Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
More specialised gynaecological surgeons needed
- David J R Hutchon (28 June 2005)
-
Progression to Vaginal Hysterectomy a personal Odyssey
- Peter R. Pedlow, Hitchin, Herts, SG4 9QZ (29 June 2005)
-
Methods of hysterectomy: should women have a say?
- Vikki A Entwistle, Siladitya Bhattacharya, Graeme MacLennan, Zoe Skea, and Brian Williams (29 June 2005)
-
Methods of hysterectomy
- Adam Magos, Lynne Chapman, Vasiliki Varela, Pangiotis Papalampros, Pietro Gambadauro, and Ramesan Navaratnarajah (7 July 2005)
-
Interesting Perspectives on Level 1 Evidence for Surgical Approach to Hysterectomy
- Neil P Johnson (28 July 2005)
-
Oophorectomy at vaginal hysterectomy ?
- David JR Hutchon (31 July 2005)
-
Re: Oophorectomy at vaginal hysterectomy ?
- Adam Magos (1 August 2005)
-
Methods of hysterectomy and ureteric injury.
- Roger A. McMaster-Fay (8 January 2006)
-
Surgical Approach to Hysterectomy and Ureteric Injury
- Neil P Johnson (5 March 2006)
|
|
|||
|
David J R Hutchon, Locum Cons Gynaecologist on sabatical leave Grey Base Hospital, Greymouth, New Zealand
Send response to journal:
|
This review and metanalysis by Johnson et al is most welcome. Hopefully the conclusions will influence the Luddite gynaecologist who only offers hysterectomy by the abdominal route (unless there is prolapse) to seek further training in vaginal and/or laparoscopic surgery. Such a surgeon is likely to be highly skilled in abdominal hysterectomy but such skill is not immediately transferable to vaginal or laparoscopic surgery. However, what is the evidence that laparoscopic surgery requires a higher degree of surgical skill than abdominal or vaginal hysterectomy? Indeed the higher bladder and ureter injury reported with laparoscopic surgery suggests that laparoscopic surgeons could have less surgical skill or at least poorer clinical judgment. Laparoscopic surgery requires a new generation of instruments and theatre facilities different from those required for the traditional instruments of abdominal or vaginal hysterectomy. Just as a highly skilled rugby player does not expect to be inevitably skilful at cricket, a highly skilled laparoscopic should not expect to be skilled at abdominal surgery. The laparoscopic surgeon however does need to be competent at abdominal surgery as the default approach in case of problems is abdominal. With the reduction in the need for hysterectomy the opportunity to gain experience in any form of hysterectomy is reducing and at now needs to be concentrated into the hands of fewer surgeons. The ntrainees for these positions need to gain experience in abdominal and all stages of vaginal/laparoscopic hysterectomy. The studies of hysterectomy do not identify how many patients had a diagnostic laparoscopy in the previous months. The information gained at laparoscopy may well influence which method of hysterectomy is offered and if vaginal surgery is planned should be classified as a staged LAVH (stage 0). Personally I would much rather be aware of any problems prior to vaginal hysterectomy and find it difficult to understand why such a safe procedure as diagnostic laparoscopy could have an adverse effect. With large fibroids the view from above can give encouragement to continue the vaginal surgery. With a scarred utero-vesical peritoneum after repeated Caesarean Section opening into the peritoneum can be difficult, and a few from above helps to confirm the progress. Neither of these approaches requires more than normal laparoscopy skills but in my own experience makes vaginal hysterectomy more likely to be completed safely. The title of the paper by Robertson et al was unfortunate and is often used to discredit the middle of the road approach of LAVH (Laparoscopic assisted vaginal hysterectomy) stage 0 and 1. The conclusion of the authors however that laparoscopic surgery should be converted to vaginal as soon as this is feasible does not in itself show that laparoscopic assistance for hysterectomy is a waste of anything. In a busy gynaecological practice in the UK I have done no more than a handful of abdominal hysterectomies in the past few years. If in doubt, I am quite comfortable at carrying out a stage 0 LAVH. I am currently on sabatical leave working in New Zealand Competing interests: None declared |
|||
|
|
|||
|
Peter R. Pedlow, Consultant Obs.and Gyn. (retired) Pinehill Hospital, Benslow Lane,, Hitchin, Herts, SG4 9QZ
Send response to journal:
|
