Practice Patterns and Complications of Benign Hysterectomy Following the FDA Statement Warning Against the Use of Power Morcellation

Key Points

Question  Did 30-day complication rates change among women undergoing hysterectomy for benign gynecologic indications following the US Food and Drug Administration warning against power morcellation?

Findings  In a cohort of 75 487 patients who underwent hysterectomy for benign gynecologic indications, statistically and clinically significant increases in major and minor complications were observed following the warning in a subset of 25 571 women with uterine fibroids.

Meaning  That potential harms were associated with decreased rates of minimally invasive surgery following reduced morcellation use should be considered by clinicians and patients during shared decision making and by regulatory bodies and medical societies issuing communications on minimally invasive hysterectomy and morcellation among women with uterine fibroids.

Importance  In November 2014, the US Food and Drug Administration (FDA) issued a black box warning against the use of power morcellation for excision of uterine fibroids to decrease the risk of disseminating malignant cells and worsening survival outcomes of patients with unexpected malignant neoplasms. After the FDA statement was issued, studies showed decreased rates of minimally invasive surgery and increased rates of open abdominal hysterectomy. However, there are limited and controversial data on the association of these changed rates with 30-day hysterectomy complications.

Objective  To assess changes in the rates of 30-day major and minor complications of hysterectomy for benign gynecologic indications following the FDA-issued statement.

Design, Setting, and Participants  This retrospective cohort study used the American College of Surgeons National Surgical Quality Improvement Program database and included 603 hospitals participating between January 1, 2013, and December 31, 2015. Thirty-day posthysterectomy complications were compared before and after the FDA-issued warning. Women who underwent hysterectomy for benign gynecologic indications (n = 75 487), including hysterectomies with indication of uterine fibroids (n = 25 571), were included. Complication rates and procedure distributions between the periods were compared with χ² tests and multivariable logistic regression controlling for patient and operative factors.

Main Outcomes and Measures  Major and minor 30-day complication rates before (from 2013 through the first quarter of 2014) and after (from the fourth quarter of 2014 through 2015) the FDA-issued warning.

Results  Of 75 487 women (mean [SD] age, 47.8 [10.7] years) who underwent hysterectomy for benign gynecologic indications, 32 186 (42.6%) were treated before and 43 301 (57.4%) after the FDA-issued warning. Non-Hispanic white women comprised most (59.4%) of the total population, followed by African American women (15.1%). Overall, major and minor complications remained stable before and after the FDA-issued warning. By contrast, among a subset of 25 571 women (33.9%) who underwent hysterectomy for uterine fibroids, major complications significantly increased after the FDA-issued warning from 1.9% to 2.4% (adjusted odds ratio [OR], 1.23; 95% CI, 1.04-1.47; P = .02), and minor complications significantly increased from 2.7% to 3.3% (adjusted OR, 1.21; 95% CI, 1.04-1.40; P = .01). In this subgroup, the rate of open abdominal surgery increased from 37.2% to 43.0%, and the rate of minimally invasive surgery (total laparoscopic hysterectomy, laparoscopic supracervical hysterectomy, and laparoscopic-assisted vaginal hysterectomy) decreased from 56.1% to 49.7% (P < .001).

Conclusions and Relevance  Major and minor 30-day complication rates among women undergoing hysterectomy for uterine fibroids increased following the FDA-issued statement. This increased risk corresponding with a decreased use of minimally invasive surgery should be balanced against the potential harms of morcellation during a shared decision-making process between clinician and patient. Regulatory bodies and medical societies should consider these findings when issuing relevant communications.

Hysterectomy is the most commonly performed gynecologic procedure, with more than 600 000 performed annually in the United States.^1,2 Uterine fibroids represent the most common indication, accounting for approximately 40% of all hysterectomies performed.³ By contrast, uterine leiomyosarcoma is a rare and aggressive cancer, occurring in 0.4 to 0.64 cases per 100 000 women,^4,5 that can present as a myometrial mass resembling a benign uterine fibroid. Presumed benign fibroids can be treated by medical and surgical approaches.

