Background Acetaminophen (paracetamol) is recommended as initial therapy for osteoarthritis of the hip or knee according to guidelines of the American College of Rheumatology (ACR) prior to use of nonsteroidal anti-inflammatory drugs (NSAIDs), athough patient surveys indicate that most patients regard NSAIDs as having greater efficacy than acetaminophen.

Objectives To compare responses of patients with osteoarthritis of the hip or knee in a randomised, controlled clinical trial of diclofenac/misoprostol (Arthrotec, ARTH) compared to acetaminophen (ACET) according to severity at baseline of scores of the Western Ontario McMaster (WOMAC) osteoarthritis scale and visual analogue scale (VAS) for pain, the two primary outcomes in the clinical trial.

Methods 227 patients with radiographic stage 2–4 osteoarthritis of the hip or knee received either 6 weeks of ARTH 75 mg twice a day or ACET 1,000 mg four times a day after a washout period, and then the other drug for 6 weeks, in a randomised, double-blind, multicenter, cross-over clinical trial. Changes in scores for the two primary outcomes while taking ACET or ARTH were analysed accoring to tertiles at baseline for each scale.

Results Overall, WOMAC pain scores were improved by 7.8 of 100 units (p < 0.001), and pain VAS scores were improved by 14.6 of 100 units (p < 0.001), while taking ARTH verus ACET. Differences between changes in WOMAC scores while taking ARTH versus ACET were 3.1 for patients in the lowest tertile at baseline (p = 0.30), compared to 9.5 for the middle tertile (p < 0.001), and 12.2 for those in the highest baseline tertile (p < 0.001). Differences between changes in pain VAS scores while taking ARTH versus ACET were 4.9 for patients in the lowest tertile at baseline (p = 0.25), compared to 16.6 for the middle tertile (p < 0.001), and 21.0 for those in the highest baseline tertile (p < 0.001).

Conclusion Patients with mild osteoarthritis of the hip or knee have similar responses to ARTH and ACET, possibly because patients in the lowest tertile have the least capacity for improvement. Patients with more severe osteoarhtritis appear much more likely to benefit from ARTH.