Regional multidisciplinary team approach to the management of placenta accreta spectrum disorder

Abstract Objective Patients with suspected placenta accreta spectrum (PAS) disorder are often referred to specialized medical centers for antepartum management and definitive treatment via cesarean hysterectomy. In 2019, our institution formed the only multidisciplinary team for the management of PAS within two of the largest counties in California. The purpose of this study was to evaluate the effects of the multidisciplinary team on patient volume and surgical outcomes for patients with PAS. Methods This was a single center retrospective cohort study, based in the only tertiary referral center within two of the largest counties in California. Patients who underwent cesarean hysterectomy for suspected PAS from January 2014 to April 2021 were included and divided into two groups, based on management by the multidisciplinary team from January 2019 and onward or routine care prior to that time. The outcomes of interest were quantitative blood loss, total units of packed red blood cell transfusion, referral volume, and diagnostic accuracy as well as ICU admission, bladder injury, and postoperative length of stay. Furthermore, we wanted to determine if patient’s distance to the hospital impacted outcomes. Normally distributed variables were compared between groups using the t-test. Categorical variables were compared between the two groups using the chi square test. Results A total of 114 patients were included in the cohort, 59 patients were from January 2014 to December 2018 and 55 patients were from January 2019 to April 2021. Since the establishment of the multidisciplinary center, there was a 2.5-fold increase in the total patient volume (0.8 case/month to 2 cases/month) and a 2.8-fold increase in the referred patient volume. Patients undergoing cesarean hysterectomy since the establishment of the multidisciplinary team had less quantitative blood loss (1500 mL vs 2000 mL, p = .005) and required less units of packed red blood cell transfusion (2 vs 4 units, p < .001). In addition, blood loss of ≥2000 mL decreased from 57.6% to 38.2% (p = .04) and diagnostic accuracy improved from 35.6% to 83.6% (p < .001). Furthermore, we found that patient distance to the hospital did not significantly impact surgical outcomes. Conclusions Since the establishment of the multidisciplinary team, our center experienced an increase in PAS volume and was able to demonstrate a statistically significant improvement in patient outcomes.


Introduction
The incidence of placenta accreta spectrum (PAS) disorder is estimated to be 0.3% in patients with one prior cesarean delivery (CD) and increases up to 6.74% in women with five or more CDs, with even higher incidence in the presence of placenta previa [1,2]. Available data suggest that up to 95% of cases require blood transfusions [3] and nearly all PAS cases require hysterectomy for definitive treatment [4]. PAS is associated with a peripartum mortality rate of 8% [5], however it can reach 30% in the absence of accurate antenatal diagnosis [6]. Even more distressing is the fact that only half of PAS cases are diagnosed antenatally based on large population studies [7][8][9][10]. Thus, outcomes of patients with PAS are directly related to accurate antenatal identification and antepartum care.
In order to improve clinical outcomes of patients diagnosed with this condition, the ACOG/SMFM published a checklist and guidance for clinical management and surgical planning for cases of PAS [11]. Many individual components of a multidisciplinary care team had previously been shown to improve maternal outcomes [12][13][14][15][16]. However, most of these studies were conducted in established centers located in urban settings, in which a concentrated population density meant easier patient follow-ups and decreased transportation times in the event of emergencies. In a rural/suburban setting, the geographic distance to established PAS centers and lack of reliable transportation may hinder access to clinical expertise. As a result, management of PAS may fall on local hospitals with limited resources and experience. Furthermore, despite data showing the benefits of delivering at highvolume centers with multidisciplinary teams, many patients with PAS delivered outside of these centers [17]. A survey study of Fellows of the American College of Obstetricians and Gynecologist found that only 23.8% of general practitioners referred patients to a sub-specialist with an overall 20.4% referred patients to the nearest tertiary center and 7.1% to a regional center [17]. This study illustrated the wide variation in care for patients with PAS in the community. We believe that some of the low referral rate is due to lack of access to specialized centers for PAS for patients living in rural and suburban areas. As an example, in Southern California, the centers for PAS are concentrated in the coastal regions. Thus, patients living more inland from the coast face the challenge of long commutes, up to four hours or more, to see the specialists there. In the event of emergencies, transporting patients to those centers may not be feasible. Therefore, in 2019, we established a multidisciplinary center for the management of PAS in a tertiary medical center located in the inland suburban and rural communities of Southern California.
