Placenta accreta spectrum is a disorder of abnormal adherence of placental trophoblasts to the uterine myometrium.1 The incidence of placenta accreta spectrum has increased significantly over time, tracking with the increased rate of cesarean delivery.2,3 Many cases of placenta accreta spectrum are managed surgically by planned hysterectomy at the time of cesarean delivery.1 However, cesarean hysterectomy in the setting of placenta accreta spectrum is technically complex and can result in significant surgical morbidity, including genitourinary injury.1 Reported risks of urinary tract injury vary, with estimates of cystotomy in 18–76% of surgeries,4–7 ureteral injury in 0–18%,4–7 and genitourinary fistula in 0–5%.5–7

Whether genitourinary injury can be prevented with prophylactic ureteral stent placement is unclear, and previous work provides conflicting data.4,5,7 The International Society for Placenta Accreta Spectrum states that prophylactic stents may be beneficial in some cases, but they do not recommend routine placement of stents in placenta accreta spectrum surgery.8 Given this knowledge gap, our objective was to evaluate whether prophylactic ureteral stent placement was associated with genitourinary injury during hysterectomy for placenta accreta spectrum.

METHODS

We conducted a retrospective cohort study of all individuals in our institution's Placenta Accreta Clinical Database meeting inclusion criteria between 2001 and 2021. We included individuals who underwent hysterectomy during or immediately after delivery at one of our two participating institutions and had clinically or histopathologically confirmed placenta accreta spectrum.

The Placenta Accreta Clinical Database includes all pregnant individuals at risk for placenta accreta spectrum (including those with multiple prior cesarean deliveries or placenta previa). Demographic, clinical, and surgical data are abstracted for each case from the medical record by trained abstractors using identical REDCap secure data-capture forms hosted at each institution.9,10 All variables used in this study were abstracted from the medical record for the database unless otherwise noted. For this study, we performed individual chart abstraction to confirm the primary exposure and outcomes.

The exposure was prophylactic ureteral stent placement, defined as ureteral stent placement occurring before hysterectomy, either before or after delivery of the neonate. This was determined based on operative reports and captured during medical record abstraction. If prophylactic ureteral stents were attempted but placement was unsuccessful, the case was classified as no ureteral stent placement. At our institutions, the use of stents was not dictated by protocol before September 2018, so use of stents was determined on a case-by-case basis and at the discretion of the attending surgeon. In September 2018, our surgical group standardized the approach and agreed to perform ureteral stents routinely when feasible. Attending surgeon subspecialties included general obstetrics and gynecology, minimally invasive gynecology, maternal–fetal medicine, and gynecologic oncology.

Our primary outcome, genitourinary injury, was a composite of any bladder injury or ureteral injury identified intraoperatively or postoperatively or vesicovaginal fistula recognized in the first 6 months after delivery. Secondary outcomes included intraoperative estimated blood loss, intentional and unintentional cystotomy, coagulopathy (as defined by the clinical team), operating room time (from start of first procedure to skin closing), maternal death, acute kidney injury (creatinine level 1.0 or doubling from baseline), urinary tract injury, need for dialysis, postpartum length of stay, postoperative hospital discharge with an indwelling catheter, and secondary urologic surgery in the 6 months after delivery for fistula repair or delayed recognition of genitourinary tract injury. These outcomes were determined by review of all operative reports and postpartum notes. Patients recovering from placenta accreta spectrum surgery routinely have multiple postoperative visits with a member of the surgical team, and at least one postoperative follow-up note was available for all patients. Patients in this cohort also complete postoperative surveys that screen for ongoing genitourinary problems and additional surgeries at 6 and 12 months postpartum and annually for 5 years.

