Placenta accreta occurs when the placenta is abnormally adherent to the uterine lining.1 Consequently, the placenta does not properly separate from the uterus after delivery, leading to maternal hemorrhage. The morbidity from placenta accreta can be dramatic and includes disorders associated with massive bleeding such as disseminated intravascular coagulation, multi-organ failure, and even death.2–4 In most cases, hysterectomy is required to stop the bleeding.
The rate of placenta accreta is rising, likely as a result of the increasing cesarean delivery rate. The incidence is now reported to be 1 in 533, considerably more than the 1 in 2,510 noted in a large center from 1985 to 1994.5,6 Previous studies have shown a strong correlation between the number of previous cesarean deliveries and an increased risk of placenta accreta, particularly in cases of placenta previa.7 Placenta accreta has now become the most common reason for cesarean hysterectomy in developed countries.8,9
Peripartum hysterectomy, particularly in cases of placenta accreta, is often technically challenging and is associated with a higher risk for complications than abdominal hysterectomies performed for benign indications.10 These cesarean hysterectomies often require difficult dissection of poorly defined tissue planes, particularly of the bladder interface, and partial bladder resection is often required. In addition, large-volume transfusion of blood products is frequently necessary. Finally, specialized interventional radiology procedures may be useful in some cases. We hypothesized that maternal outcomes are improved in cases of placenta accreta managed by a multidisciplinary team with expertise and experience in the care of women with this condition. Thus, our objective was to compare maternal morbidity in cases of placenta accreta managed by a multidisciplinary care team including maternal fetal medicine physicians, gynecologic oncologists, interventional radiologists, and a tertiary care blood bank with those managed by standard obstetric care.
MATERIALS AND METHODS
This was a retrospective cohort study of pregnancies complicated by placenta accreta in the state of Utah from October 1996 until December 2008. Approval was obtained from the institutional review boards at the University of Utah Health Sciences Center and Intermountain Healthcare, for cases occurring at Intermountain Medical Center and Latter Day Saints Hospital. We have previously reported on the maternal outcomes in cases of placenta accreta managed at the University of Utah Health Sciences Center and Latter Day Saints Hospital in Salt Lake City, Utah, in a descriptive case series.3 In this study, we included cases occurring in the interim since that publication and further broadened our search to include all identifiable cases of accreta occurring within the state of Utah during the aforementioned time period. In addition, permission was obtained from the Utah Department of Health to further review records at all hospitals within the state. The health department provided a database of possible placenta accreta cases encompassing all 28 hospitals within the state based on ICD-9 codes, which included placenta accreta, postpartum hysterectomy, postpartum hemorrhage (immediate and delayed), postpartum coagulation defects, and retained placenta or membranes (with or without complications or hemorrhage or both).
Once the records had been obtained and de-identified, a composite database was compiled. Placenta accreta was defined as the placenta being adherent to the uterine wall without easy separation and included the spectrum of placenta accreta, increta, and percreta.1 The diagnoses were confirmed by histopathologic evidence of placental invasion into the myometrium, by clinical assessment of abnormal adherence of the placenta, or by evidence of gross placental invasion at the time of surgery. After review of available data, several cases were excluded because they clearly did not involve placenta accreta. The remaining cases were subject to chart abstraction for maternal medical, obstetric, and gynecologic history; timing of diagnosis; antepartum and intrapartum management; maternal postpartum course; and complications occurring within 6 months of delivery.
A delivery was considered “scheduled” if planned at least 1 day in advance and performed nonurgently because of either documented fetal lung maturity or clinical concerns for risks associated with expectant management such as eventual hemorrhage or labor. Early and delayed reoperations were defined as surgical procedures occurring less than or greater than 7 days after delivery, respectively. Delayed ureteral stent removal procedures were not considered delayed reoperations as they would be anticipated in cases of ureteral stenting and would therefore not represent an additional complication. Cystotomy included both intentional and unintentional intraoperative surgical entry into the bladder.