We hypothesize that establishing such a center for the management of PAS in an underserved suburban/rural region will help improve clinical outcomes of patients diagnosed with PAS. Furthermore, the establishment of such center would improve outcomes for even the most remote patients with access to our center. The objective of this study is to evaluate the clinical outcomes of PAS before and after the establishment of a rural/suburban multidisciplinary center.

Materials and methods
This was a single-center retrospective cohort study of patients who underwent cesarean hysterectomy for suspected PAS based on antenatal sonographic evaluation, using a pre-and post-exposure analysis of a multidisciplinary team. This study was based at Loma Linda University Children's Hospital, the only university-affiliated tertiary referral center in an underserved suburban and rural area of Southern California -Riverside and San Bernardino counties, which span from the inland Southern regions to the Mojave Desert regions with a combined population of 4.56 million people [18]. Patients who underwent cesarean hysterectomy for suspected PAS between January 2014 and April 2021 were included in this study. This study was approved by the institutional review board (IRB# 520052).
Prior to 2019, our medical center, though being the tertiary referral center of the region, did not have a specialized multidisciplinary team dedicated to the management of patients with PAS. The closest specialized centers caring for this population were 60-100 miles away, thus patients living in Riverside and San Bernardino counties, two of the most populous counties in California, often face the geographic challenge to assessing specialized medical care for this condition. Because of this, in 2019, our medical center situated within San Bernardino County, formalized a dedicated multidisciplinary team from existing faculty members for the management of PAS. Antenatal diagnosis and surgical management conformed to the Society for Maternal Fetal Medicine Obstetric Care Consensus recommendations [11]. The team included designated members from Maternal-fetal Medicine (MFM), Obstetrical care services, Gynecologic Oncology, Obstetric Anesthesiology, Acute Care Surgery, and Radiology. The MFM member of the team was responsible for the evaluation, diagnosis, and coordination of care of all suspected PAS cases. The antepartum diagnosis of PAS was made primarily based on ultrasound evaluation of the placenta (Supplementary Table 1). MRI was not routinely used in all cases. The team met at regular interval and suspected PAS cases were reviewed for surgical planning with the entire team. Surgical planning included the type and length of abdominal incision, hysterotomy entry site and technique, evaluation for the need for optimizing preoperative hemoglobin, intraoperative cell salvage by MFM and Gynecologic Oncology, as well as the placement of resuscitative endovascular balloon occlusion of the aorta by Acute Care Surgery [19]. All PAS patients received antepartum consultations with Gynecologic Oncology, Obstetric Anesthesiology and Neonatology teams. Gynecologic Oncology team was notified of all admitted PAS patients with daily monitoring of vaginal bleeding and contraction symptoms. MFM and Gynecologic Oncology surgeons jointly performed all cesarean hysterectomies. The multidisciplinary team met at regular intervals to review recent and upcoming cases and to identify areas for improvement. Supplementary Table  2 includes the resources crucial to the multidisciplinary team care approach for PAS. Additionally, the director of the program routinely reached out to community obstetricians to provide education regarding the diagnosis and management of PAS as well as to update the providers on their referred patients.
In order to evaluate the effects of coordinated multidisciplinary management of PAS, the cohort was divided into two groups for analysis. The control group included cases managed between January 2014 and December 2018, representing outcomes prior to the inception of the multidisciplinary team treatment approach. Cases managed from January 2019 and thereafter represented outcomes since the multidisciplinary team approach. Medical records were reviewed to abstract basic demographic information as well as clinical outcomes data. Maternal demographics were compared between the two comparison groups using the appropriate univariate statistical tests. Linear variables were tested for normal distribution using the Shapiro-Wilk test. Normally distributed variables were compared between groups using the t-test, otherwise the Kruskal-Wallis test was used. Categorical variables were compared between the two groups using the chi square test. Statistical significance was defined as pvalue <.05.