Covariates included urgency of delivery (defined as emergent, urgent, or scheduled), whether placenta accreta spectrum was diagnosed antenatally, number of prior cesarean deliveries, clinical or histopathologic disease severity (accreta [International Federation of Gynecology and Obstetrics (FIGO) 1, placenta accreta spectrum (PAS) 1], increta [FIGO 2, PAS 2], or percreta [FIGO 3, PAS 3]),11 delivery type (vaginal or cesarean), and presence of a gynecologic oncologist as the primary surgeon. Emergent cases were defined as those in which patients required immediate transport to the operating room for delivery or planned cesarean delivery that required immediate conversion to hysterectomy for placenta accreta spectrum that was previously undiagnosed. All vaginal deliveries that were converted to hysterectomies were considered to be emergent cases. Urgent cases were defined as those in which patients were deemed to require delivery or cesarean hysterectomy within several hours of the decision to deliver but were not emergent, requiring immediate intervention. We could not ascertain the severity of disease suspected antenatally in all cases during the study period based on medical record review, so instead we relied on clinical (surgical) or histopathologic grade as a marker for disease severity.

We evaluated demographic and clinical differences between groups using χ2 tests for categorical measures and t tests for continuous measures. To evaluate differences in the primary outcome (genitourinary injury) between patients with and without stent placement, we used logistic regression to report unadjusted odds ratios and 95% CIs. To control for clinically important characteristics, multivariable logistic regression was used, from which adjusted odds ratios (aOR) and 95% CIs are reported. Two adjusted models were estimated. The first adjusted model (adjusted model 1) included covariates that were statistically different between groups and the presence of a gynecologic oncologist as the primary surgeon (as an a priori–defined covariate). The second model (adjusted model 2) added placenta accreta spectrum grade (two categories: PAS 2 [increta] or PAS 3 [percreta] vs PAS 1 [accreta]), because a priori guidelines and studies suggest that ureteral stents may be more useful in more severe placenta accreta spectrum surgical cases. In a sensitivity analysis, we evaluated the same associations excluding intentional cystotomy from the primary outcome.

To evaluate trends in stent use and genitourinary injury over time, we performed a Cochran-Armitage χ2 trend test. Time was grouped into 3-year-increments before 2012, when deliveries meeting inclusion criteria were less frequent, and 1-year increments after 2012. Data analysis was completed using SAS 9.4. Graphics were created using GraphPad Prism 9.1.2. The University of Utah and the Intermountain Medical Center IRBs approved this study.

RESULTS

In total, 236 patients were included in this analysis (Table 1). Stent placement was significantly associated with three or more prior cesarean deliveries and urgency of delivery, with scheduled cesarean deliveries occurring more frequently among patients with stents. Vaginal delivery was less common in the stent group. Placenta accreta spectrum discovered at delivery without an antenatal diagnosis was less common in the stent group. Other baseline maternal and obstetric characteristics did not differ between groups.

Table 1.:

Clinical Characteristics

The primary outcome (genitourinary injury) occurred less frequently in the stent group compared with the no stent group (28% vs 51%; odds ratio 0.37, 95% CI 0.21–0.65, Table 2). The association between ureteral stent placement and the primary outcome persisted after controlling for clinical covariates including urgency of delivery, three or more prior cesarean deliveries, and whether a gynecologic oncologist was the primary surgeon (multivariable aOR 0.27, 95% CI 0.14–0.52). When controlling for disease severity in addition to the covariates in the first adjusted model, the multivariable aOR was 0.21 (95% CI 0.10–0.43).

Table 2.:

Primary and Secondary Outcomes Among Those With and Without Prophylactic Ureteral Stent Placement at the Time of Hysterectomy for Placenta Accreta Spectrum

Stent placement was attempted but was unsuccessful in five patients in the no stent group. Of these cases, four (80%) were complicated by genitourinary injury, including two patients with bladder injury, one with bladder injury and vesicovaginal fistula, and one with bladder injury and ureteral injury.

Among 84 instances of bladder injury, 54% were intentional and 48% were unintentional, with one case having both intentional and unintentional injury. In a sensitivity analysis excluding cases of intentional cystotomy, differences were in the same direction and remained significant but were of a smaller magnitude (Fig. 1). Components of the primary outcome (bladder injury and ureteral injury) occurred less frequently in the stent group (Table 2). The incidence of vesicovaginal fistula was not statistically different between groups. Stent use was associated with lower intraoperative estimated blood loss but longer mean operative time. It was also associated with a shorter mean postpartum length of stay (Table 2).

