An Open Letter to the Editor of The BMJ
Dear Dr. Godlee:
I have previously corresponded with you about this subject. For the record, I would like to state I am absolutely against unnecessary surgery and there is no doubt that, in the past, arthroscopy of the knee has been an overused intervention. This is an open letter and is being copied to a variety of interested parties.
I was asked to speak about meniscal surgery a couple of weeks ago in London and whilst researching my lecture I came across the series of articles in The BMJ, that culminated in “Arthroscopic Surgery for Degenerative Knee Arthritis and Meniscal Tears: A Clinical Practice Guideline” published in May 2017.1
Unfortunately, it is my view that you have once again published opinions that cannot be justified by the evidence and that seem to have been produced by an erroneous interpretation of the data (detail below). This is likely to adversely affect a substantial number of patients in the United Kingdom, especially as yours is such an influential journal.
First, I would question the makeup of the panel that produced the guidelines, which are likely to be applied in the U.K. It is indeed an international panel but there seems to be no input from the British Orthopaedic Association, British Association for Surgery of the Knee, or the Royal College of Surgeons. The only British input, as far as I can make out, is from a general practitioner who is also an Associate Editor of The BMJ and a physiotherapist from the spinal unit in Oxford, neither of whom would seem to have any significant experience in knee surgery.
The orthopaedic input seems to be (1) Professor Harris in Australia whose “areas of research interest relate to surgical outcomes (trauma and elective, orthopaedic and non-orthopaedic), predictors of patient satisfaction after surgery, psychosocial factors predicting outcome after injury and surgery, systematic reviews and randomized trials of orthopaedic interventions, compensation research, and aspects of clinical epidemiology (including quality and safety) related to surgery,” and whose clinical work seems related to arthroplasty; (2) Rudolph Poolman, a Dutch orthopaedic surgeon who seems to be heavily involved in undertaking systematic analyses and meta-analyses on a variety of subjects; and (3) Gunnar Knutsen who has done some good work in knee surgery but mainly to do with articular cartilage repair. The rest seem to be professional epidemiologists and physiotherapists, apart from the 3 “patient representatives,” who seem to be health bloggers from Canada.
I question the wisdom of not including any orthopaedic, rheumatology, or patient representatives from the U.K. when producing practice guidelines, which are most likely to be influential in our National Health System. Our health system is different from many in the world. For instance, it is unusual for a patient to arrive in my clinic having not previously been triaged by a “musculoskeletal service” and most patients have already had symptoms for many months, despite prolonged conservative treatment. This moves them into “potential for surgery” based on the already published ESSKA guidelines. The British Association for Surgery of the Knee is currently in the process of producing recommendations for knee arthroscopy for meniscal tears, led by Professor Andy Price from Oxford who is both an academic and a busy and respected knee surgeon.
Now I will address the evidence. This started with the article by Thorlund (a physiotherapist) et al. in The BMJ in 2015.2 A response to this was published in the Bone and Joint Journal (formerly the British Journal of Bone & Joint Surgery).3 A full version of this was published as a rapid response when Thorlund's article was republished in the British Journal of Sports Medicine.
I had the honor of speaking in a debate with Professor Lohmander (senior author of the above article) last year, who stated that there is never an indication for MRI or arthroscopy in anyone over 40. He lost all credibility with the audience when he was asked, “if you left this meeting, slipped on the steps, twisted your knee and it locked in flexion, what would you do?” His reply was he would go to a physiotherapist and would not consider an arthroscopy as it is never indicated. This is clearly an extreme stance to take and is perhaps an indication of his level of bias. It is noteworthy the majority of the quoted articles come from a small area of Scandinavia with many of the authors interlinked on many publications.
In 2016, you published an article “Exercise Therapy Versus Partial Meniscectomy for Degenerative Meniscal Tears in Middle Aged Patients” by Kise et al.4 Their inclusion criteria for the study were “unilateral knee pain more than two months; medial degenerative meniscal tear verified by MRI.” This was then defined as “an intrameniscal linear magnetic resonance imaging signal penetrating one or both surfaces of the meniscus.”
When lecturing in London, I asked the assembled audience of approximately 300 orthopaedic knee surgeons, “who would regard this as an indication for surgery?” There was not a single positive response, which I am sure you will find reassuring. No reputable knee surgeon would operate on patients based on these criteria. We have known for a long time that these types of changes on MRI scanning are often an incidental finding. Unless closely tied in with symptoms and signs, these findings are irrelevant. Even then, 20% of the “exercise therapy” group crossed over to have surgery.
This seems to be a common finding in many studies, no matter how poor the decision-to-operate criteria are, with between 20% and 30% of patients switching from conservative treatment to surgery, in the case of Herrlin's study,5 because of “disabling symptoms,” who were then cured by their operation.
Perhaps more concerning was the article “Knee Arthroscopy Versus Conservative Management Inpatients With Degenerative Knee Disease: A Systematic Review” by Brignardello-Petersen et al.6 This is essentially a reprise of all the points made and raised in the article by Thorland et al. and, as a reviewer of manuscripts for The Bone & Joint Journal and several other orthopaedic journals, it is very likely I would have rejected this latest review on the basis that it is strikingly similar in form, content, and conclusion to the earlier paper. From my review, it appears to contain essentially the same information and detail as the Thorlund article, which is simply analyzed with a few additional references commenting on the comparison between injection therapy with arthroscopy. I don't believe this latest article adds anything new to the previously published study and, perhaps not surprisingly, the authors (who as far as I can tell are largely not involved in knee surgery) reach the same conclusions. Incidentally, the authors quote themselves no fewer than 12 times including a publication on symptomatic aortic stenosis, a good way to improve your citation index!