On completing my training in Obstetrics and Gynaecology in the early 1960's and becoming a Consultant at two District General Hospitals in North Hertfordshire it seemed self evident to me that if the uterus could be removed safely through the natural orifice of the vagina (as then advocated by a very few gynaecologists in this country), rather than via the unnecessary abdominal incision commonly used, it must be the preferable approach for both surgeon and patient. In those days it was possible to train oneself and gradually increase one's experience without putting the patients at risk - after all it was perfectly feaseable to convert to the abdominal approach in the rare case where unexpected problems were encountered during the procedure. However,postoperative vaginal vault infections were unacceptably commoner after the vaginal approach, a recognised fact which discouraged many other surgeons from operating vaginally. These infections resolved themselves by discharging through the vagina after a few days but were unpleasant for all concerned. I was convinced that the infection started in small vault haeamatomas, an excellent culture medium, so I not only scrupulously uprated my haemostasis but also administered an intramuscular injection of Ampicillin just prior to operating and continued for a few days, despite the fact that the use of "prophylactic" antibiotics was generally frowned upon in those days (but has now become common practice). Hence any haematoma would contain antibiotic and was therefore not a good culture medium. Eureka! postoperative infections virtually disappeared. Before retiring I audited 451 unselected consecutive hysterectomies carried out by myself during the ten year period from 1984 to 1993. Of these 78% were vaginal (in several cases minor ovarian was also dealt with),and 22% abdominaldue to anticipated gross pelvic pathology. There were 331 of these who did not have any utero-vaginal prolapse and 71% of this sub-group had their hysterectomy completed entirely by the vaginal route. Three hysterectomies were commenced vaginally but completed abdominally due to unforseen intra-operative problems. Fortunately there were virtually no post-operative complications, and most were pleased to go home in about 4 days. It was amusing to me, and predictable, that many of the Laparoscopic Assisted Vaginal Hysterectomists who embraced the technique after the advent of laparoscopic surgery in the 1980's soon to realise that the laparoscope could safely be discarded and the operation! But there are still too many hysterectomies carried out through abdominal incisions or with laparoscopic punctures in this country, leaving scars to remind the woman that she has "lost her womb". But I'm pleased to see that the message is now getting across that removing a uterus through the natural orifice is not a daunting or risky procedureif carried out with patience and care. After all the E.N.T. surgeons do not remove benign tonsils through the side of the neck! Competing interests: None declared |
|||
|
|
|||
|
Vikki A Entwistle, Reader Health Services Research Unit, University of Aberdeen, Foresterhill, AB25 2ZD, Siladitya Bhattacharya, Graeme MacLennan, Zoe Skea, and Brian Williams
Send response to journal:
|
The review shows that vaginal hysterectomy outperforms (open) abdominal hysterectomy on all outcomes for which there is evidence from randomised controlled trials, and laparoscopic hysterectomy outperforms abdominal hysterectomy on all except injuries to the bladder or ureter.(1) In practice, abdominal hysterectomy dominates, so the editorial reasonably advocates more training in vaginal surgery and the development of evidence -based guidelines about choice of surgical method.(2) Questions must also be asked about how women should be informed about, and enabled to influence, the selection of a method for their hysterectomy. We recently found that 25% of women surveyed prior to a hospital admission for hysterectomy had not been told what method they would have.(3) Fewer than half had been told about the advantages or disadvantages of different methods. Women knew, or learned as they discussed their forthcoming hysterectomy with friends, that there are different methods. Those women whose gynaecologists had told them that vaginal or keyhole surgery was not feasible because of their particular pathology (for example large fibroids) apparently accepted this. But some women whose gynaecologists did not discuss alternative methods of hysterectomy wondered whether the selection was made in their interests or their gynaecologist’s. None expressed awareness that some gynaecologists only perform certain methods of hysterectomy. Decisions between hysterectomy methods may be preference sensitive. Although the review team consider laparoscopic surgery preferable to abdominal surgery,(1) some women may be more concerned to avoid higher risks of bladder or ureter injury than to obtain other benefits associated with the laparoscopic method. Especially in the context of renewed calls for greater patient choice about type of treatment,(4) the nature and acceptability of constraints on individual choice between hysterectomy methods need careful consideration, as do the desirability and feasibility of revising consultation and referral procedures to give women more say about their surgical procedures. 