Fig. 1.:

Decreased odds of genitourinary injury with prophylactic ureteral stent use. Adjusted model 1 includes delivery urgency (scheduled, urgent, or emergent), history of cesarean delivery (0–2 vs 3 or more), and whether the primary surgeon was a gynecologic oncologist. Adjusted model 2 adds disease severity (accreta [International Federation of Gynecology and Obstetrics (FIGO) 1, placenta accreta spectrum (PAS)] 1, increta [FIGO 2, PAS 2], or percreta [FIGO 3, PAS 3]). The sensitivity analysis included the same components, but the outcome excluded intentional cystotomy. Vaginal delivery and missed antenatal diagnosis were not included in the modeling because these occurred infrequently in one or both groups (cell count less than 5). Circles indicate primary outcomes; diamonds indicated primary outcomes with intentional cystotomy.

Over time at our institutions, prophylactic ureteral stent placement increased and urinary tract injury decreased (Fig. 2). Stent use increased from 50% in 2000–2002 to 94% in 2020 (P<.001 for test of trend). During the same time, urinary tract injury rates decreased (P<.001 for test of trend).

Fig. 2.:

Stent use (blue dashed line) and genitourinary tract injury (black solid line) by year, represented as percentage of total cases. Stent use increased over time (P<.001). The composite genitourinary injury outcome decreased over time (P<.001).

DISCUSSION

In this large retrospective cohort, placement of prophylactic ureteral stents was associated with a decrease of nearly 80% in the odds of genitourinary injury during hysterectomy performed for placenta accreta spectrum, even after controlling for differences in delivery urgency, number of prior cesarean deliveries, primary surgeon subspecialty, and disease severity.

Previous studies come to conflicting conclusions regarding the association between ureteral stent placement and genitourinary injury. In 57 patients with placenta accreta spectrum, Eller et al4 found no difference in ureteral injury (0% vs 7%, P=.31). In 44 patients with placenta accreta spectrum, Crocetto et al5 found no difference in urinary tract injury at the time of hysterectomy (25% vs 10%, P=.21). A recent case–control study of 292 patients with and without urologic injury at the time of hysterectomy for placenta accreta spectrum found no protective association of prophylactic ureteral stents (which had been used in the majority of placenta accreta spectrum cases).6 In contrast, a 2012 systematic literature review of 90 cases found a significant reduction in genitourinary injuries with ureteral stent placement compared with without ureteral stent placement (6% vs 33%, P=.01).7

Notably, we found decreased occurrence of bladder injury in addition to ureteral injury with placement of ureteral stents. The cystoscopy necessary for stent placement allows for assessment of the bladder, which is often significantly distorted, potentially giving the surgical team anatomic information about disease location and severity. Additionally, ureteral stents are easily palpated, allowing surgeons to avoid ureteral damage during difficult and sometimes rapid pelvic surgery. Finally, we have experienced many cases in which ureteral stents help to define the lateral and inferior angle of the bladder, which aided in safer dissection of the bladder off the lower uterus and cervix. It is worth noting that, of our five unsuccessful ureteral stent placements, 80% had genitourinary injury. Therefore, those in whom stent placement is unsuccessful may be at high risk of injury, although there were too few instances to support this supposition statistically.

The longer operating room time in the stent group is worth noting. Longer surgery due to cystoscopy and stenting is an important consideration, because it may be associated with additional morbidity and cost. This could be due to unmeasured confounding (antenatal suspicion for anticipated difficult surgery might be positively associated with both the exposure and the outcome). One solution to this time problem may be to place stents concurrently with the start of the cesarean delivery, which we have begun to do more often.

We included cystotomy in our primary outcome because, although it is relatively minor compared with ureteral injury or fistula, bladder injury is meaningful. Those with bladder injury require prolonged bladder catheterization, increased follow-up appointments, and likely increased cost of care.