Composite early morbidity was defined as the occurrence of one or more of the following: maternal admission into the intensive care unit (ICU) for more than 24 hours, transfusion of 4 or more units of packed red blood cells, coagulopathy (platelets 100,000 or less, international normalized ratio 1.2 and higher, or fibrinogen 200 or less), ureteral injury, early reoperation, or death. Composite late morbidity was defined as the occurrence of one or more of the following: intraabdominal infection (defined as persistent fever 38.3°C or higher, leukocytosis, and abdominal pain), hospital readmission within 6 weeks, or delayed reoperation. Cystotomy was not included as a major morbidity because it is performed intentionally in some cases of placenta accreta to facilitate dissection of the posterior bladder in hopes of avoiding bladder resection or to confirm ureteral patency after an urgent cesarean hysterectomy.
The hospitals themselves were divided into multidisciplinary care centers and standard care centers. Multidisciplinary care centers were defined as institutions with 24-hour in-house obstetrician gynecologists, anesthesiologists, fully stocked blood banks, immediate availability of a gynecologic oncologist, and interventional radiology. There were two hospitals that met criteria for multidisciplinary care. The standard care centers included a total of 26 other hospitals: three tertiary care centers (not affiliated with obstetrics and gynecology residencies) with some, but not all components of the multidisciplinary care center requirements as well as 23 smaller, community hospitals. A patient's transport status indicated that the patient was transported to a higher level of care, usually from a standard care center to a multidisciplinary care center.
Descriptive statistics were used to describe the cohort. Outcomes among women delivered at multidisciplinary care centers were compared with those delivered at standard care centers using Student t test, χ2 analysis, and Fisher exact test as appropriate. The Mann-Whitney test was used to compare medians between groups for nonparametric data. The Fisher-Pitman permutation test for independent samples was used to compare estimated blood loss between groups.11 A multivariable regression model was used to compare early composite morbidity among groups while controlling for placenta previa, number of cesarean deliveries, percreta, and antepartum hemorrhage.
Using the database compiled by the Utah Department of Health, 362 potential cases of placenta accreta were identified. Once the charts had been reviewed, 141 cases met the criteria of clinical documentation or histopathologic diagnosis of placenta accreta or both (Fig. 1). Seventy-nine cases were identified at two multidisciplinary care centers and 62 cases at 26 standard care centers.
Demographics and clinical characteristics of all cases are shown in Table 1. Overall, the two patient groups were similar with respect to maternal age, cesarean and other uterine surgical history, and the presence of placenta previa. Women delivering at multidisciplinary care centers had higher gravidity and parity compared with those delivering at standard care centers. The median gestational age at delivery was lower at multidisciplinary care centers compared with standard care centers for all cases of accreta and also for the subgroup of antenatally suspected cases (33.9 weeks compared with 36.0 weeks, P=.003).
Within the entire cohort, there were four stillbirths that occurred before labor and in the absence of vaginal bleeding at 23, 31, 36, and 37 weeks of gestation. Two women delivered previable fetuses before 21 weeks of gestation secondary to ruptured membranes complicated by intraamniotic infection, one woman delivered a previable fetus at 22 weeks due to excessive maternal hemorrhage with preterm labor, and one woman was found to have abnormal placentation after severe hemorrhage with dilation and evacuation of an intrauterine fetal demise at 15 weeks of gestation.