1. Johnson N, Barlow D, Lethaby A, Tavender E, Curr L, Garry R. Methods of hysterectomy: systematic review and meta-analysis of randomised controlled trials. BMJ 2005; 330: 1478-81. 2. Edozien LC. Hysterectomy for benign conditions. BMJ 2005; 330: 1457-1458. 3. Entwistle V, Williams B, Skea Z, Maclennan G, Bhattacharya S. Which surgical decisions should women participate in and how? Reflections on women’s recollections of discussions about different types of hysterectomy. Soc Sci Med (in press) 4. Coombes R. Gap between NHS policy makers and doctors is “enormous” (news). BMJ 2005; 330: 1468. Competing interests: None declared |
|||
|
|
|||
|
Adam Magos, Consultant Gynaecologist University Department of Obstetrics and Gynaecology, Royal Free Hospital, London NW3 2QG, UK, Lynne Chapman, Vasiliki Varela, Pangiotis Papalampros, Pietro Gambadauro, and Ramesan Navaratnarajah
Send response to journal:
|
EDITOR – We read the systematic review by Johnson et al on methods of hysterectomy with interest. 1 There has long been a debate amongst gynaecologists about which is the best route of surgery, but it is only with the advent of laparoscopic hysterectomy and its variants that proper scientific methods have been applied to compare different techniques. As the review makes clear, the medical literature is now extensive, and sufficient to answer most issues. We agree with the conclusion that vaginal hysterectomy should be the default route of surgery unless there are obvious contra-indications. Amongst all the careful science, it is a pity that the authors have allowed a statement to creep into their otherwise excellent article which is not backed up by fact. After stating that “vaginal hysterectomy is preferable to abdominal hysterectomy, provided it can be done safely”, they add “a laparoscopic approach may be appropriate if an oophorectomy is needed”, the implication being that oophorectomy is better done laparoscopically. This is simply not the case. Evidence shows (a) the adnexa are anatomically easily accessible vaginally in most cases, (b) vaginal oophorectomy or salpingo-oohorectomy is feasible in most women undergoing vaginal hysterectomy, and (c) vaginal adnexectomy is safe. 2-4 As the majority of women undergoing hysterectomy do not have adnexal pathology, the majority can be managed by vaginal surgery alone without the need for an unnecessary laparoscopy, with its inherent risks of bowel or vascular injury. 5 Randomized trials maybe relatively scant in this area, but that is no reason to ignore what we already know. The need for oophorectomy should not be considered a contra-indication to vaginal hysterectomy. Dentists remove molar teeth and ENT surgeons carry out tonsillectomy through the mouth, and neither would dream of operating via the cheeks or the neck! Gynaecologists too should be encouraged to utilise the vagina, the obvious route of access not only to the uterus but the ovaries at hysterectomy. If we do not, we will deny a significant proportion of our patients the benefits of this route of surgery. 6 1.Johnson N, Barlow D, Lethaby A, Tavender E, Curr L, Garry R. Methods of hysterectomy: systematic review and meta-analysis of randomised controlled Trials. BMJ 2005; 330: 478-81. 2.Kovac SR, Cruikshank SH. Guidelines to determine the route of oophorectomy with hysterectomy. Am J Obstet Gynecol 1996; 175: 1483-8. 3.Sheth SS. The place of oophorectomy at vaginal hysterectomy. Br J Obstet Gynaecol 1991; 98: 662-6. 4.Davies A, O'Connor H, Magos A. A prospective study to evaluate oophorectomy at the time of vaginal hysterectomy. Br J Obstet Gynaecol 1996; 103: 915-20. 5.Magrina JF. Complications of laparoscopic surgery. Clin Obset Gynecol 2002; 45: 469-80. 6.Davies A, Vizza E, Bournas N, O'Connor H, Magos A. How to increase the proportion of hysterectomies performed vaginally. Am J Obstet Gynecol 1998; 179: 1008-12. Competing interests: None declared |
|||
|
|
|||
|
Neil P Johnson, Associate Professor & Consultant University of Auckland & National Women's Health @ Auckland Hospital, New Zealand
Send response to journal:
|
