Objective
The objective of the study was to compare outcomes between patients who did and did not receive preoperative uterine artery balloon catheters in the setting placenta accreta.
Study Design
This was a retrospective case-control study of patients with placenta accreta from 1990 to 2011.
Results
Records from 117 patients with pathology-proven accreta were reviewed. Fifty-nine patients (50.4%) had uterine artery balloons (UABs) placed preoperatively. The mean estimated blood loss (EBL) was lower (2165 mL vs 2837 mL; P = .02) for the group that had UABs compared with the group that did not. There were more cases with an EBL greater than 2500 mL and massive transfusions of packed red blood cells (>6 units) in the group that did not have UABs. Percreta was diagnosed more often on final pathology in the group with UABs. Surgical times did not differ between the 2 groups. Two patients (3.3%) had complications related to the UABs.
Conclusion
Preoperative placement of UABs is relatively safe and is associated with a reduced EBL and fewer massive transfusions compared with a group without UABs.
The incidence of placenta accreta is rising rapidly in the United States, affecting an estimated 1 in 533 pregnancies. The strongest factor contributing to this rapid rise is the increased incidence of cesarean deliveries over the past 20 years. In addition to cesarean deliveries, a history of prior dilatation and curettage, surgical termination, resection of uterine septum, and myomectomies violating the endometrial cavity also increase the risk of developing placenta accreta. Nonsurgical factors include the presence of a placenta previa, advanced maternal age, grand multiparity, and smoking.
Placenta accreta increases the likelihood of significant morbidity and mortality for both mother and newborn, especially if diagnosed at the time of delivery. Antenatal diagnosis of placenta accreta with ultrasound or magnetic resonance imaging in high-risk populations has been shown to significantly improve both maternal and neonatal outcomes through such interventions as early delivery prior to labor, planned cesarean hysterectomy without disruption of the placenta, and multidisciplinary presurgical planning. A recent decision analysis suggests that the ideal timing for delivery in cases in which placenta accreta is highly suspected is 34 weeks; however this remains controversial. In addition to the timing of delivery, a lack of consensus remains for the preoperative and intraoperative management of these suspected cases of placenta accreta.
Prophylactic placement of occlusive uterine artery balloons (UABs) for the purpose of managing intraoperative hemorrhage in cases of placenta accreta has been a topic of debate for nearly 2 decades. Case reports and small series have offered mixed results with regard to surgical outcomes and tend to be observational or descriptive, thus not allowing for formal statistical analysis of their findings. Shrivastava et al published the largest series to date on this topic. In their case-control study based on 69 cases of placenta accreta, 19 of which received intravascular balloon catheters, the investigators found no benefit in terms of surgical outcomes. Additionally, they reported a 15.8% complication rate related to balloon placement.
The purpose of this study was to compare surgical outcomes and complication rates at our institution for patients that underwent cesarean hysterectomy with or without placement of intravascular balloon catheters.
Materials and Methods
This was a retrospective case-control study performed from our ongoing placenta accreta database at the University of California, San Diego, Health System. The study period was from 1990 to 2011 and has been approved by our institutional review board and Human Research Protection Program. All subjects were identified by their pathologic diagnosis obtained from hysterectomy specimens.
Data collected on each patient included standard demographic data, obstetrical and surgical history, antenatal radiologic workup and diagnosis, preoperative balloon placement and intraoperative use, anesthesiologist and surgeon operative reports, postoperative hospital course and disposition, neonatal outcome and disposition, and final pathologic diagnosis.
In all but 3 cases, balloon catheters were placed via the common femoral arteries. In the first 3 cases, vascular access was achieved via axillary/high brachial arteries. Once vascular access was obtained, either a 6 or 7 French vascular sheath was then placed. The internal iliac arteries were selected and either 5.3 mm (flow-directed balloon catheter, FDB 5.3-35-80; Cook Inc, Bloomington, IN) or 8.5 mm (occlusion balloon catheter OB-8.5-5-65; Boston Scientific, Boston, MA) occlusion balloon catheters were placed in the proximal internal iliac arteries. The balloons were inflated briefly and contrast was injected to verify stasis of flow. The balloons were then deflated. The 3 mL syringes, used to inflate the balloons, are marked to indicate how much fluid should be injected to create stasis in the arteries. These syringes were then attached to the catheters with a stopcock. Steristrips were placed on the catheters to mark the appropriate position at the entrance to the sheath. The catheters were taped to the skin. The patient was then taken to the operating room. Intraoperatively, the balloons were inflated at the surgeon’s discretion and the timing of inflation was recorded.
Primary outcomes included estimated blood loss (in milliliters), the need for transfusion of blood products, the number of units of blood products transfused, and operative time. Estimated blood loss (EBL) and amount of packed red blood cells (PRBCs) transfused were further dichotomized into cases with estimated blood loss (EBL) exceeding 2500 mL and cases of massive transfusion requiring greater than 6 units of PRBCs. Complications related to the balloon catheters were also reported.