The conditions prompting delivery are depicted in Table 2. Indications for delivery included scheduled elective delivery, vaginal bleeding, spontaneous preterm or term labor, preterm premature rupture of membranes, suspected chorioamnionitis, preeclampsia, and intrauterine fetal demise. Overall, reasons for delivery differed between multidisciplinary and standard care centers (P=.021). Women cared for in a multidisciplinary care center were more likely to be delivered for vaginal bleeding compared with those delivering in a standard care center. Regional anesthetic was used in a similar proportion of cases at multidisciplinary and standard care centers (26% compared with 23%, P=.637); however, a higher proportion of cases initially managed with regional anesthesia were converted to general anesthesia at standard compared with multidisciplinary care centers (36% compared with 8%, P<.001). Placental removal was less likely to be attempted in cases managed at a multidisciplinary compared with a standard care center (42% compared with 81%, P<.001). Preoperative internal iliac balloon catheters were used in 9 cases (15%) managed at standard care centers and no interventional radiology procedures were used in any of the cases managed at multidisciplinary care centers. However, hypogastric artery ligation was performed intraoperatively in 23 cases (29%) managed at multidisciplinary care centers and the procedure was not performed in any cases managed at standard care centers. Gynecologic oncologists were involved in 52 cases (66%) managed at multidisciplinary care centers and 3 cases (5%) managed at standard care centers. The proportion of cases with histopathologic confirmation of accreta was similar between multidisciplinary and standard care centers (71% compared with 82%, P=.132). Among those cases in which no placental removal was attempted before hysterectomy, histopathologic examination did not confirm accreta in 12 cases: 11 (24%) at multidisciplinary care centers and 1 (8%) at a standard care center (P=.223).
Table 3 summarizes the maternal morbidity for the entire cohort, regardless of antenatal suspicion for placenta accreta. Estimated blood loss values were wide-ranging with no difference in median estimated loss. The most commonly documented estimated loss was 2.0 liters at multidisciplinary care centers (24%) and 2.5 liters at standard care centers (15% of cases). There was one maternal death at a standard care center. Among the women delivering at a multidisciplinary care center, two women (3%) initially delivered by cesarean required early reoperation for bleeding complications compared with 16 women (36%) initially delivered by cesarean at a standard care center (P<.001). Both women at the multidisciplinary care centers had originally undergone cesarean hysterectomy and required reoperation for hemoperitoneum. At standard care centers, 13 of the 16 women underwent early reoperation for hysterectomy to control bleeding following an attempt at uterine preservation at the time of delivery. Other indications for reoperation at standard care centers included ureteral obstruction (1), hemoperitoneum (1), and retroperitoneal hematoma (1), Fewer women delivered at multidisciplinary care centers required large volume blood transfusion (4 or more units of packed red blood cells) compared with women delivered at standard care centers (43% compared with 61%, P=.031). Women at multidisciplinary care centers received 0–46 units of packed red blood cells with a median value of 2 units (mode of 2 units transfused in 29.1%), whereas women at standard care centers received 0–51 units of packed red blood cells with a median value of 5 units (mode of 2 units transfused in 19.7%). Infectious complications were similar among groups, as was the median postoperative length of stay. There was no significant difference in composite early or late morbidity between groups. A multivariable logistic regression model was used to control for the presence of placenta previa, percreta, number of previous cesarean deliveries, and antenatal vaginal bleeding. In this model, delivery at a multidisciplinary care center was protective for composite early morbidity with an odds ratio (OR) of 0.46 (95% confidence interval [CI] 0.22–0.95, P=.04). There was no difference in the incidence of cystotomy (intentional or unintentional) between women managed at multidisciplinary compared with standard care centers (32% compared with 23%, P=.232).
Table 4 summarizes maternal morbidity in the subgroup of women with antenatally suspected placenta accreta. Early reoperation occurred less frequently among patients delivered at multidisciplinary compared with standard care centers (3% compared with 41%, P<.001) and the incidence of large-volume blood transfusion was also lower (42% compared with 70%, P=.023). Fewer women delivered at a multidisciplinary care center experienced early morbidity compared with those delivered at a standard care center (47% compared with 74%, P=.026). In univariable analysis, delivery at a multidisciplinary care center was protective for early morbidity relative to delivery at a standard care center (OR 0.31, 95% CI 0.11–0.89, P=.030). Multivariable logistic regression analysis showed that this protective effect remained after controlling for the presence of placenta previa, percreta, the number of previous cesarean deliveries, and the occurrence of antenatal vaginal bleeding (OR 0.22, 95% CI 0.07–0.70, P=.010).
