Eligibility criteria in knee osteoarthritis clinical trials: systematic review

There is an increasing concern over generalizability of trial results. We investigated eligibility criteria of knee osteoarthritis clinical trials. Eligible trials were randomized, placebo-controlled trials that were identified by searches in MEDLINE, SCOPUS, and the Cochrane Central Register of Controlled Trials. We then attempted to extract data on the eligibility criteria by employing predetermined criteria. From 355 randomized knee osteoarthritis trials, we reviewed data categorized by non-osteoarthritis-related and osteoarthritis-related factors. A variety of items were used in the eligibility criteria. Regarding the non-osteoarthritis-related factors, ethical considerations, such as inability to give an informed consent (79.4 %) or medical conditions potential for risks by test treatments (56.0 %) or by participation in a trial (57.2 %), were the common reasons for excluding patients from a trial. Concerning the osteoarthritis-related factors, most of the trials did not specify age and symptom severity. When trials specifying these items were investigated, patients with 40 to 80 years in age and grade 2 in the Kellgren-Lawrence scale were mostly included into a trial. For the pain intensity, patients with a parts per thousand yen20-40 in the 100-mm visual analog scale and a parts per thousand yen6-8 in the WOMAC pain subscale were commonly enrolled into a trial. These findings warrant further investigation on the generalizability of trial results.