http://www.ncbi.nlm.nih.gov/pubmed/8666628
I like this old study by Hubbard from the JBJS. It is a well designed study with a clearly defined patient cohort and they showed a clear benefit to debriding focal osteoarthritic lesions of the medial femoral condyle.
Since this study there have been three sham surgical trials on knee arthroscopy, each greeted with a media fanfare, the problem for me is that despite these studies being interesting to a degree, I am not convinced that knee arthroscopy is as dead in the water as some physicians seem to want it to be. Why?
Well, there are some serious flaws in all these studies in terms of patient groups (heterogeneity/symptoms/stage of OA/type of OA lesions etc), outcome measures and study power. For example in the trials of Moseley and Kirkley they took heterogeneous groups of patients with diffuse OA of different stages and showed that a general bit of non-specific debridement (ie no chondroplasty/microfracture etc) did not improve outcome. While in the recent study by Sihvonen patients had no mechanical symptoms, and isolated knee pain with no arthritic lesions.
http://www.ncbi.nlm.nih.gov/pubmed/18784099
http://www.nejm.org/doi/full/10.1056/NEJMoa013259
From what I can see we have learnt some interesting things from these sham surgical trials, for example we know that a general bit of non-specific debridement is probably not a good thing for people with knee pain related to diffuse osteoarthritis and no mechanical symptoms, and we have also learnt that a meniscectomy is not a good thing in terms of relieving pain in patients with an isolated degenerate meniscus, no osteoarthritis and no mechanical symptoms.
It will be interesting in the years ahead to see the role of knee arthroscopy become better defined, in my opinion the baby should not be thrown out with the bath water at this stage. We know focal osteoarthritis lesions respond better with surgery, and the role of modern treatments like chondroplasty/MACI/ACI needs to be better determined. The problem is I don't think that knee arthoscopy is of no value in the treatment of knee pain as a result of more diffuse osteoarthritis, I just think the evidence is not good enough to tell us the answer yet, it all depends on what one does once inside the knee and in which specific patient groups one intervenes.........