Nov. 29, 2004 — Nonsteroidal anti-inflammatory drugs (NSAIDs) offer only a small, short-term benefit that may not be clinically significant for people with knee osteoarthritis (OA), according to the results of a meta-analysis published in the Nov. 30 Online First issue of the BMJ. Due to the long-term harm and lack of demonstrated benefit, the investigators recommend against long-term use of NSAIDs for OA.

"Treatment guidelines for knee osteoarthritis recommend pharmacological intervention, initially with paracetamol and subsequently with an NSAID," write Jan Magnus Bjordal, MD, from University of Bergen in Norway, and colleagues. "The recent introduction of [cyclo-oxygenase-2 inhibitors] (coxibs) seemed to promise a reduction in serious adverse events related to NSAIDs, but this remains controversial. Guidelines from the European League Against Rheumatism (EULAR) state that both pharmacological and nonpharmacological interventions are needed for optimal treatment of knee osteoarthritis."

Of 10,845 patients enrolled in 23 randomized trials reviewed for this meta-analysis, 7,807 patients received adequate doses of NSAIDs and 3,038 received placebo. Median age was 62.5 years, the mean weighted baseline pain score was 64.2 mm on a 100-mm visual analog scale (VAS), and the average duration of symptoms was 8.2 years. The primary outcome was the change in overall intensity of pain.

Although the methodologic quality of trials was acceptable, 13 trials excluded patients before randomization if they did not improve while receiving NSAIDs. In one trial providing long-term data, NSAIDs had no significant effect compared with placebo at one to four years.

In all included trials, the pooled difference for pain on the VAS after two to 13 weeks was 10.1 mm (95% confidence interval [CI], 7.4-12.8) or 15.6% better than placebo. The results were heterogeneous, with the effect size for pain reduction 0.32 (95% CI, 0.24-0.39) in a random effects model. Ten trials that did not exclude nonresponders to NSAID treatment had homogeneous results, with an effect size for pain reduction of 0.23 (95% CI, 0.15-0.31).

"NSAIDs can reduce short-term pain in osteoarthritis of the knee slightly better than placebo, but the current analysis does not support long-term use of NSAIDs for this condition," the authors write. "As serious adverse effects are associated with oral NSAIDs, only limited use can be recommended."

Study limitations are the inclusion of nine trials in which outcomes were recorded with fewer than the five pain dimensions covered by the Western Ontario and McMaster Universities osteoarthritis index (WOMAC) pain subscale and possible selection bias related to low average age.

"It may be reasonable to assume that the benefits of NSAIDs may be less and the harmful effects more common in an unselected population of patients with knee osteoarthritis compared with the patients in these studies," the authors conclude.

This study received no external funding, and the authors report no competing interests.

BMJ. Posted online Nov. 30, 2004.