When an operation is required, most women prefer minimally invasive surgery (MIS) that is less invasive, has a short recovery time,⁶ and has lower associated complications compared with those following open abdominal surgical procedures.⁷ Minimally invasive surgical procedures, however, may require the morcellation of the uterus inside the peritoneal cavity to remove the uterus through a smaller incision or through the vagina. The process of morcellation carries the risk of disseminating benign or malignant disease.⁸ In the latter case, dissemination of malignant tissue would increase the risk of recurrence and shorten survival.^9,10 As a consequence, in April 2014, the US Food and Drug Administration (FDA) issued a safety communication statement “discouraging” the use of laparoscopic power morcellation for patients with uterine fibroids.¹¹ Subsequently, in November 2014, the FDA “warned against” the use of uterine power morcellation,¹² issuing “contraindications” for its use for “peri- and post-menopausal” women and for patients with “suspected or known malignancy.”

Following the issue of the FDA warning, many institutions discouraged the use of any morcellation and banned the use of the power morcellator; there was a subsequent decrease in the use of power morcellation during surgical procedures among women who underwent a minimally invasive hysterectomy from 13.7% in early 2013 to 2.8% in early 2015.¹³ Moreover, the results of 2 surveys of gynecologists evaluating the influence of the FDA warning on management strategies in hysterectomy and myomectomy showed that a large proportion of respondents shifted from the use of MIS to the use of a larger incision or open abdominal procedures, raising concerns about an increase of surgical complications.^14,15 Furthermore, the FDA decision was criticized for overestimating the risk associated with using power morcellation compared with the risk associated with preventing its use in a large group of women who could receive benefits from morcellation.¹⁶ However, evaluating the contribution of the FDA statement to complications of hysterectomy to date has provided controversial results.^13,17-19

Most importantly from a patient perspective, a survey distributed to more than 300 women in gynecologic waiting rooms regarding the consequence of limiting access to MIS and limiting women’s choices indicated that at the 1 in 350 risk of a malignant tumor reported by the FDA, the majority of women would still choose an MIS approach.²⁰

The present study used the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database to overcome the limitations of previous studies because data on complications up to 30 days after the index surgery are collected. We evaluated 30-day major and minor complications associated with hysterectomy before and after the FDA issued the warning statement about the use of power morcellators.

Details about sampling methods, data abstraction procedures, variables, and outcomes collected by the ACS NSQIP database are described elsewhere.²¹ In brief, the ACS NSQIP database is a prospective, validated, multicenter, clinical data registry designed to facilitate, measure, and enhance the quality of surgical care by providing hospitals with detailed reports of their risk-adjusted outcomes. The Mayo Clinic institutional review board deemed this study exempt from requiring approval because the data were deidentified.

Data on more than 150 variables, including demographic characteristics, preoperative risk factors, laboratory test results, operative details, and postoperative outcomes up to 30 days after the index operation are collected from 603 participating academic and nonacademic hospitals. Data collection is performed directly from the medical record by trained data abstractors at each participating hospital, with regular onsite audits aimed to ensure data reliability.

The analysis was restricted to patients who underwent hysterectomy for a benign gynecologic indication. Current Procedural Terminology codes were used to identify patients who underwent hysterectomy. Indications for hysterectomy were identified using the International Classification of Diseases, Ninth and Tenth Revision diagnosis codes. We excluded patients with associated procedures indicating the presence of malignant neoplasms or those who underwent hysterectomy with an indication of malignant neoplasm or uncertain behavioral diseases, as well as patients who underwent a pregnancy-related hysterectomy or an emergent hysterectomy, in which emergent was identified as patients with American Society of Anesthesiologists class 5; ventilator dependence; preoperative systemic inflammatory response syndrome, sepsis, or septic shock; preoperative open wound; preoperative acute renal failure; more than 4 transfusions with packed red blood cells 72 hours prior to surgery; or preoperative dialysis. An MIS was defined as total laparoscopic hysterectomy, laparoscopic supracervical hysterectomy, or laparoscopic-assisted vaginal hysterectomy. Details of study cohort selection are reported in eTables 1 to 5 in the Supplement.

