Randomized clinical trials of osteoarthritis : a review

Aim:  The objective of this study was to review published randomized clinical trials (RCTs) of osteoarthritis (OA) in three time periods to determine the characteristics of RCTs and the types of outcome measures used and whether any changes have occurred.Methods:  We identified RCTs assessing clinical efficacy of treatments for osteoarthritis published in English in 1987/1988, 1997/1998 and 2001/2002, using MEDLINE. RCTs were then assessed for baseline disease characteristics, treatment type and outcome measures utilized. We classified outcome measures into 10 broad subgroups.Results:  The number of RCTs increased with each time period (31 in 1987/1988; 36 in 1997/1998; 91 in 2001/2002). The majority of RCTs were of knee, hip or both (94% in 1987/1988, 69% in 1997/1998, 71% in 2001/2002). The median duration of RCTs for all three time periods was ≤ 3 months. In 1987/1988, nonsteroidal anit-inflammatory drugs (NSAIDs) accounted for 65% of RCTs. In 1997/1998, NSAID/COX-2 trials accounted for 41.7%, with glucosamine (14%) and hyaluronan (11%). In 2001/2002, 28% of RCTs were related to NSAID/COX-2 inhibitors and 25% to complementary therapies. The median number of participants was unchanged over the three time periods. The median number of outcome measures used at all three time periods was four (range 1–8). The most commonly used outcome measures were: (i) symptoms (pain, stiffness); (ii) function; (iii) global assessment (patient, physician); and (iv) composite scores (such as Western Ontario and McMaster [WOMAC] Osteoarthritis index) in order of use. There has been significant increase in use of function and composite score as outcome measures.Conclusion:  There has been a recent increase in both the number and variety of therapeutic interventions for OA. This study shows that currently a wide variety of outcome measures are used. This reinforces the importance of using standardized responder criteria, such as the OMERACT-OARSI set of responder criteria, which incorporate pain, function and patient's global assessment.