Data that were not normally distributed were log transformed for analysis using parametric statistical tests and retransformed for presentation. Continuous variables were analyzed using Student t test, whereas categorical outcomes were analyzed using χ 2 or Fisher exact tests. All analyses were performed using SPSS Statistical Software (version 16.0; SPSS Inc, Chicago, IL) and significance was considered at P < .05.
Results
Records from 117 patients with pathology-proven accreta were reviewed. Fifty-nine patients (50.4%) had UABs placed. In 30 of these cases (50.8%), the balloons were inflated intraoperatively. In 29 of these 30 cases, the balloons were inflated once excessive bleeding was encountered.
There were no differences found in maternal characteristics, including age, gravidity, and parity between the 2 groups ( Table 1 ). Significant differences were found in the number of prior cesarean deliveries (2-3) patients had undergone in the group that had UAB placed compared with the group that did not. Also, a greater percentage of those with UABs had a predelivery diagnosis of invasive placentation (98% vs 29%; P < .01). Specifically, placenta percretas were more likely to be diagnosed prenatally than placenta accreta (95.3% vs 45.9%; P < .01), and there were significantly more cases of placenta percreta, as opposed to accreta, diagnosed pathologically in the group that had UABs placed preoperatively (59.3% vs 13.8%; P < .01).
| Demographic | UAB catheters placed (n = 59) | No UAB catheters (n = 58) | P value |
|---|---|---|---|
| Age, y | 32.6 ± 5.8 | 33.2 ± 6.3 | .75 |
| Gravidity | |||
| 4 or fewer | 32 (54.2%) | 34 (58.6%) | .39 |
| 5 or more | 27 (45.8%) | 24 (41.4%) | |
| Parity | |||
| 4 or fewer | 51 (86.4%) | 49 (84.5% | .49 |
| 5 or more | 8 (13.6%) | 9 (15.5%) | |
| Prior cesarean sections, n | |||
| 0 | 1 (1.7%) | 15 (25.9%) | < .01 |
| 1 | 19 (32.2%) | 19 (32.8%) | .55 |
| 2 or 3 | 32 (54.2%) | 19 (32.8%) | .02 |
| 4 | 7 (11.9%) | 5 (8.6%) | .39 |
| Diagnosed prenatally | 58 (98%) | 17 (29%) | < .01 |
| Final pathology | |||
| Accreta | 24 (40.7%) | 50 (86.2%) | < .01 |
| Percreta | 35 (59.3%) | 8 (13.8%) |
Estimated blood loss in the group with balloon placement was 2165 mL compared to 2837 mL in the group without balloons ( P = .02). The number of cases with EBL greater than 2500 mL was significantly higher in the group without UABs placed. Although there was no significant difference in the need for blood transfusion or the mean amount of blood products transfused ( Table 2 ), there were significantly more cases of massive transfusion requiring more than 6 units of PRBCs in the group without UABs.
| Variable | UAB catheters placed (n = 59) | No UAB catheters (n = 58) | P value |
|---|---|---|---|
| GA at delivery, wks | 33.9 ± 1.1 | 34.1 ± 1.2 | .85 |
| Mean EBL, mL | 2165 ± 245 | 2837 ± 288 | .02 |
| EBL >2500 mL | 16 (27%) | 32 (55.2%) | < .01 |
| Transfused | 46 (78%) | 46 (79%) | .37 |
| PRBC, units | 4.7 ± 2.1 | 5.9 ± 1.7 | .14 |
| FFP, units | 3.9 ± 2.1 | 5.2 ± 2.3 | .17 |
| Platelets, units | 2.1 ± 2.1 | 2.1 ± 1.9 | .89 |
| Massive transfusion (≥6 u PRBCs) | 18 (31%) | 30 (52%) | .03 |
| OR time, min | 199 ± 1.5 | 173 ± 1.7 | .12 |
| Maternal LOS, d | 7.6 ± 1.8 | 5.7 ± 1.7 | .02 |
| Surgical complications | 15 (25%) | 12 (21%) | .44 |
| UAB complications | 2 (3%) | 0 | N/A |
There were 17 cases in which accreta was diagnosed prenatally but the patients did not receive UABs. These patients delivered at a significantly earlier gestational age (31.1 vs 34.0 weeks). Although the mean EBL trended higher (3025 mL vs 2605 mL; P = .19), this was not statistically significant.
In the subgroup that had the balloons inflated intraoperatively because of hemorrhage, the EBL of 2731 mL compared with 1702 mL in the group without inflation of the balloons. There was also a greater need for transfusion (28 vs 18; P = .005) and larger amounts of packed red blood cells transfused in the subset of patients in which the balloons were inflated (5.7 u vs 3.4 u, P = .02). No significant difference in the incidence of percreta versus accreta was seen in the group that had UABs inflated.
No significant difference in operative time was seen in cases in which balloons were placed. Hospital length of stay was significantly longer for the group that received UAB (7.6 vs 5.7 days; P = .02). Surgical complications were not different between groups and 2 patients (3.3%) had complications specific to the UAB: 1 with a femoral artery clot requiring thrombectomy without bypass and 1 superficial catheter site hematoma treated supportively.
Neonatal birthweight and neonatal intensive care unit length of stay did not differ between groups ( Table 3 ).