The primary outcomes of interest included quantitative blood loss (QBL), units of transfused packed red blood cells (PRBCs), referral volume, and diagnostic accuracy; secondary outcomes include intensive care unit (ICU) admission, postoperative length of stay, unintentional bladder injury. The QBL was determined based on the suction canister volume, laparotomy sponge count, and clinical assessment of any other unaccounted blood loss. The accuracy of increta/percreta diagnosis was evaluated by comparing the final pathologic diagnosis to the antenatal sonographic diagnosis. The rates of antenatally missed PAS and suspected PAS that were not confirmed by pathology were separately determined. The primary outcomes were compared between the two groups using the appropriate univariate statistical tests. The rate of blood loss 2000 mL, PRBC transfusion of 4units, and cases that did not require transfusion of any blood product were further evaluated. The risks of these outcomes were estimated using logistic regression and adjusted for statistically significant clinical variables. Furthermore, the rate of surgical complications were separately calculated for patients living more than 50 miles (80.5 km) and patients living within that distance. Statistical significance was defined as p-value < .05. Statistical analyses were performed using the Stata 14 (College Station, TX).

Results
Between January 2014 and April 2021, there were 114 patients who underwent cesarean hysterectomy for suspected PAS. Of these cases, 59 (51.3%) were performed between January 2014 and December 2018, and 55 (48.7%) were performed between January 2019 and April 2021. The average number of PAS cases increased from 0.8 case per month before January 2019 to 2.0 cases per month since the institution of the multidisciplinary PAS program, an increase of 2.5 folds. The referred patient volume increased from 0.5 case per month to 1.4 case per month, an increase of 2.8 folds. Comparing the two groups, cases performed since 2019 were more likely to have placenta increta or percreta on final pathology diagnosis (92.9% vs 72.0%, p ¼ .005) and the median gestation age at delivery was earlier by 1 week (33 weeks vs 34 weeks, p ¼ .04). Patient distance from the hospital were similar before and after implementing the multidisciplinary program, with the furthest patient referral from 175 miles (282 km) away. All other demographic variables were similar between the two groups (Supplementary Table 3).
Since the establishment of the multidisciplinary team at our center, a statistically significant reduction in intraoperative median blood loss were noted, 1500 mL vs 2000 mL, p ¼ .005. Percentage of patients requiring 4 units of red blood cells was significantly lower than those undergoing surgery prior to January 2019 (34.6% vs 59.3%, p ¼ .01) with the median total transfused PRBCs of 2 units vs 4 units (p < .001). Postoperative hospital length of stay was also shorter (3 days vs 4 days, p < .001). Furthermore, the accuracy rate of increta/percreta diagnosis was significantly improved since January 2019 (83.6% vs 35.6%, p < .001), as confirmed by final pathology reports.
In addition, while the rate of undiagnosed PAS cases were similar in both groups, there were two cases of cesarean hysterectomy before January 2019 that did not yield histopathologically confirmed PAS diagnosis, and none since 2019 (Table 1).

Discussion
This study of 114 patients undergoing cesarean hysterectomy prior to and after the establishment of a multidisciplinary team demonstrates the increase in the referred patient volume, as well as the improvement in intraoperative blood loss, units of transfused PRBC, and antenatal diagnostic accuracy. Specifically, compared to prior PAS cases, those managed by the multidisciplinary team had reduced median blood loss of 2000 mL or more by nearly 20%, and reduce the need for transfusing 4 or more units of PRBCs by 42%. Furthermore, the rate of patient not requiring any blood products increased by 2.3 times. These improvements were observed despite the higher incidence of percreta and increta that the medical center had cared for.
PAS remains one of the most challenging pregnancy complications faced by obstetricians and colleagues for multiple reasons. These cases provide numerous technical challenges due to the neovascularization and hypervascularity of pelvic organs, complex fetal delivery, pelvic adhesive disease, maternal body habitus, and more. Specifically, available studies suggested that more than 95% of all PAS cases require blood transfusion [3] with the median estimation of blood loss of 2-3 L, median units of transfused PRBCs of 3.5-4.5, and up to 40% of the cases requiring largevolume blood transfusion of >10 L [5].
