Research Review by Dr. Shawn Thistle©



Study Title:

Efficacy of topical non-steroidal anti-inflammatory drugs in the treatment of osteoarthritis: meta-analysis of randomised controlled trials


Lin J, Zhang W, Jones A, Doherty M

Publication Information:

British Medical Journal 2004; July 30


Osteoarthritis (OA) is the most common type of arthritis and causes a great amount of disability, particularly in the elderly. As such, OA imparts a severe burden not only on patients, but the health care system that must deal with it. Non-steroidal anti-inflammatory (NSAID) drugs are commonly prescribed for OA, but carry a risk of gastrointestinal ulcers (twice the risk in users of NSAIDs vs. those who don't use them).

When applied topically, NSAIDs bypass the gastrointestinal system and therefore reduce the risk of this severe side effect.

The aim of this meta-analysis was to determine the utility of topical NSAIDs in the treatment of OA. Studies included in the analysis had to be randomized controlled trials comparing the use of topical NSAIDs to placebo or oral NSAIDs in patients with clinical or radiographic evidence of OA. No language restrictions were applied to the literature search. From a pool of 77 trials identified through the literature search, 18 met the inclusion criteria, resulting in analysis of data for 1983 patients. All of the trials, except for one, were completely or partially sponsored by pharmaceutical companies.

Analysis of the pooled data revealed the following: 1) topical NSAIDs were superior to the placebo for the first two weeks of treatment but not thereafter 2) topical NSAIDs were not as effective as oral NSAIDs at any time 3) topical NSAIDs improved function and stiffness more that placebo in the first two weeks but not thereafter 4) topical NSAIDs resulted in fewer adverse events than oral NSAIDs, but those using the topical medications experienced a higher incidence of local reactions (skin irritation etc.)

Of note is the authors' finding of a statistically significant asymmetrical funnel plot for this analysis….In English, this means that negative studies were less likely to be published and smaller studies are more likely to produce larger effect sizes.

This combined with the lack of longer term studies (i.e. greater than one month) and the fact that studies using different topical NSAIDs were included in the analysis, weaken the results of this study somewhat, but also highlight the need for longer, well-designed trials to investigate this question.

Conclusions & Practical Application:

The evidence from this study indicates that the current recommendation for long-term use (i.e. greater than one month) of topical NSAIDs for the treatment of osteoarthritis cannot be supported.

Some major organizations, such as the American College of Rheumatology, should revise their current guidelines to reflect this finding, and aim to fund future research in this area.

To the average patient with OA, trying a topical NSAID is still reasonable, but perhaps not for a lengthy time period. As always, consult your doctor before changing your current treatment regimen, or beginning a new one.