Getting square pegs out through round holes: A survey of Australian and New Zealand Gynaecologists regarding specimen retrieval.

The Centre for Advanced Reproductive Endosurgery supplied $700AUD to enable the authors to complete the RANZCOG email survey process.

The survey was emailed to 2300 RANZCOG Fellows; of these, 438 (19.04%) RANZCOG Fellows began the survey, and 258 (11.22%) completed the entire survey. Analysis was conducted on the responses of the Fellows who completed the entire survey. The demographic characteristics of respondents are listed in Table  1 . Demographic characteristics of the study population ASG‐RA, Australian Statistical Geography Standard – Remoteness Area. Basic laparoscopic surgeon = confident and competent to perform (for example) ovarian cyst aspiration, excision of stage I–II endometriosis. Intermediate laparoscopic surgeon = confident and competent to perform (for example) ovarian cystectomy, oophorectomy, laparoscopic‐assisted vaginal hysterectomy, excision of stage III endometriosis. Advanced laparoscopic surgeon = confident and competent to perform (for example) total laparoscopic hysterectomy, laparoscopic myomectomy, sacrocolpopexy, excision of stage IV endometriosis. There were 208 (80.62%) respondents who reported currently practising in Australia, and 47 (18.22%) in New Zealand. The sample was broadly representative of all 2300 RANZCOG Fellows: the proportions of respondents classified according to age, gender, location, remoteness, and public vs private practice were similar to the complete sample of Fellows. 13 In comparison to RANZCOG Fellows generally, respondents were less likely to have sub‐specialised in ultrasound, maternal fetal medicine, or fertility, and more likely to have sub‐specialised in gynae‐oncology. Respondents' self‐assessed surgical mastery (as per RANZCOG's levels of laparoscopic scope of practice) was: basic for 20 (7.75%) respondents, intermediate for 101 (39.15%), and advanced for 137 (53.10%) respondents. 14 There were 27 (10.47%) respondents who had completed or were currently completing a two‐year Fellowship in advanced laparoscopic gynaecology through the Australasian Gynaecological Endoscopy and Surgery society (AGES). Respondents reported performing an average of 16.98 operations annually that required morcellation, with 145 (56.20%) performing fewer than ten such cases annually, and ten (3.88%) performing more than 50 procedures. Respondents were asked to clarify which specimen retrieval method they currently utilise, when performing an operation that requires morcellation (eg laparoscopic sub‐total hysterectomy, laparoscopic myomectomy). The options provided included: performing an open (abdominal) operation from the start of the operation; contained/uncontained power morcellation; contained/uncontained manual morcellation; conversion to laparotomy intra‐operatively (having started the operation laparoscopically); and colpotomy with some combination of contained/uncontained power or manual morcellation. Respondents were asked the frequency with which they employed each specimen retrieval method. The average frequencies (in order) were: performing an open operation from the get‐go (35.92%); contained manual morcellation (33.65%); converting to laparotomy intra‐operatively (12.29%); contained power morcellation (8.04%); uncontained manual morcellation (4.66%); colpotomy plus any method of morcellation (3.62%); and uncontained power morcellation (1.82%). Adding together 35.92% of cases that have a laparotomy from start of the case, to the 12.29% who convert from a laparoscopic to a laparotomic approach intra‐operatively, 48.21% of patients having operations requiring morcellation undergo a laparotomy. One hundred sixty five (63.95%) respondents worked as a gynaecology consultant prior to 2014. Figure  1 summarises their use of uncontained power morcellation before and after the 2014 FDA warnings, in addition to the percentage change. The use of uncontained power morcellation prior to 2014 and since 2015, among respondents who worked as gynaecology consultants prior to 2014. For this question, respondents were asked: ‘Of the operations you perform that required morcellation: for what percentage of cases did you employ uncontained power morcellation, prior to 2014 versus since 2015?’ Thirty one (18.79%) respondents reported using uncontained power morcellation for every case requiring morcellation prior to 2014: this proportion fell by 80.65% to six (3.64%) respondents following 2015. At the other end of the spectrum, 100 (60.6%) respondents reported never using uncontained power morcellation prior to 2014: this proportion increased by 36.00% to 136 (82.42%) since 2015. This represents a clear trend away from uncontained power morcellation since the 2014 FDA warnings. As detailed in Table  2 , 244 (94.57%) of all 258 respondents reported using uncontained power morcellation in 0% of their operations requiring morcellation. Four (1.55%) use uncontained power morcellation in 1–10% of cases, and six (2.33%) in 11–20% of cases. No respondents used uncontained power morcellation in more than 80% of their cases requiring morcellation. Use of uncontained power morcellation For this question, respondents were asked: ‘When you perform an operation that requires morcellation, how often do you employ uncontained power morcellation?’ Respondents were asked to select reasons why they were disinclined to perform uncontained power morcellation; they could choose any number of options provided. The responses (listed in order) were: awareness of 2014 FDA warnings (104, 40.31%); risk of adverse outcomes (87, 33.72%); recommendations from speciality colleges such as the American College of Obstetricians and Gynecologists (ACOG), RANZCOG or the Royal College of Obstetricians and Gynaecologists (70, 27.13%); recommendations from sub‐specialty interest groups such as the American Association of Gynecologic Laparoscopists (AAGL), AGES, or the British Society for Gynaecological Endoscopy (BSGE) (62, 24.03%); inability to source a power morcellator (39, 15.12%); word‐of‐mouth (34, 13.18%); personal experience of adverse outcomes (nine, 3.49%); and patient request (eight, 3.10%). There were no significant differences in specimen retrieval practices between sub‐groups of respondents based on: age, years of RANZCOG Fellowship, country of training, and public vs private practice. As can be seen in Table  3 , there was a significant correlation between respondents' self‐assessed surgical mastery, and the frequency with which they employed a laparotomic approach from the beginning of an operation requiring morcellation. Surgical mastery vs open surgical approach from the start of an operation requiring morcellation The P ‐value was <0.001, and Cramer's V correlation coefficient 0.352 (ie strong association). For this question, respondents were asked: ‘Do you ever employ an open approach (from the start of an operation requiring morcellation)?’ Respondents who were currently completing, or had previously completed, an advanced laparoscopic gynaecology Fellowship were more likely to use uncontained power morcellation compared to other participants ( P ‐value <0.001, Phi correlation coefficient 0.211, weak association). They were less likely to use an open approach from the beginning of an operation ( P ‐value <0.01, Phi correlation coefficient 0.164, weak association), and more likely to use contained manual morcellation ( P ‐value <0.05, Phi correlation coefficient 0.124, weak association). As can be seen in Table  4 , survey participants who perform more operations annually requiring morcellation were significantly less likely to employ an open approach ( P ‐value <0.001, Cramer's V correlation coefficient 0.328, strong association). Number of operations performed per annum requiring morcellation vs open surgical approach from the start of an operation requiring morcellation

RANZCOG Fellows' use of uncontained power morcellation has significantly decreased since the 2014 FDA warnings against power morcellation. Key drivers of this change include an awareness of the FDA warnings, and recommendations of specialty colleges and interest groups (such as AGES). Similar research performed by clinicians in North America and Europe has shown similar results: Lum et al performed a similar survey among AAGL members in 2015, and concluded that the FDA warnings had led to a statistically significant decrease in power morcellation. 8 In contrast, Sankaran et al 10 performed a survey of BSGE members, and found that 85% of respondents retrieved specimens using power morcellation. Sampling clinicians with an interest in advanced laparoscopic gynaecology, and the potential selection bias that results, may explain the discrepancy between their data and ours. The likelihood of a presumed ‘fibroid’ actually being a LMS is imprecise and unreliable, ranging from one in 350 to one in 8300. 4 , 6 Risk factors for LMS include advancing age, previous pelvic radiation, and African American ethnicity. 15 There are no definitive tests to confirm the diagnosis of LMS pre‐operatively. Given the difficulties estimating a patient's risk of LMS, selecting appropriate patients for whom morcellation is appropriate can be challenging. Iatrogenic dissemination of malignant tissue during uncontained morcellation has serious clinical implications, including a worse prognosis. 16 Park et al undertook a retrospective study comparing the outcomes of patients with LMS, who had or had not undergone morcellation. There was increased disseminated disease (44% vs 12%), and decreased five‐year disease‐free survival (40% vs 65%), and five‐year overall survival (46% vs 73%) in patients in whom tissue morcellation had been utilised. 16 Published in 2020, FDA guidance recommends that power morcellation should not be used in patients over 50 years old, or in patients for whom specimens can be retrieved en bloc through a colpotomy or mini‐laparotomy. 