Primary outcomes were 30-day major and minor complications. For evaluating the 30-day outcomes before and after the FDA statement, complications following hysterectomies performed from the first quarter of 2013 through the first quarter of 2014 (defined as pre-FDA) were compared with those from the fourth quarter of 2014 through the fourth quarter of 2015 (defined as post-FDA). Hysterectomies performed during the period of transition between the 2 FDA statements (the second through the third quarters of 2014) were excluded. Major complications were defined as unplanned intubation, wound disruption, ventilator use of more than 48 hours, sepsis, septic shock or systemic inflammatory response syndrome, pneumonia, deep incisional surgical site infection, acute renal failure, organ space surgical site infection, progressive renal insufficiency, pulmonary embolism, myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation, stroke or cerebrovascular accident with neurological deficit, deep vein thrombosis, or thrombophlebitis. Minor complications were defined as any urinary tract infection or superficial surgical site infection. Routes of hysterectomy for benign gynecologic indications were compared between pre- and post-FDA periods.

Two-tailed χ² tests and unpaired, 2-tailed t tests were used to compare complication rates, procedure distributions, and patient factors between pre-FDA and post-FDA periods among all women who underwent hysterectomy for benign gynecologic indications as well as in the subset of women with an indication of uterine fibroids. Multivariable logistic regression compared odds of complications before the FDA statement (combining all hysterectomies performed from the first quarter of 2013 through the first quarter of 2014) and after the FDA statement (combining all hysterectomies performed from the fourth quarter of 2014 through the fourth quarter of 2015) after controlling for age, race/ethnicity, body mass index (≥30 vs <30; calculated as the weight in kilograms divided by the height in meters squared), functional status, corticosteroid use for chronic conditions, preoperative platelet count (reference range of 150-450 × 10³/μL vs lower vs higher [to convert to ×10⁹ per liter, multiply by 1.0]), preoperative hematocrit (reference range of 34.9%-44.5% vs lower vs higher [to convert to proportion of 1.0, multiply by 0.01]), preoperative serum albumin level (reference range of 3.5-5.0 g/dL vs lower vs higher [to convert to grams per liter, multiply by 10]), hypertension requiring medications, preoperative serum creatinine level (reference range of 0.6-1.1 mg/dL vs lower vs higher [to convert to micromoles per liter, multiply by 88.4]), diabetes with oral agent or insulin administration, smoking status within 1 year, American Society of Anesthesiologists class (1-2 vs 3-4 [healthy to mild systemic disease and severe systemic disease to constant life threat, respectively]), and abdominal, extra-abdominal, upper-abdominal, or other associated procedures (yes vs no). Models, including that for all women with benign gynecologic indications, were also controlled for hysterectomy indication. To verify that increases in the complication rates observed among women with uterine fibroids differed compared with the increases among women with all other benign indications, we performed a secondary analysis among all women with benign conditions that included the interaction between indication (uterine fibroids vs all other benign indications) and period (pre-FDA vs post-FDA) in a multivariable logistic regression model of major complications, controlling for the aforementioned factors. Statistical analysis was performed using SAS, version 9.4 (SAS Institute Inc). All P values were 2-sided and were considered statistically significant at P < .05.