I have welcomed the flurry of rapid responses to our systematic review of randomised trials comparing surgical approaches to hysterectomy (Hutchon 28 June 2005; Pedlow 29 June 2005; Entwhistle et al 29 June 2005; Magos et al 7 July 2005) – it is pleasing to see that at least some of my colleagues internationally do care about level 1 evidence that should be of interest to every practicing gynaecologist! David Hutchon highlights the increased incidence of urinary tract injury accompanying laparoscopic hysterectomy compared to abdominal hysterectomy. It is currently unclear whether this relates to poor clinical judgement amongst surgeons performing laparoscopic hysterectomies, as David suggests, is an inherent risk with laparoscopic approaches to hysterectomy, or is simply a function of the learning curve. Extensive case series by individual surgeons are accumulating and these suggest the learning curve with laparoscopic hysterectomy and its associated injuries may run into many hundreds of cases. Peter Pedlow is to be congratulated – his personal odyssey of progression to vaginal hysterectomy shows, in the face of many institutions (including my own in Auckland) struggling to change their majority of hysterectomies for benign disease being performed abdominally, that with a commitment to vaginal hysterectomy, rates in excess of 70% vaginal hysterectomy can be achieved for non-prolapse hysterectomy with benign disease. Like Adam Magos et al, Peter promotes the concept of ovarian surgery vaginally. David and Adam may wish to consider one another’s diametrically opposed arguments. At one pole, David advances an increasingly popular argument amongst surgeons committed to laparoscopic approaches, extolling the virtues of “such a safe procedure as diagnostic laparoscopy” being a useful adjunct that in his hands “makes vaginal hysterectomy more likely to be completed safely”. At the opposite pole, Adam remains an ardent vaginal oophorectomist. These expressed views, neither of which are supported nor refuted by the best available randomized trial evidence in our review, demonstrate that individuals’ prior beliefs (dare I say biases, or is this more to do with personal experience and expertise with vaginal and laparoscopic surgery?) strongly colour how we interpret available evidence. Furthermore these differences of opinion highlight the challenges facing us when we systematically review randomised trials of surgical interventions using methodology better suited to comparisons of medical interventions. How generalisable are such results and how are they squared with individual surgeons’ expertise with various surgical approaches? Drug A versus drug B will inevitably have little of the clinical and statistical heterogeneity we have seen in the systematic review of surgical approaches to hysterectomy, meaning that, even though we had 27 randomised trials in our review, it still suffers to some extent from limited generalisability. Adam’s group and others have had extensive vaginal surgical experience and high success rates with removing ovaries vaginally. Even in expert hands, however, there is an appreciable percentage of ovaries that cannot be safely removed vaginally. Surgeons surely need to operate within the limits of their expertise and many find oophorectomy difficult with the vaginal approach. Why struggle with a difficult oophorectomy vaginally when laparoscopic hysterectomy as an adjunct to vaginal hysterectomy is a procedure with a low complication rate. Frankly, I agree with Adam that “the need for oophorectomy should not be considered a contra-indication to vaginal hysterectomy”, but may I underscore heavily the statement that “a laparoscopic approach may be appropriate if an oophorectomy is needed” (or if surgeons like David wish to view Caesarean adhesions, or other pelvic pathology such as fibroids, with the option of dealing with the pathology laparoscopically or vaginally). An RCT comparing BSO vaginally versus laparoscopically in conjunction with a vaginal hysterectomy is the only way to know for sure, but the sample size would need to be powered carefully. The data of Vikki Entwhistle et al concerning women’s views on the surgical approach to their own hysterectomy are extremely powerful. They re-emphasise the imperative of clear communication of the rationale underpinning the choice of surgical approach to hysterectomy with our patients. Vikki is outlining what is surely inevitable – that we need to explain the benefits and hazards of the various approaches and encourage women to be involved in the decision of the surgical approach to their hysterectomy. And what about all this talk of ENT surgeons and dentists? They, like us, are gradually realising that many of the structures they have removed, happily through the mouth and not the side of the neck, maybe didn’t need to be removed at all. Should we remove the tongue through the cheek or not? Maybe just leave it in there! Competing interests: None declared |