17 Both AAGL and ACOG have published position statements regarding power morcellation. 18 , 19 Both societies support power morcellation in patients who are appropriately selected, and informed of the potential risks. Despite these learned societies' guarded support of power morcellation, many clinicians and hospitals have interpreted the 2014 FDA warnings as a complete ban on power morcellation. Contained morcellation techniques represent creative and potentially safer methods of specimen retrieval. Contained morcellation aims to decrease both the dissemination of occult malignancies and parasitic leiomyomas, and internal organ injury. In 2020, the FDA recommended the use of a containment bag whenever morcellation (whether manual or power) is performed. 17 This guidance acknowledged that containment bags cannot prevent the spread of tissue that occurs when manoeuvring the specimen into the bag, or malignant spread that has occurred pre‐operatively. It also acknowledged that the recommendations were based on ‘bench and animal testing’ of containment bags. Health Canada also recommends that containment bags be used whenever morcellation is performed. 20 There is no evidence that using a containment bag changes the natural course of the disease and/or improves clinical outcomes. Furthermore, potential complications inherent to contained morcellation have not yet been assessed. While there is limited (if any) high‐grade evidence supporting the use of containment bags, intuition and biological plausibility support their use. There is emerging evidence that the risks of laparotomy outweigh the risks of inadvertent morcellation of an undiagnosed LMS. 21 Compared to laparoscopic hysterectomy, abdominal hysterectomy is associated with higher rates of venous thromboembolism, blood transfusion, bowel injury, myocardial infarction, prolonged length of stay, and death. 19 AAGL estimates that converting all hysterectomies that would otherwise use the power morcellator to laparotomy would result in an additional 17 women dying annually in the US, and a substantial increase in the aforementioned morbidities. 18 Contained manual morcellation may represent current best practice, in that it optimises the benefits of laparoscopy (over laparotomy), while mitigating the risks inherent to both uncontained and power morcellation. Our respondents lacking the requisite morcellation skillset is one of several possible drivers behind their use of laparotomy during operations that necessitate specimen retrieval. Hence, any move away from uncontained power morcellation must be complemented by a stronger evidence‐base and training skillset for alternative specimen retrieval techniques, such as contained manual morcellation. Colpotomy was the least preferred method of specimen retrieval among our survey participants. This route was first described more than 100 years ago (Kelly 1896); our survey has demonstrated that colpotomy is not a popular method of specimen retrieval in the Antipodes. However, this method may be used more frequently in future, with increasing adoption of vNOTES procedures. A response rate of 11.22% is in keeping with many medical questionnaires. 22 Measures were taken to improve the response rate, such as keeping the survey concise and easy to understand; an email reminder was sent. There may have been an (understandable) lack of interest in our survey among Fellows who only perform obstetrics; this may explain why our response rate was lower than previous studies which only invited laparoscopic gynaecologists. 9 , 10 This survey, like many others, is limited by the response rate, leading to potential non‐response bias. Nevertheless, we believe these data are important, as they provide valuable insights into specimen retrieval practices among Antipodean gynaecologists. Participants' responses about the number of procedures they perform annually may be prone to recall bias: respondents may have over‐ or under‐estimated the number of procedures performed, and/or the proportion of specimens retrieved by various methods. In addition, an awareness that uncontained power morcellation is no longer de rigour may have led respondents to down‐play this element of their practice. Nevertheless, the results are of interest to the gynaecological community, and should prompt gynaecologists to reflect on their current practice. In summary, our survey of RANZCOG Fellows has found a clear trend away from uncontained power morcellation since the 2014 FDA warnings. There were 31 (18.79%) respondents who reported using uncontained power morcellation for every case requiring morcellation prior to 2014: this proportion fell by 80.65% to six (3.64%) respondents, when reporting their use of uncontained power morcellation since 2015. The most common reason for this change was an awareness of the 2014 FDA warnings against uncontained power morcellation.

