Research question Does diclofenac gel work better than placebo for people with moderately painful osteoarthritis of one knee?
Answer Diclofenac gel (1.16%) relieves pain and stiffness somewhat better than a placebo, but it takes about a week to start working.
Why did the authors do the study? European and American guidelines recommend oral or topical non-steroidal anti-inflammatory drugs as second line treatment for osteoarthritis. The authors, supported by the drug manufacturer Novartis, wanted to evaluate the topical gel in a clinical trial with modern robust methods.
What did they do? 237 German patients with osteoarthritis of one knee took part in a double blind, placebo controlled trial. They had moderate pain and radiographic changes, including osteophytes. After a washout period, during which they could take only paracetamol (up to 2 g a day), participants used 4 g of diclofenac gel (1.16%) four times a day, or an identical placebo, for three weeks. They kept daily diaries of symptoms and use of paracetamol. The authors assessed participants' pain, stiffness, and physical function at the end of each week using a 100 mm visual analogue scale of pain intensity and the Western Ontario and McMaster osteoarthritis index questionnaire. At the end of three weeks, participants were asked to rate their treatment overall on a scale from 0 (poor) to 4 (excellent). The main measure of outcome was patient's daily record of pain on movement, averaged over the first 14 days.
What did they find? Patients using the diclofenac gel reported a slightly greater improvement in pain on movement than those using the placebo. Averaged over 14 days, their scores fell by 14 mm on the visual analogue scale compared with 10 mm for patients using the placebo (absolute difference 4 mm, P = 0.02). The difference increased a little during the study, averaging 6 mm in the second and third week (P = 0.005).
Patients using the diclofenac improved significantly more than controls in their scores on the osteoarthritis index. During weeks 2 and 3 they reported an extra 7-9 point improvement on the three 100 point scales measuring pain, stiffness, and physical function.
At the end of the study, 69% of patients using diclofenac and 58% of controls rated their treatment as good, very good, or excellent (P = 0.03). Few patients reported side effects. One patient in each group developed contact dermatitis.
What does it mean? Diclofenac gel worked better than placebo in this study, but the benefits were modest and took about a week to develop. The authors note that oral non-steroidal anti-inflammatory drugs work much faster but are associated with substantially more side effects. This is the first study to suggest that topical treatments can be effective beyond two weeks. But the results should not be extrapolated to patients with severe pain or extensive radiographic changes. Two thirds of the patients in this study were well enough not to need paracetamol during the washout period.