Placenta accreta is associated with considerable morbidity including coagulopathy, ureteral injury, infection, and the need for reoperation, and these cases present a challenge to obstetrician-gynecologists. The rate of placenta accreta appears to be rising, almost assuredly as a consequence of the rising rate of cesarean delivery. Accordingly, all obstetric providers need to have a high index of suspicion and be aware of risk factors for placenta accreta. Given the substantial morbidity and increasing incidence of placenta accreta, it is paramount that we evaluate and implement strategies to improve outcomes. One such strategy is to use multidisciplinary teams with experience and expertise in the management of placenta accreta. In this cohort, delivery at a medical center with a multidisciplinary care team resulted in a more than 50% risk reduction for composite early morbidity among all cases of accreta and a nearly 80% risk reduction among those cases wherein accreta was suspected before delivery.
Potential advantages of a multidisciplinary team include surgical expertise, access to blood banks capable of managing massive transfusion, and readily available intensive care units. Perioperative morbidity and maternal mortality in cases of major obstetric hemorrhage for any indication have been shown to be lower in women delivering in high-volume hospitals compared with those delivering in low-volume hospitals.12 It makes intuitive sense that expertise and experience are useful when managing difficult and somewhat uncommon problems. Although our cohort is too small to evaluate the specific role of gynecologic oncologists in improving outcomes, we suspect that experienced surgeons who frequently perform complex pelvic surgery likely do contribute to reduced maternal morbidity, especially for cases requiring extensive bladder resection or in cases where the placenta invades the broad ligament and retroperitoneal space. In our centers, we attempt to have gynecologic oncologists perform hysterectomies for placenta accreta because of their extensive surgical training and experience.
Our study had several limitations. There was insufficient sample size to exclude small differences between groups for individual rare complications such as ureteral injury. This prompted the use of composite morbidities, which bias toward a positive result. However, there were also significant differences in individual morbidities such as large-volume transfusion and the need for early reoperation. A greater proportion of women delivering in a multidisciplinary care center underwent hysterectomy without an attempt at placenta removal and this difference in operative technique by experienced providers may also have contributed to improved outcomes. We found that among 11 cases managed at a multidisciplinary care center wherein the placenta was left in situ before hysterectomy, histopathologic examination of the uterine specimen did not confirm placenta accreta compared with one such case managed at a standard care center. In many of these cases, the pathology report indicated less than 3 mm of intervening myometrium between the placenta and the overlying serosa and there was no comment made regarding the presence of Nitabuck's layer; therefore, we chose to include these cases because we believe they represent accreta as it is encountered in clinical practice. Although the difference in the proportion of histopathologically unconfirmed cases was not significant between groups (P=.223), we must acknowledge that this may bias our results toward more favorable outcomes at multidisciplinary care centers. Some records could not be located for review, introducing another potential source of bias, the direction of which is unknown. Cases were not managed in a uniform fashion at either the multidisciplinary care centers or the standard care centers over the entire time period of the study. Thus, we cannot define which component of the multidisciplinary team may have contributed to the relatively decreased morbidity. Finally, some standard care centers had components of a multidisciplinary care center, possibly minimizing the differences noted between groups. Unfortunately, the number of patients cared for in each standard care center was too small to allow for meaningful subgroup analysis.
There were also numerous strengths of the study. Utilization of public health records allowed identification of all potential cases of placenta accreta across the state of Utah rather than only at select institutions. Cases were identified at a mix of urban, rural, tertiary care, and community hospitals. In addition, there was a relatively large number of well-characterized cases of placenta accreta. Finally, all cases had extensive medical record abstraction with uniform data collection by two M.D. investigators.
In summary, maternal morbidity was decreased in women with placenta accreta delivering at tertiary care hospitals with multidisciplinary teams experienced in the care of patients with the condition. The benefit was more profound in cases of suspected placenta accreta. When possible, patients with suspected placenta accreta should be cared for by a multidisciplinary team in a tertiary care hospital.