The characteristics of women who underwent hysterectomy for benign indications and for uterine fibroids in the pre- and post-FDA periods are reported in Table 1. In total, 75 487 women underwent hysterectomy for benign indications, of whom 32 186 (42.6%) were treated in the pre-FDA period and 43 301 (57.4%) were treated in the post-FDA period. The most common indications for hysterectomy were uterine fibroids (33.9%), irregular bleeding (24.6%), and pelvic organ prolapse or urinary incontinence (15.8%). The mean (SD) age was 47.8 (10.7) years, and the mean (SD) body mass index was 30.4 (7.4). Non-Hispanic white women comprised most (59.4%) of the population, followed by African American women (15.1%).

In a subgroup of 25 571 women (33.9%) treated for an indication of uterine fibroids, 10 903 (42.6%) underwent hysterectomy in the pre-FDA period and 14 668 (57.4%) in the post-FDA period. For this entire subgroup, the mean (SD) age was 46.9 (7.2) years, and mean (SD) body mass index was 30.6 (7.2). The proportion of African American women was almost double (27.0%) compared with the entire cohort of women treated for a benign indication (15.1%), whereas the proportion of Non-Hispanic white women was lower (44.9%).

Among 75 487 women who underwent hysterectomy for benign gynecologic indications, open abdominal hysterectomy increased from 24.0% during the pre-FDA period to 26.1% during the post-FDA period (Figure 1). The use of MIS decreased from 57.3% during the pre-FDA period to 55.4% during the post-FDA period, while vaginal hysterectomy decreased from 18.7% to 18.4%, respectively. The change in surgical route was significant (P < .001).

Among 25 571 women who underwent hysterectomy for uterine fibroids, the change in route of hysterectomy was more prominent (Figure 2). The rate of open abdominal hysterectomy increased from 37.2% during the pre-FDA period to 43.0% during the post-FDA period, the rate of MIS decreased from 56.1% during the pre-FDA period to 49.7% during the post-FDA period, and the rate of vaginal hysterectomy increased from 6.7% during the pre-FDA period to 7.3% during the post-FDA period (P < .001).

Among 75 487 women who underwent hysterectomy for benign gynecologic indications, 1639 (2.2%) experienced 30-day major complications, and 2752 (3.6%) experienced 30-day minor complications (Table 2). The rates of major complications were significantly higher in women undergoing open abdominal surgery (3.5%) compared with women undergoing MIS (1.7%) or vaginal hysterectomy (1.7%) (P < .001). The minor complication rates were higher in women undergoing vaginal hysterectomy (4.5%), followed by women undergoing open abdominal surgery (4.1%) and women undergoing MIS (3.2%) (P < .001).

Similarly, among women who underwent hysterectomy for uterine fibroids, 556 (2.2%) experienced 30-day major complications, and 777 (3.0%) experienced 30-day minor complications (Table 2). The rates of major complications were significantly higher in women undergoing open abdominal surgery (2.8%) compared with women undergoing MIS (1.8%) and women undergoing vaginal hysterectomy (1.8%) (P < .001). Minor complication rates were slightly higher in women undergoing open hysterectomy (3.3%) and women undergoing vaginal hysterectomy (3.2%) than in women undergoing MIS (2.8%), although the difference was not statistically significant (P = .07).

Among women who underwent hysterectomy for benign indications, 30-day major and minor complication rates remained relatively stable from the pre-FDA to post-FDA periods (Figure 3). Major complication rates were 2.1% during the pre-FDA period and 2.2% during the post-FDA period (adjusted odds ratio [OR], 1.03; 95% CI, 0.93-1.13; P = .62). Minor complications were 3.6% pre-FDA and 3.7% post-FDA (adjusted OR, 1.00; 95% CI, 0.92-1.08; P = .97).

By contrast, among women who underwent hysterectomy for uterine fibroids, 30-day major and minor complication rates significantly increased after the FDA statement (Figure 3). The major complication rate increased from 1.9% during the pre-FDA period to 2.4% during the post-FDA period (adjusted OR, 1.23; 95% CI, 1.04-1.47; P = .02), and the minor complication rate increased from 2.7% during the pre-FDA period to 3.3% during the post-FDA period (adjusted OR, 1.21; 95% CI, 1.04-1.40; P = .01). The full multivariable models for both major and minor complication rates among all women with benign gynecologic indications and also among women with uterine fibroids are reported in eTables 6 to 9 in the Supplement.

