products in a short period of time. An analysis from the United Kingdom reported a median transfusion requirement of 10 units of packed red cells and 4 units of fresh frozen plasma among women who underwent peripartum hysterectomy. The large volume transfusion requirements of these women necessitate the availability of adequate blood bank support services.
Many hospitals lack the intensive support services that are needed to manage the substantial blood component requirements of women who undergo peripartum hysterectomy for placenta accreta. The goal of our study was to determine predictors of massive blood loss in patients with placenta accreta who undergo hysterectomy. Specifically, we sought to identify factors that are associated with large-volume blood component requirements to develop referral recommendations to tertiary facilities for women with abnormalities of placentation.
Materials and Methods
After institutional review board approval was obtained, a retrospective study of women with placenta accreta was performed. We identified women with a pathologically confirmed placenta accreta (including placenta increta and percreta) who delivered between 2000 and 2010. Patients who had clinical characteristics of placenta accreta, but without pathologic documentation of placental invasion into the uterus, were excluded from the analysis. Institutional databases were queried to determine the demographic and clinical characteristics of each patient. For each subject, the details of the preoperative, intraoperative, and postoperative treatment were reviewed. The imaging results (ultrasound scans and magnetic resonance imaging) were collected. Loss of the endometrial-myometrial interface, invasion into the myometrium or surrounding tissues, or the presence of lacunar spaces within the placenta were considered evidence of abnormal placentation.
The estimated blood loss (EBL) of each patient was identified from operative reports. The EBL was estimated and documented at the time of the surgical procedure. We determined the number of units of packed red blood cells (PRBCs) that each patient received during the surgical procedure. During the study period, cell salvage was commonly used in women who underwent peripartum hysterectomy. The number of units of red cells that were collected and reinfused from cell salvage was included in the quantitation of PRBCs that were transfused.
The primary endpoints of the study were massive blood loss and large-volume PRBC transfusion. A priori, we selected cutpoints of ≥5000 mL to define massive blood loss and transfusion of ≥10 units of red cells to define large volume transfusion.
Statistical analysis was performed to determine clinical characteristics that are associated with massive blood loss and large volume transfusion. Frequency distributions between categoric variables were compared with the use of the χ 2 test. Multivariable logistic regression models were developed to examine predictors of massive EBL and large volume transfusion while controlling for other clinical variables. A probability value of < .05 was considered statistically significant. All analyses were performed with SAS software (version 9.2; SAS Institute Inc, Cary, NC).
Results
A total of 77 patients who met the inclusion criteria were identified. Hysterectomy was required at the time of delivery in 68 women (88.3%). At the time of delivery, 44 women (57.1%) were ≥35 years old, and 19 women (24.7%) were at a gestational age of ≥37 weeks. The delivery was elective in 39 women (50.6%). Most of the patients (72.7%) had a placenta previa. Pathologic examination of the placenta and uterus revealed placenta accreta in 52 women (67.5%), and placenta increta or percreta in 25 women (32.5%).
The median EBL for the cohort was 3000 mL. Figure 1 graphically displays the blood loss of the cohort. The EBL was ≤2500 mL in 31 patients (40.3%); an EBL of ≤5000 mL was noted in 60 of the women (77.9%). Ten patients (13.0%) had a blood loss of ≥10,000 mL.
The mean transfusion requirement was 5 units of packed red cells ( Figure 2 ). Overall, 34 patients (44.2%) required <5 units of blood; 52 patients (67.5%) required <10 units of blood, and 60 women (77.9%) required <15 units. A total of 12 subjects (15.6%) received ≥20 units of blood, and 6 patients (7.8%) required ≥30 units of PRBCs at the time of surgery.
An analysis of factors that were associated with an EBL of ≥5000 mL is given in the Table . There was no association among maternal age, gravidity, number of previous cesarean deliveries, or antenatal bleeding and blood loss of ≥5000 mL ( P > .05 for all). Women with a diagnosis of placenta accreta before delivery were more likely to experience massive blood loss. Among women with a diagnosis of abnormal placentation on ultrasound scanning, 41.7% had an EBL of ≥5000 mL, compared with 12.0% of those women who were not diagnosed antenatally on ultrasound scanning (p = .01). Women who were treated more recently were more likely to experience heavy bleeding; 41.9% of patients who delivered between 2006 and 2010 had an EBL of ≥5000 mL, compared with 17.7% of those women who were treated between 2000 and 2005 ( P = .02). Massive blood loss was seen in 26.9% of patients with placenta accreta and in 40.0% of those patients with either a placenta increta or placenta percreta ( P = .25). There was no association between the presence or absence of placenta previa or the placental location and heavy blood loss ( P > .05 for each). Among women who delivered electively, 28.2% had an EBL of ≥5000 mL, compared with 34.2% of those women who had an unscheduled delivery. Likewise, an EBL of ≥5000 mL was noted in 44.8% of those women with a gestational age of <34 weeks at delivery, compared with 5.3% who delivered at a gestational age of >37 weeks ( P = .01).