This study demonstrated that the multidisciplinary team approach to PAS management at our suburban/rural referral center improved outcomes from several aspects. Successful and safe maternal and fetal outcomes are directly related to antenatal identification of PAS. A recent metanalysis study had shown that antenatal diagnosis of PAS was associated with significant reduction in morbidities associated with blood loss, compared to cases that were detected only at the time of delivery [20]. Near-missed PAS cases were previously noted in cases involving the lower uterine segment, posterior bladder and parametria [21]. Early first trimester sonographic assessment of PAS in patients with history of previous cesarean sections in combination of second and third trimester ultrasounds  had been shown to reliably predict PAS and stratify surgical risks [22]. While ultrasonography is the standard diagnostic modality for PAS, its ability to accurately detect PAS is at least partly dependent on a high clinical index of suspicion. We speculate that establishing a multidisciplinary center increased awareness and index of suspicion for PAS within our own institution. It provided frequent reminders for the sonographers and reading physicians regarding this diagnosis. By consistently using established ultrasound protocol for PAS evaluation, we increased the accuracy of detecting severe PAS disease from 36% to 84%, a 2.3-fold increase. Through regularly scheduled meetings and case reviews, intraoperative finding were correlated with radiologic features to improve diagnostic accuracy of future cases. We further suggested that the improved diagnostic accuracy, a result of a consistent ultrasonography reader and cumulative experience over the years, had allowed for more detailed surgical planning.
Prior to the establishment of a multidisciplinary approach to antepartum and intraoperative management of PAS, the counties of San Bernardino and Riverside did not have a specialized center. Community-based obstetricians either managed these cases in local hospitals or inconsistently referred patients to our institution or other academic institutions further away. Our monthly number of cases has more than doubled since the program started, with an increase in referral volume by 2.8 times. We attribute this overall increase in total case volume and referral volume to the increased awareness of our program by obstetricians and MFM subspecialists in the community. Furthermore, we noted an increase in PAS cases with severe disease including increta and percreta. We speculate that the increase in disease severity could be explained by a preference by community physician or patients for a closer referral center. The increase in disease severity likely explained the lower median gestational age at delivery, as they are more likely to experience complications leading to earlier deliveries.
It is a well-known fact that repetition and experience improve outcomes over time. Thus, the increased volume and increased experience and practice over time could have positive contributory factors to the observed outcomes. We further attribute the improvement in surgical outcome of decreased blood loss and blood transfusion to the consistent team members, expertise, assurance of availability of necessary surgical equipment, and efficiency in teamwork of MFM, Gynecology Oncology, Obstetric Anesthesiology, and Acute Care Surgery intraoperatively. Our MFM and Gynecologic Oncology physicians reviewed all imaging to appreciate the extent of suspected disease, planed surgical approach accordingly, and co-performed all cesarean hysterectomies. Regular case reviews also assisted the team in identifying areas of improvement. These efforts in-turn helped the team anticipate and minimize future surgical complications.
This study demonstrated that establishing a specialized PAS multidisciplinary center in a tertiary referral hospital in a suburban/rural setting is achievable. Such a center in a relatively remote and resource-limited area could increase referral for management of PAS. Furthermore, despite limitations in transportation and other constraints for patients, such a center could demonstrate improved clinical outcomes for the management of PAS comparable to findings from urban centers. In this study, we found that patients living in remote areas benefited from improved outcomes as did patients living closer to the center (Supplementary Table 4). We would like to acknowledge that our institution is a tertiary referral center, thus we have blood bank availability, subspecialties such as Gynecologic Oncology and Acute Care Surgery, allowing for highly specialized care for this high risk condition. In community hospitals or rural hospital who do not have these resources, establishment of multidisciplinary care for this condition could be challenging.
One of the main strengths of this study is that the sample size was large enough to demonstrate the differences in surgical outcomes associated with instituting a multidisciplinary team for PAS management. This is one of the first studies to evaluate outcomes in patient living near an established PAS center and remote from the center. Furthermore, all 114 patients had complete data including demographic, sonographic, surgical and pathology records, allowing for detailed analysis of patient outcomes. Additionally, the dedicated multidisciplinary team consisted of members who were already involved in PAS pregnancies, although without formal organization and division of care and responsibilities. The organization of the team did not require additional resources or extensive time commitment. Therefore, such care models could be replicated in other tertiary centers with similar resources.
This study does have limitations. Given the relatively rare incidence of PAS and some of the severe maternal or perinatal outcomes, we could not evaluate for all of the adverse outcomes reported in literature. Furthermore, we were not able to determine the outcomes of PAS patients within the region who were managed by other local or regional hospitals. Therefore, we cannot conclusively compare and determine if establishing our PAS program reduced overall maternal morbidity within the entire region.