Prior data suggest that patients with more severe histopathologic disease are more likely to benefit from prophylactic stent placement.6 In contrast, our study found that the odds of urinary tract injury were essentially unchanged when controlling for disease severity, suggesting that patients with all degrees of disease severity benefit from prophylactic stents.

The strengths of this study include its cohort design and relatively large number of patients compared with prior studies. The conclusion of our primary analysis is strengthened by two additional analyses showing a similar protective association: 1) a time-trend analysis and 2) a sensitivity analysis of the primary outcome excluding intentional cystotomy. Finally, trained abstractors collected and verified all data from the electronic medical record, increasing the reliability of both the outcome and the exposure measures.

Our study has important limitations, too. First, as a retrospective study, there is potential for bias and unmeasured confounding. Although we used statistical modeling to control for important covariates, we cannot account for unmeasured confounding variables, particularly those related to improved surgical experience and expertise over time. However, the most likely direction of systematic bias in this setting would be toward the null, because prophylactic ureteral stents are likely placed in the most complex cases. Second, it is possible that some complications, including vesicovaginal fistula, were not detected. Reassuringly, as the major referral centers for complex surgical care in our region, it is likely that postoperative complications would have been referred back to one of our two large health systems. Finally, our analysis may not be generalizable to smaller centers or regions that differ substantially from our geographic location.

In summary, prophylactic ureteral stent placement was associated with decreased odds of genitourinary injury during hysterectomy for placenta accreta spectrum.

REFERENCES 1. Silver RM, Branch DW. Placenta accreta spectrum. N Engl J Med 2018;378:1529–36. doi: 10.1056/NEJMcp1709324 2. Wu S, Kocherginsky M, Hibbard JU. Abnormal placentation: twenty-year analysis. Am J Obstet Gynecol 2005;192:1458–61. doi: 10.1016/j.ajog.2004.12.074 3. Publications Committee, Society for Maternal-Fetal Medicine, Belfort MA. Placenta accreta. Am J Obstet Gynecol 2010;203:430–9. doi: 10.1016/j.ajog.2010.09.013 4. Eller AG, Porter TF, Soisson P, Silver RM. Optimal management strategies for placenta accreta. BJOG 2009;116:648–54. doi: 10.1111/j.1471-0528.2008.02037.x 5. Crocetto F, Esposito R, Saccone G, Della Corte L, Sarno L, Morlando M, et al. Use of routine ureteral stents in cesarean hysterectomy for placenta accreta. J Matern Fetal Neonatal Med 2021;34:386–89. doi: 10.1080/14767058.2019.1609935 6. Erfani H, Salmanian B, Fox KA, Coburn M, Meshinchiasl N, Shamshirsaz AA, et al. Urologic morbidity associated with placenta accreta spectrum surgeries: single-center experience with a multidisciplinary team. Am J Obstet Gynecol 2022;226:245.e1–5. doi: 10.1016/j.ajog.2021.08.010 7. Tam Tam KB, Dozier J, Martin JN. Approaches to reduce urinary tract injury during management of placenta accreta, increta, and percreta: a systematic review. J Matern Fetal Neonatal Med 2012;25:329–34. doi: 10.3109/14767058.2011.576720 8. Collins SL, Alemdar B, Van Beekhuizen HJ, Bertholdt C, Braun T, Calda P, et al. Evidence-based guidelines for the management of abnormally invasive placenta: recommendations from the International Society for Abnormally Invasive Placenta. Am J Obstet Gynecol 2019;220:511–26. doi: 10.1016/j.ajog.2019.02.054 9. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)- a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377–81. doi: 10.1016/j.jbi.2008.08.010 10. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O'Neal L, et al. The REDCap consortium: building an international community of software platform partners. J Biomed Inform 2019;95:103208. doi: 10.1016/j.jbi.2019.103208 11. Hecht JL, Baergen R, Ernst LM, Katzman PJ, Jacques SM, Jauniaux E, et al. Classification and reporting guidelines for the pathology diagnosis of placenta accreta spectrum (PAS) disorders: recommendations from an expert panel. Mod Pathol 2020;33:2382–96. doi: 10.1038/s41379-020-0569-1