Comparisons of individual complication rates during the pre- and post-FDA periods among patients who underwent hysterectomy for benign gynecologic indications and for uterine fibroids are reported in eTable 10 in the Supplement. A model among all women with benign gynecologic indications, including a term for the interaction between indication (uterine fibroids vs nonuterine fibroids) and period (pre-FDA vs post-FDA period), showed that the association of increased complications in the period following the FDA statement was different between women with uterine fibroids and women with nonuterine fibroid indications (P = .009).

Our study, using a large nationwide data set, showed that major and minor complication rates among women who underwent hysterectomy for uterine fibroids increased following the FDA statement warning against the use of power morcellation. This 20% increase in the odds of major and minor complications could translate into a large number of additional complications among the 200 000 hysterectomies performed annually for uterine fibroids in the United States.^1,2

As one might expect, our results showed that open abdominal hysterectomy was associated with significantly more major and minor complications than MIS irrespective of the study period. Considering this superiority of MIS in terms of morbidity over open abdominal hysterectomy, the increased number of complications observed after the FDA statement was likely associated with the shift in the route of hysterectomy. Moreover, a decrease in the rate of open abdominal hysterectomies and an increase in the rate of MIS observed from 2003 to 2012^1,22 support our hypothesis that the inverse trend in the route of hysterectomies observed after 2013 was attributable to the FDA statement and not to other reasons.

Similar to our study, previous studies have shown that the FDA statement on morcellation resulted in a shift in approach to hysterectomy, with decreased use of MIS and increased use of open abdominal surgery.^13,17-19 However, to our knowledge, the only study that reported an increase in the number of complications after the FDA statement is a retrospective study by Harris et al¹⁸ evaluating postoperative complications of hysterectomy for benign indications in Michigan. They showed an increase in the rate of nontransfusion major complications from 2.2% to 2.8% and in the rate of 30-day readmissions from 3.4% to 4.2%; however, those study results were limited by a lack of generalizability because they examined only hospitals in the state of Michigan immediately after the FDA statement. By contrast, to date, no other study dealing with this important issue in gynecologic surgery has observed an increase in the number of complications after the FDA statement. In particular, to our knowledge, the only study evaluating the influence of the FDA statement on complications across the United States was by Wright et al,¹³ and it showed no changes in major postoperative complications. The reasons for the discrepancies between our study results and those of Wright et al¹³ could be their use of an administrative database (that did not differentiate between types of complications and was not designed to evaluate complications), their inclusion of patients with malignant neoplasms (which could be confounders), and their examination of only the immediate period after the FDA statement. Moreover, a study recently published by Ottarsdottir et al¹⁷ examining patients with indication of uterine fibroids who underwent hysterectomy from 2013 to 2015 at Brigham and Women’s Hospital showed an increase in the rate of open surgery from 19% to 29% and a decrease in the rate of laparoscopy from 71.3% to 65.3%. However, except for an increase of blood loss in the later years, other perioperative outcomes remained similar. As recognized by those authors, the generalizability of the study was limited by its single-institution design, and the study was underpowered to detect a small difference.

The large sample size in the present study allowed for the detection of a significant 0.5% increase in the rate of major complications among women with an indication of uterine fibroids. Although this increase may appear small, it is larger than the 0.2% risk of unsuspected leiomyosarcoma estimated by the FDA, which was the motivation for the FDA statement and the impetus for using open techniques instead of morcellation. Furthermore, given the large number of women who undergo hysterectomy for uterine fibroids in the United States, a small increase in the rate of complications would translate to a large increase in the number of individual women who experience a major complication. Moreover, a decision-tree analysis including the risk of death from leiomyosarcoma and the differences in morbidity between laparoscopy and laparotomy estimated that the overall mortality would increase if laparoscopic cases were converted to laparotomy cases.²³

The strengths of the present study include a large sample size from a prospective multicenter database, including centers from across the United States. Moreover, standardized protocols and central auditing ensured reliability of data collection across the participating institutions.

