Randomized Controlled Trials for Arthroscopy in Degenerative Knee Disease: Was Conservative Therapy Appropriately Tried Prior to Arthroscopy?

PURPOSE:

We aimed to determine if the randomized controlled trials (RCTs) evaluated in the most recent meta-analysis on arthroscopic surgery for degenerative knee arthritis included documented trials of appropriate conservative treatment prior to randomization.

METHODS:

We selected all RCTs of the most recent meta-analysis by Brignardello-Petersen and recorded for each RCT, if physiotherapy prior to randomization was mandatory. We compared the treatment effect of arthroscopy in studies in which physiotherapy prior to randomization was mandatory versus studies in which it was not. This review was registered in the PROSPERO database (CRD42017070091).

RESULTS:

Of the 13 RCTs in the meta-analysis, there were 2 in which physiotherapy prior to randomization was mandatory. In 1 additional multicenter RCT, prior conservative treatment was mentioned as mandatory in the publication, but not in the protocol. The treatment effects attributed to arthroscopy in terms of short-term pain (P = .0037), short-term function (P = .0309), and long-term function (P = .0012) were larger in studies in which prior physiotherapy was mandatory.

CONCLUSIONS:

Although the most recent meta-analysis claims that it is based "on patients who do not respond to conservative treatment," physiotherapy was mandatory prior to randomization only in 2 of the 13 studies. As several orthopaedic guidelines recommend that the first line of treatment in patients with degenerative arthritis of the knee should be conservative, for instance with physiotherapy, and the question of performing arthroscopy arises once conservative treatment fails, 11 of the 13 RCTs failed to adhere to these accepted guidelines. Therefore, patient selection in these 11 studies may not represent the typical indications for arthroscopy, where patients have tried conservative management prior to being offered surgery. When comparing studies where prior physiotherapy was mandatory to studies in which it was not mandatory, there were statistically significant effects favoring arthroscopy in terms of pain in the short term, and for function both in the short and the long term. These findings suggest that the treatment effects attributed to arthroscopy were higher when prior physiotherapy was mandatory. Given these findings, the external validity of most of these RCTs, and the resulting "strong recommendation against the use of arthroscopy in nearly all patients with degenerative knee disease," is called into question.