Arthroscopy for degenerate meniscal tears of the kneeBMJ 2014; 348 doi: https://doi.org/10.1136/bmj.g2382 (Published 02 April 2014) Cite this as: BMJ 2014;348:g2382
- Andrew Price, professor of orthopaedic surgery and consultant knee surgeon1,
- David Beard, professor of musculoskeletal sciences and codirector RCS Surgical Trials Unit (Oxford)1
- 1Botnar Institute of Musculoskeletal Sciences, NIHR Oxford Biomedical Research Unit, OUH NHS Trust, University of Oxford, Oxford OX3 7LD, UK
The NHS performs around 150 000 arthroscopic knee operations a year, with more than half involving resection of the meniscus. Therefore, close scrutiny of this intervention in the United Kingdom is entirely appropriate, particularly in the context of the ongoing drive towards providing evidence based and value based care.
Considering such high rates of surgery, it would be natural to assume that this operation is backed up by adequate evidence. However, unlike knee replacement surgery, which is supported by population based patient reported outcomes (PROMs) data and the National Joint Registry, healthcare commissioners lack the necessary data to allow informed decision making for knee arthroscopy. Detailed indications for its use need further refinement. With this backdrop, any results from high quality randomised controlled trials in this area are welcome.
The recent study by Sihvonen and colleagues therefore makes interesting reading.1 In 2005, when the trial was started, investigators were worried about the increasing numbers of arthroscopic meniscectomies performed for patients with degenerative meniscal tears. At this time it was common practice to perform arthroscopy on patients with a history of joint line pain in the presence of an isolated degenerative meniscal tear. These inclusion criteria were therefore used for a sham surgery trial that investigated the efficacy of arthroscopy. The researchers found no difference in outcome between actual and sham surgery and concluded that the practice of resection for degenerative meniscal tears should be challenged, if not discontinued.
Their study provides good evidence to challenge the indiscriminate selection of patients for arthroscopic meniscectomy, and fewer patients will probably undergo surgery for this indication in future. In addition, the researchers are to be commended on completing a placebo controlled trial of surgery. Unlike drug trials, randomised trials of surgical interventions are rare. The addition of a sham or placebo, commonplace in drug trials, is rare in surgical trials and adds a further layer of complexity.
But does the study completely answer the question of whether degenerate tears should be resected? Perhaps not. Closer inspection identifies some methodological problems that may influence interpretation.
The trial began in 2005 and was designed before this. The long duration of the trial is important because it affects the external validity and generalisability of the findings. Generalisability is a key requirement for an assessment of the impact on healthcare systems. Trials that investigate outdated or non-contemporary interventions seldom have any effect.
Whether the treatment used in the trial represents current practice is questionable, with the standard approach for this population having moved on since the trial’s inception. In general, the identification of candidates for arthroscopic surgery is now more sophisticated, with some subgroups potentially more suitable for meniscectomy than others. For example, patients who sustain a traumatic cartilage tear that results in mechanical symptoms (catching or locking) may benefit from arthroscopy.2 Similarly, a further subset of patients with degenerative tears, initially managed non-operatively, may have recurring symptoms over a longer period, as seen in recent crossover trials of arthroscopic meniscectomy in other groups.3 4 For this patient group, where discrete unstable meniscal fragments are identified on magnetic resonance imaging, meniscectomy may also be an effective treatment. Further trials are needed to evaluate the worth of arthroscopy in these distinct and separate populations.
The work also highlights some interesting problems in the design of placebo controlled surgical trials. In this study the placebo or sham procedure involved a joint washout. Sceptics will claim that joint washout is not an inert intervention and may have a therapeutic element (despite the much cited sham surgery study of Moseley and colleagues, which showed a lack of efficacy for arthroscopic washout).5 The inclusion of a washout for both groups makes interpretation of the trial more difficult. Although meniscal resection seems to provide little extra benefit over arthroscopic washout, only a third “no treatment” group could delineate the effects of a washout only. A more appropriate conclusion for this study might have been that arthroscopic meniscectomy (for degenerative tears) offers no more benefit than washout alone. This is very different from stating that arthroscopy is ineffective—one of the messages publicised in the media.6 Through no fault of the authors, the media and deliberate troublemakers might exploit a sensational headline and use this research to condemn arthroscopy for all populations. Our objective in raising these issues is not to criticise laudable efforts, but to caution against the propagation of selective or erroneous interpretations. Every worthwhile randomised controlled trial should be given the chance to influence clinical practice in an appropriate manner.
Sihvonen and colleagues’ study is a good example of the progress made in introducing rigorous and inventive methods to surgical trials. The work provides some valuable information about the indications for knee arthroscopy that will hopefully guide a change in practice, perhaps reducing the numbers of unnecessary procedures performed. At the same time, the work reinforces the urgent need for the orthopaedic community to confirm the potential, and often assumed, efficacy of meniscectomy in other patient subgroups.
This type of research is pivotal to the creation of evidence based surgical practice. We need sensible interpretation of well constructed trials to provide the necessary support for all effective surgical interventions (new or established), but equally to challenge the practice of ineffective surgery.
Cite this as: BMJ 2014;348:g2382
Competing interests: We have read and understood the BMJ Group policy on declaration of interests and declare the following interests: None.
Provenance and peer review: Commissioned; not externally peer reviewed.