The present study has several potential limitations. First, its retrospective design might have introduced bias inherent with such a data set. Second, although the ACS NSQIP is accurate in capturing 30-day complications, it does not follow-up with patients beyond 30 days; therefore, we were unable to assess long-term complications. Third, owing to the lack of data on the use of power morcellation and on the prevalence of abnormal uterine pathology during the pre- and post-FDA periods among women who underwent MIS, we were unable to assess whether the shift in the route of hysterectomy consequent to the FDA statement was associated with a reduction in the prevalence of spreading unexpected malignant neoplasms. However, Wright et al¹³ showed that the prevalence of abnormal uterine pathology among woman who underwent MIS with electric power morcellation after the post-FDA period did not change compared with that in the pre-FDA period. Fourth, the increase in the number of participating hospitals between the pre- and post-FDA periods (435 hospitals in 2013, 517 hospitals in 2014, and 603 hospitals in 2015) and the change in the sampling method that occurred for hysterectomy procedures during the study period may have affected our results; however, this limitation cannot be overcome with the available data. Finally, although we could speculate that an increase in the number of complications could have resulted in increased costs and delayed return to work, the lack of data on these variables prevented us from addressing these aspects.

The FDA statement warning against the use of power morcellation was associated with an increased risk in the number of major and minor complications among women undergoing hysterectomy for uterine fibroids. This increased risk corresponded with an abrupt shift in the route of hysterectomy; we believe that many hospitals and payers reduced the use of both forms of morcellation (power and hand), which decreased options for MIS. Although caution is required to avoid morcellation of unexpected uterine malignant neoplasms, our results should be considered by women and clinicians during the process of shared decision making and by medical societies and regulatory bodies when issuing safety communications. Until new studies discover either preoperative techniques to identify patients with sarcomas or alternative intraoperative technology, such as contained power morcellation to remove a large uterus from the abdominal cavity without disseminating tissue, the decision regarding the surgical approach should be guided by the values and informed preferences of the patient. Further research is warranted focusing on the refinement of these alternative techniques of uterine morcellation and on the identification of women who would benefit from MIS.

Accepted for Publication: January 14, 2018.

Corresponding Author: Shannon K. Laughlin-Tommaso, MD, Department of Obstetrics and Gynecology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (laughlintommaso.shannon@mayo.edu).

Published Online: April 11, 2018. doi:10.1001/jamasurg.2018.0141

Author Contributions: Drs Multinu and Laughlin-Tommaso had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Multinu, Casarin, Angioni, Mariani, Habermann, Laughlin-Tommaso.

Acquisition, analysis, or interpretation of data: Multinu, Hanson, Mariani, Habermann.

Drafting of the manuscript: Multinu, Mariani, Laughlin-Tommaso.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Multinu, Hanson, Habermann.

Administrative, technical, or material support: Casarin.

Study Supervision: Multinu, Casarin, Mariani, Habermann, Laughlin-Tommaso.

Conflict of Interest Disclosures: None reported.

Funding/Support: This publication was made possible by a Clinical Translation Science Award (UL1 TR000135) to the Mayo Clinic Graduate School of Biomedical Sciences from the National Center for Advancing Translational Sciences, a component of the National Institutes of Health. Dr Casarin is supported by the University of Insubria, Varese, Italy, and by Fondo Miglierina, Provincia di Varese, Italy.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The contents are solely the responsibility of the authors and do not necessarily represent the official view of the National Institutes of Health. The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) and the hospitals participating in the ACS NSQIP have not verified and are not responsible for the statistical validity of the data analysis or the derived conclusions.