While laparoscopic surgery has become the gold standard for many gynaecological procedures, the retrieval of (sometimes very large) specimens through small incisions presents several challenges. Specimen retrieval methods available to clinicians include: employing an open (abdominal) approach from the beginning of the operation; power morcellation, using an electromechanical morcellator; and manual morcellation through either an extended abdominal incision or a colpotomy, a containment bag may or may not be utilised when performing power or manual morcella. Optimal specimen retrieval has been a subject of much debate over the last decade. Ideally, specimen retrieval should optimise laparoscopic surgery's many well‐documented benefits (eg smaller scars, expedited recovery and inpatient stay, and decreased postoperative pain and venous thromboembolism risk), while minimising the risks inherent to the morcellation process itself (eg disseminated peritoneal leiomyomatosis, spread of malignant tissue, internal organ injury). Morcellation is the division of large specimens into smaller fragments, to permit removal from the peritoneal cavity. 1 Morcellation is used during minimally invasive gynaecology to retrieve specimens (eg leiomyomas, uterus) through relatively small (eg 10–30 mm) surgical incisions. Power morcellation was described by Semm in 1991, 2 and a laparoscopic morcellator gained Food and Drug Administration (FDA) approval in 1995. 3 Following a landmark case in 2013, the FDA issued warnings against power morcellation, citing the potential risks of intra‐abdominal spread of previously unsuspected uterine leiomyosarcoma (LMS). 4 , 5 The FDA notification precipitated a debate regarding the incidence of LMS, and the pros and cons of morcellation. 3 Debates (both in public and among the medical community) over the last decade regarding morcellation are remarkable in gynaecological surgery's history; these discussions changed the landscape around specimen retrieval forevermore. 7 Several research groups have surveyed clinicians to assess changes to morcellation practices following the 2014 FDA warnings 8 , 9 , 10 , 11 , 12 ours is, to our knowledge, the first such study performed outside North America, Europe, and/or the United Kingdom. The primary objective of our survey was to ascertain the frequencies with which different specimen retrieval techniques are utilised by Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) Fellows. Secondary objectives included assessing: changes in specimen retrieval practices since the 2014 FDA warnings; clinicians' rationales for utilising various specimen retrieval methods; and any significant differences based on respondents' demographic variables. We hypothesised that we would find a significant decrease in the use of power morcellation, and a significant increase in the use of containment systems, when comparing practice prior to, and after, 2014.

Ethics approval was gained from the Northern Sydney Local Health District Human Research and Ethics Committee; application number 2020/ ETH01534 . Survey questions were developed based on previous published research on the topic, and were pilot‐tested by experts in minimally invasive gynaecology. The survey was approved by the RANZCOG continuing professional development committee, who also invited all 2300 RANZCOG Fellows to participate. An initial email was sent in April 2021, plus a reminder email two weeks later. No incentives were offered. The structured survey was conducted through a SurveyMonkey online survey ( www.surveymonkey.com ); there were both multiple choice and free text response options for each question. Internet protocol addresses were the only identifying information attached to the participants' responses. The information below was collected from respondents: demographic data such as: age; gender; country where they completed obstetric and gynaecology training; location of practice; whether or not sub‐specialist training is being undertaken/had been completed; proportion of public vs private practice scope of practice, such as: self‐assessed laparoscopic mastery; number of cases performed annually necessitating morcellation; utilisation rates of specimen retrieval methods surgeons' motivations when choosing between specimen retrieval methods respondents who were gynaecology consultants prior to 2014 were asked about their use of uncontained power morcellation prior to and since 2014, and their rationale for any change in practice demographic data such as: age; gender; country where they completed obstetric and gynaecology training; location of practice; whether or not sub‐specialist training is being undertaken/had been completed; proportion of public vs private practice scope of practice, such as: self‐assessed laparoscopic mastery; number of cases performed annually necessitating morcellation; utilisation rates of specimen retrieval methods surgeons' motivations when choosing between specimen retrieval methods respondents who were gynaecology consultants prior to 2014 were asked about their use of uncontained power morcellation prior to and since 2014, and their rationale for any change in practice After allowing three months for respondents to complete the survey, data analysis commenced. As this was a descriptive study, data were reported descriptively using percentages. All analyses were performed using a combination of Microsoft Excel™ and SPSS version 25 (IBM, Armonk, NY, USA).