Obstetric hemorrhage survey: Attitudes and practices of maternal-fetal medicine fellows

1 Introduction

Postpartum hemorrhage (PPH), which occurs in approximately 1 to 5% of pregnancies, has increased by 26% over the past few decades [1, 2]. PPH may be either unplanned or anticipated, depending on the clinical situation. PPH often occurs in situations of persistent uterine atony or abnormal placentation (i.e. placenta accreta/increta/percreta). Modern management of severe PPH initially includes use of uterotonic agents and uterine tamponade devices. There are case series that describe usage of the uterine tamponade device in Europe but they do not provide exact details on how providers would deflate the device [3, 4]. Alternative interventions for severe PPH include surgical ligation/compression techniques and assistance from interventional radiology.

If conservative methods fail during a PPH, then surgeons will resort to a peripartum hysterectomy. The background rate of peripartum hysterectomy in a review of Maternal-Fetal Medicine Units Network studies that included over 70000 women was as high as 0.6% in the setting of primary cesarean delivery [5]. There are no studies that examine physician comfort to deal with situations of peripartum hysterectomy. Furthermore, there is no information about management practices of severe PPH specifically at higher volume tertiary care centers. Given that most patient referrals for abnormal placentation diagnosed by imaging are sent to these centers, practice patterns at these centers would be of particular interest.

Our study aimed to determine the attitudes and practices of current maternal-fetal medicine (MFM) fellows at tertiary care centers across the US in situations of unplanned or anticipated obstetric hemorrhage.

2 Materials and methods

We performed a cross-sectional anonymous survey of first year MFM fellows at the society of MFM (SMFM) annual retreat. Permission was obtained from SMFM and the institutional review board (IRB) was considered exempt from full review by the Duke University IRB. The survey was validated for use by one of the authors (LRB) and a fellow from the corresponding author’s institution who did not respond to the final survey administered. Of the 101 first year MFM fellows, 92 were in attendance of this retreat. Six out of the nine who did not attend the retreat were sent the survey electronically via SMFM administrative staff approximately two weeks after the retreat.

Fellows were asked about management strategies and experiences involving unplanned and anticipated obstetric hemorrhage. Questions focused on three main areas: 1) conservative measures such as uterine tamponade device or antifibrinolytic therapy; 2) preoperative consultations, imaging and interventions; and 3) comfort level with peripartum hysterectomies. Questions were directed at fellow experiences, how-ever, questions about complex management of abnormal placentation or use of adjuvant therapies were direct reflections of institutional practice. MFM fellows almost uniformly had clinical responsibilities during their first year of fellowship and this survey took place about three-quarters into the first year (in March). This was the optimal time period to ask about current institutional practices as they were directly involved in the clinical experiences.

Survey responses were compared based on region of academic institution, prior clinical experience (None, 1–3 years, 4–6 years and 7–10 years), and fellow experience with hysterectomies. Regions were self-selected by responders as East (Northeast and Southeast), Mid (Midwest) and West (Northwest and Southwest). Analysis of Variance was used to test for differences in continuous variables and Fisher’s exact test was used to test for differences in categorical variables. Logistic regression was used to examine the effect of previous clinical experience on responders’ feeling of adequate training to act as attending. A two-sided significance level of 0.05 was used for all statistical tests. All statistical analyses were conducted using SAS version 9.3 (SAS Institute, Cary, NC).

3 Results

The response rate for completed surveys was 56% (55/98). Demographic data of survey responders is presented in Table 1. No participant had used an antifibrinolytic agent (such as tranexamic acid or 6-aminocaproic acid) in the setting of a severe PPH. Almost all participants (98%) responded that their institutions had ‘Hemorrhage’ or ‘Massive Transfusion Protocols’. The majority of participants responded that they had used a uterine tamponade device (specifically asked as ‘Bakri balloon^®’ in the survey) during fellowship in the setting of postpartum hemorrhage to help avert a periparturm hysterectomy (None 16%, 1–4 balloons 71% and >4 balloons 13%). More specifically, when asked how the Bakri balloon^® was used, there was a wide variation in the frequency and the management of the device once placed (Fig. 1). The median incremental time for balloon deflation was every 5 hours (IQR 2–12).

In situations of anticipated hemorrhage due to an abnormal placentation, preoperative consultations were obtained most commonly from the gynecology oncology and anesthesia services (Table 2a). Interventional radiology (IR) was either ‘Never’ or ‘Rarely’ consulted by 35% of fellows in the preoperative planning phases. An MRI was obtained as a complement to antenatal ultrasound diagnosis ‘Often’ or ‘Always’ in 35% of the time (Table 2b). IR involvement either prior to or during the surgery was variable. Preoperative ureteral stents were ‘Never’ or ‘Rarely’ used by 60% of responders. Fifty-six percent of participants responded that their institution never recommended planned delayed hysterectomies. On the other hand, if a placenta was left in situ for conservative management, 36% responded that they gave methotrexate in the postpartum period.