I question how the authors have the required knowledge base to critically analyze the articles they have chosen to review. Once again they seem to be predominately epidemiologists, with the only orthopaedic input coming from an “orthopaedic resident” whose major interest seems to be research methodology. I personally would not have the confidence to cast judgment on a paper from a different specialty of orthopaedics, let alone a subject about which I do not have an intimate knowledge or extensive professional background.
I note that the article was peer reviewed, but as your chosen reviewer is once again from the same area of Scandinavia, has previously published an article7 saying that surgery for meniscal tears is no better than sham surgery, and is again quoted in this publication, it is at least questionable whether the review can be seen to be totally unbiased or independent.
Degenerative knee disease is very different from symptomatic degenerative meniscal tears and they should not be lumped together, but the authors may not appreciate the difference nor to be able to critically assess the evidence in front of them, given that their skill sets are in different areas. They have committed what I believe to be a cardinal error of accepting the conclusions of a randomized trial without delving into the details to see if the conclusions are justified, which in many cases they are not.
The authors have ignored a systematic review by Lamplot and Brophy in the The Bone & Joint Journal last year,8which shows that meniscal surgery in the middle aged is often beneficial. They also appear to have missed at least one prospective, randomized, single-surgeon trial of arthroscopic treatment for degenerative disease, which clearly shows benefit from arthroscopic intervention.9 In the study by Herrlin et al.,5 30% of patients crossed over to surgery because of “disabling symptoms” and the recent study by Gauffin et al. published in the American Journal of Sports Medicine10 concludes that “knee arthroscopic surgery may be beneficial for middle-aged patients with meniscal symptoms.”
This last article is interesting in that it is very similar to another study by Thorlund et al., published in The BMJ11where they come to an entirely opposite conclusion based on similar outcomes! Thorlund's argument goes that both traumatic and degenerative tears improve after surgery but as it has previously been shown that surgery for degenerative tears is no better than placebo (based on a single study7 in which the conclusions are not justified, as outlined by Krych et al. in Arthroscopy),12 surgery for any meniscal tear is not justified!
Based on the above evidence, The BMJ has then published guidelines which conclude, “we make a strong recommendation against the use of arthroscopy in nearly all patients with degenerative knee disease, based on linked systematic reviews: further research is unlikely to alter this recommendation.”
I strongly believe that these conclusions cannot be justified based on the evidence presented and that they are wrong. I would be happy to discuss my detailed reasoning with you further and to introduce you to some of my patients.
I appreciate that this is anecdotal but in the last two weeks I have seen a 50 year old joiner who was struggling to work because every time he knelt down his knee locked and in desperation had come to see me privately as he had been denied surgical referral after a “normal MRI.” After taking out his degenerate bucket handle tear he was back at work after a week.
Another factory worker whose job involved a lot of kneeling who had been off work for 6 weeks with a lot of pain and swelling – back to work 10 days after surgery, pain free immediately postoperatively.
A 48-year-old marathon runner came to see me about his left knee. I operated on his right knee a year ago when he was unable to run because of pain and swelling that had not responded to conservative treatment. He had since run 2 marathons but now had developed identical symptoms in the other knee.
A 60-year-old lady for whom I did a microfracture and partial meniscectomy 11 years ago because she was unable to go hill walking, came to see me about her knee as it was starting to become symptomatic again. She has been pain free and extremely active over the intervening period, walking many mountains – not bad for a placebo.
Figure 1 is an MRI of a true symptomatic medial meniscal tear in a middle-aged golfer who had endured months of conservative treatment and pain and an inability to crouch and line up his putts! After I took out the tear and decompressed the meniscal cyst, he woke up pain free and told me he was able to sleep through the night for the first time in many months.
MRI of a true symptomatic medial meniscal tear in a middle-aged man who had endured months of pain and unsuccessful conservative treatment. After removal of the tear and decompression of the meniscal cyst, the patient was pain free.
I could go on. The decision to operate is based on a clinical evaluation with MRI having a limited place to play in decision making. It is decision making that is the key to getting good results from any surgery, and in this instance with an operation that, in skilled hands, is quick, atraumatic, relatively painless and with a very, very low complication rate.
Degenerative disease and degenerative meniscal tears in the knee are entirely different entities and approached in a different fashion. I do not treat average patients, I treat the patient in front of me, predominately based on the history and findings on clinical examination and taking into account his or her individual circumstances and aspirations.
In summary, I object to the conclusions of your publication for the reasons above, but especially the unnecessary combination of degenerate knee joint disease (osteoarthritis) and degenerate meniscal tears, as they are completely different entities. I object to the sweeping conclusion that surgery for degenerate meniscal tears is not indicated in “nearly all patients.” I agree that most do not need surgery, but a carefully selected minority benefit greatly after at least 3 months of conservative treatment has failed.
The statement that further research is unlikely to change the conclusion is way off the mark. Only when proper studies are carried out, with proper inclusion criteria and surgery carried out with appropriate expertise, can this statement be justified. Separate studies will need to be carried out with regard to degenerative disease, as there are certainly observational studies showing that years of improved symptoms and delay to knee replacement can be achieved with surgical intervention when conservative measures have failed.9, 13, 14 The many patients who come to see me requesting me to carry out the same surgery as I carried out on their other knee years previously, cannot all be wrong!
Article taken from www.arthroscopyjournal.org