|||
|
|
|||
|
David JR Hutchon, Locum Consultyant, on sabatical leave Greymouth Hospital, New Zealand
Send response to journal:
|
Sir, There is no doubt that oophorectomy can be carried out at the time of vaginal hysterectomy but laparoscopic assisted vaginal hysterectomy remains a safe and effective way of carrying out salpingo-oophorectomy. I am unable to understand why Adam Magos et al are so anti a laparoscopic approach to assist removing the ovaries at vaginal hysterectomy. As it is said there is more than one way to skin a rabbit. The risks of laparoscopy are known to be small but the risks of complications arising out of an attempt at vaginal oophorectomy are unknown. The results of a few enthusiasts cannot necessarily be generally applied. I have a number of issues with vaginal oophorectomy. Firstly, if the ovaries are normal and healthy why is it necessary to remove them? If they are unhealthy surely It must be helpful to get a good view from above before any attempt at a vaginal approach is considered. The ovaries are often adherent to the pelvic wall, The standard method for releasing ovarian adhesions is through the laparoscope. Secondly only oophorectomy rather than salpingo-oophorectomy is easily carried out through the vagina. Oophorectomy requires an excision margin through the ovarian hylum and leaves the possibility of partial oophorectomy. What is the incidence of the ovarian remnant syndrome after vaginal oophorectomy? Can the pathologist always show a margin clear round the ovarian tissue removed at vaginal hysterectomy? None of these questions have been answered because nobody has thought to answer them yet. It appears that we have at one end of the spectrum the highly skilled pure laparoscopic proponents for hysterectomy, and at the other end of the scale the highly skilled vaginal surgeons both believing that they are providing their patients with the most effective, safest and least morbid approach. In between there are ordinary surgeons using a laparoscopically assisted approach when it seems necessary. The metanalysis currently seems to support the ordinary rather than the extra-ordinary. I am ignoring the comparison with dental surgery as irrelevant. David Hutchon, Consultant Gynaecologist, Darlington Memorial Hospital, England currently on sabatical leave. Competing interests: None declared |
|||
|
|
|||
|
Adam Magos, Consultant Gynaecologist Royal Free Hospital, London NW3 2QG
Send response to journal:
|
It is unfair for David Hutchon to claim that we are “anti” laparoscopic surgery, quite the opposite. We have been running regular postgraduate courses on laparoscopic (and hysteroscopic) surgery since 1989 and will be hosting our 100th workshop in September. We have written numerous scientific papers on the subject since our very first study of the use of laparoscopic surgery to treat gynaecological emergencies which was, incidentally, published in the BMJ in 1989.1 Of the 300 or so operative procedures we carried out over the past 12 months, for instance, just under 50% were laparoscopic cases (and about 30% hysteroscopic) for indications such as endometriosis, ovarian cysts, fibroids, etc. I consider our Unit to be a proponent of laparoscopic surgery. However, we are also advocates of vaginal surgery. In the case of hysterectomy, we realised early on the advantages of vaginal surgery compared with laparoscopy.2 Firstly, the performance of salpingo- oophorectomy vaginally rarely presents a problem; the “trick” is to divide the round ligament separately to the infundibulo-pelvic ligament, a technique popularised by Shirish Sheth.3 Secondly, and irrespective of the need for oophorectomy, the only difference between vaginal and laparoscopic hysterectomy is that the operative time for the latter is considerably longer, a difference which all comparative studies have confirmed. Add to this the knowledge that one third of laparoscopic bowel and vascular injuries occur during the set-up phase of laparoscopy,4 something which is completely avoided by operating vaginally, and the case against unnecessary laparoscopic surgery seems evident. This is not to say that there is no role for laparoscopic hysterectomy or one of its variants. Certainly, in the presence of adnexal pathology, concern about endometriosis or adhesions, reduced uterine mobility or narrow vaginal access, certain types of cancer, laparoscopy is a useful alternative to laparotomy, accepting the fact that major complications appear to be more common than with open surgery as shown by the recently published eVALuate study.4 In our experience, however, these adverse factors only operate in a minority of cases, and that is why vaginal hysterectomy is our default route of surgery. 