Finally, surgeon role and comfort level was assessed. Use of a handheld cautery device such as Ligasure^® or Gyrus^® during a peripartum hysterectomy performed in fellowship occurred by 29% of fellows. For the peripartum hysterectomies done in fellowship, fellows were the primary surgeon (29%), first assistant (13%) or second assistant (22%) for the entire surgery. Figure 2 illustrates the number of respondents who stated they felt adequately trained to perform peripartum hysterectomy as the supervising attending, both with and without gynecology oncology service back up.

Fellows who had previous clinical experience outside of residency were 5 times (OR 5.1, 95% CI 1.1, 23.1, p = 0.03) more likely to be comfortable as the supervising physician for a peripartum hysterectomy. The number of years in practice prior to fellowship was not significantly associated with the responses for any of the other questions asked on this survey. The number of peripartum hysterectomies was statistically different by geography when comparing across ‘East’ ‘Mid’ and ‘West’ regions (1.2±1.2, 1.3±1.6 and 2.9±2.4, respectively, p = 0.005). No other response on this survey was significantly associated with geographical distribution.

4 Discussion

We sought to describe attitudes and practices among first year MFM fellows at tertiary care centers related to obstetric hemorrhage. Although a majority of fellows used a uterine tamponade device to manage a severe PPH, there was wide variation in terms of how they deflated it. In the American Congress of Obstetricians and Gynecologists (ACOG) practice bulletin on PPH, specific guidelines on management of a uterine tamponade device once it is placed are not provided [6]. Furthermore, the guidelines do not encourage preparation for hemorrhage, by requesting standby blood products or consideration of antifibrinolytic therapy, in situations other than suspected abnormal placentation. ACOG and SMFM additionally have opinion articles on the management of placenta accreta [7, 8]. Most of the recommendations provided in these articles are supported by level B or C evidence data. In general, the lack of knowledge on how to best manage these clinical situations is reflected by the limited information national guidelines provide about how to manage severe PPH.

A few small studies have looked at prophylactic interventions such as ureteral stents or IR balloon insertion to decrease comorbidities but these techniques are not universal [9, 10]. Our study showed that about 20 to 40% of responders did use prophylactic interventions such as ureteral stents or IR involvement. However, the frequency of IR involvement was variable across respondents. Esakof et al.described common practices among Society of Maternal Fetal Medicine (SMFM) members, both in academic and private practice settings, for the management of placenta accreta [11]. Of the 361 providers who responded, about 35% used ureteral stents and 36% used internal femoral artery balloons.

A conservative approach, whereby the placenta is left in situ and a delayed hysterectomy is performed, has been described in the US mostly in case reports or series, yet our study showed that almost half of the fellow respondents had either ‘always, often, sometimes or rarely’ used this method [12, 13]. In the Esakofet al. study, only 32% of providers attempted conservative management (i.e. leaving a portion or the entire placenta in situ) for placenta accreta [11].

Compared to this study, providers in the Esakoff study had a wider range of prior clinical experience and were of academic and private practice backgrounds. It is difficult to conclude from their study that variations in practice are not primarily due to prior level of provider experience. We showed that after selecting a less diverse cohort of individuals, variations in practice of how to manage obstetric hemorrhage still persisted.

Our study is the first of its kind to inquire about comfort level of physicians to manage postpartum hemorrhage when peripartum hysterectomy is indicated. Although our survey included only first year MFM fellows, these fellows had already had been through three-quarters of their first year in fellowship, which for most was clinically focused. The low level of comfort reported by fellows to perform a peripartum hysterectomy as the supervising attending is likely a reflection of surgical overlap with other subspecialties such as gynecology oncology. Furthermore, it signals towards a trend to even specialize within MFM subspecialty, where only certain providers at tertiary care centers are ‘designated’ to perform these cases.

Postpartum hemorrhage may be unplanned or anticipated and causes serious maternal morbidity and mortality. Preoperative preparation and intraoperative practices are variable across first year MFM fellows surveyed. Additional evidence based studies are needed to evaluate effectiveness of uterine tamponade devices, hemorrhage protocols and antifibrinolytic therapy in averting peripartum hysterectomy. Updated hemorrhage guidelines should include suggested algorithms for using uterine tamponade devices. Peripartum hysterectomy is a definitive form of treatment for persistent obstetric hemorrhage. The fellows in this survey expressed some reservations with operating independently to perform this procedure. Future work should address the role of delayed hysterectomies and how subspecialties can work together to surgically manage abnormal placentation.

Funding sources

None

Conflicts of interest

The authors declare no conflict of interest.