1. Magos AL, Baumann R, Turnbull AC. Managing gynaecological emergencies with laparoscopy. Br Med J 1989; 299: 371-374. 2. Richardson RE, Bournas N, Magos AL. Is laparoscopic hysterectomy a waste of time? Lancet 1995; 345: 36-41. 3. Sheth SS. Vaginal hysterectomy. Best Pract Res Clin Obstet Gynaecol 2005; 19: 307-32. 4. Chapron C, Querleu D, Bruhat M-A, Madelenat P, Fernandez H, Fabrice P, Dubuisson J-B. Surgical complications of diagnostic and operative gynaecological laparoscopy: a series of 29 966 cases. Hum Reprod 1998; 13: 867-72. 5. Garry R, Fountain J, Mason S, Hawe J, Napp V, Abbott J, et al. The eVALuate study: two parallel randomised trials, one comparing laparoscopic with abdominal hysterectomy, the second comparing laparoscopic with vaginal hysterectomy. BMJ 2004; 328: 129-38. Competing interests: None declared |
|||
|
|
|||
|
Roger A. McMaster-Fay, Clinical Lecturer University of Sydney. c/ PO Box 82 Emu Plains NSW 2750, Australia
Send response to journal:
|
The recently published meta-analysis by Johnson et al.1 on methods of hysterectomy found the ureteric injury at laparoscopic hysterectomy (LH) to be higher than at abdominal hysterectomy (AH); (10/796 -v- 1/512 with OR of 3.83 [0.94, 15.57]). Analysing the data of Lumsden et al.2, Johnson et al.1 makes important errors with regard to ureteric injuries quoting two ureteric injuries in the LH arm and one in the AH arm (Fig.27.1). In fact there were no ureteric injuries after LH and two after AH. Of the two urinary tract injuries in the LH arm (Table 3.2), one was a cystotomy and the other was in a patient who despite being randomized to the LH arm, “opted to have an AH after randomization but prior to being admitted for operation”. 2 Our recently published article3 presents data that indicates that ureteric injury at LH is more the result of surgical inexperience and that in experienced hands the rate of ureteric injury is probably no higher than for AH. Johnson et al.1 analysed the Garry et al.4 trial which reported five ureteric injuries in the LH arm of the ‘abdominal trial’ but most of the procedures performed as laparoscopically-assisted vaginal hysterectomies. Thus most of the patients the ‘abdominal trial’ (i.e. not suitable for the ‘vaginal trial’) and randomised to the LH arm had laparoscopically-assisted vaginal hysterectomies, indicating that the surgeons involved had little experience at LH. Johnson et al.1 included the data of Langebrekke et al.5 who reported two ureteric injuries in the LH arm in one hospital where the staff had little training in LH prior to their trial. This prompted the authors to recommend: “a need for thorough training prior to embarking on such a technically advanced operation method”.5 For meaningful comparison of methods of hysterectomy the surgeons must be experienced. References: 1. Johnson N, Barlow D, Lethaby A, Tavender E, Curr L, Gary R. Methods of hysterectomy: systematic review and meta-analysis of randomised controlled trials. BMJ 2005:330;1478-81. 2. Lumsden MA, Twaddle S, Hawthorn R, Traynor I, Gilmore D, Davies J, et al. A randomised comparison and economic evaluation of laparoscopic- assisted hysterectomy and abdominal hysterectomy. Br J Obstet Gynaecol 2000:107:1386-91. 3. McMaster-Fay RA, Jones RA. Laparoscopic hysterectomy and ureteric injuries: a comparison of the initial 275 and last 1,000 cases using staples. Gynecological Surgery 2006;in press. 4. Garry R, Fountain J, Mason S, Hawe J, Napp V, Abbott J, et al. The eVALuate study: two parallel randomised trials, one comparing laparoscopic with abdominal hysterectomy, the other comparing laparoscopic with vaginal hysterectomy. BMJ 2004: 328;129-38. 5. Langebrekke A, Eraker R, Nesheim B-I, Urnes A, Busund B, Sponland G. Abdominal hysterectomy should not be considered as a primary method of uterine removal: A prospective randomized study of 100 patients referred for hysterectomy. Acta Obstet Gynec Scand 1996:75:404-7. Competing interests: None declared |
|||
|
|
|||
|
Neil P Johnson, Associate Professor & Consultant University of Auckland & National Women's Health @ Auckland Hospital, New Zealand
Send response to journal:
|
I am grateful to Roger McMaster-Fay for his interest and response (1) to our systematic review of the surgical approaches to hysterectomy (2). Roger has pinpointed an error in one of our 67 meta-analysis graphs in the corresponding published Cochrane review (3), where a bladder injury in a woman in the laparoscopic hysterectomy arm of the trial by Lumsden et al (4) was incorrectly included in these meta-analysis graphs as a ureteric injury. This will be corrected in the next issue of the Cochrane Library, but it does not alter the statistical significance of any results, nor any conclusions of the review. Thus the incidence of ureteric injury in the review of RCTs is 1 in 86 (not 1 in 78) amongst women randomised to laparoscopic hysterectomy compared to 1 in 492 amongst women randomised to abdominal hysterectomy, still a difference that is not statistically significant; the data in our BMJ publication are otherwise correct (2). It must be emphasised that our outcome ‘urinary tract injury’ involved a pooling of bladder and ureter injuries. Hence our quote of two urinary tract injuries (one ureteric injury and one bladder injury) in the laparoscopic hysterectomy arm of the randomised controlled trial (RCT) by Lumsden et al (4) was correct (2). Roger’s other point is one which has seen statisticians arguing with clinicians until both are blue in the face – it is not an error in our review. We have presented an intention-to-treat (ITT) meta-analysis of urinary tract (bladder or ureter) injury (2). In this meta-analysis, the data of Lumsden et al (4) have correctly been included, as the authors did themselves in their ITT analysis, as showing two urinary tract injuries (one bladder injury and one ureteric injury, as above) in the laparoscopic hysterectomy arm, and one urinary tract (ureteric) injury in the abdominal hysterectomy arm of the RCT. Roger is quite right that the woman who had a ureteric injury in the laparoscopic hysterectomy arm had opted to have an abdominal hysterectomy rather than the laparoscopic hysterectomy to which she was randomly allocated (4). This phenomenon is not unique to the Lumsden RCT, as major complications have occurred in other RCTs in women who have opted not to undergo the procedure to which they were allocated, but instead to undergo the comparison procedure (5). Those of us with an intuitive prior belief that laparoscopic hysterectomy, where possible, is a ‘better option’ than abdominal hysterectomy, and who would wish to ‘conveniently move’ some of these complications from the laparoscopic hysterectomy arm, need to remember that intention-to-treat is the only statistically valid analysis of a randomised trial. Nonetheless, there is continuing debate whether adverse events should be analysed by intention to treat or ‘per protocol’: either way, RCTs have shown more ureteric injuries associated with laparoscopic versus abdominal hysterectomy, but the numbers are insufficient to show statistically significant differences. Such controversies and arguments underscore the need for yet more RCT data, powered adequately to assess important outcomes such as ureteric injury. Roger’s paper (6) is one of a growing number of observational studies that have suggested that the ‘first couple of hundred’ advanced laparoscopic procedures appear to be more hazardous than ‘the next thousand’ cases in any surgeon’s series. Although I imagine that surgeons whose data suggest they are progressively experiencing more complications would be less inclined to publish their data, I am fully persuaded that the comparative urinary tract injury rates for laparoscopic versus abdominal hysterectomy might be different if the surgeons involved in the trials had all performed at least 300 laparoscopic hysterectomies rather than, say, 20. References: 1) McMaster-Fay RA. Methods of hysterectomy and ureteric injury. bmj.com 8 January 2006. 2) Johnson N, Barlow D, Lethaby A, Tavender E, Curr L, Garry R. Methods of hysterectomy: systematic review and meta-analysis of randomised controlled trials. Brit Med J 2005; 330: 1478-81. 3) Johnson N, Barlow D, Lethaby A, Tavender E, Curr E, Garry R. Surgical approach to hysterectomy for benign gynaecological disease. The Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No.: CD003677.pub2. DOI: 10.1002/14651858.CD003677.pub2. 4) Lumsden MA, Twaddle S, Hawthorn R, Traynor I, Gilmore D, Davies J, et al. A randomised comparison and economic evaluation of of laparoscopic -assisted hysterectomy and abdominal hysterectomy. Brit J Obstet Gynaecol 2000; 107: 1386-91. 5) Garry R, Fountain J, Mason S, Hawe J, Napp V, Abbott J, et al. The eVALuate study: two parallel randomised trials, one comparing laparoscopic with abdominal hysterectomy, the other comparing laparoscopic with vaginal hysterectomy. Brit Med J 2004; 328: 129-38. 6) McMaster-Fay RA, Jones RA. Laparoscopic hysterectomy and ureteric injuries: a comparison of the initial 275 and last 1,000 cases using staples. Gynaecol Surg 2006; DOI 10.1007/s10397-006-0178-0 Competing interests: None declared |
|||