Aim: To test the evidence for a difference in analgesic efficacy and adverse effects of non-steroidal anti-inflammatory drugs (NSAIDs) given by different routes.
Methods: Systematic review of published randomised controlled trials. Relevant trials were comparisons of the same drug given by different routes. Presence of internal sensitivity was sought as a validity criterion. Analgesic and adverse effect outcomes were summarised, and synthesised qualitatively.
Results: In 26 trials (2225 analysed patients), 8 different NSAIDs were tested in 58 comparisons. Fifteen trials (58%) compared the same drug by different routes. Drugs were given by intravenous, intramuscular, intrawound, rectal and oral routes in postoperative pain (14 trials), renal colic (4), acute musculoskeletal pain (1), dysmenorrhoea (1), and rheumatoid arthritis (6). Five of the 15 direct comparisons were invalid because they reported no difference between routes but without evidence of internal sensitivity. In all 3 direct comparisons in renal colic, intravenous NSAID had a faster onset of action than intramuscular or rectal. In 1 direct comparison in dysmenorrhoea, oral NSAID was better than rectal. In the 5 direct comparisons in postoperative pain, results were inconsistent. In 1 direct comparison in rheumatoid arthritis, intramuscular NSAID was better than oral. Injected and rectal administration had some specific adverse effects.
Conclusion: In renal colic there is evidence that NSAIDs act quickest when given intravenously. This may be clinically relevant. In all other pain conditions there is a lack of evidence of any difference between routes. In pain conditions other than renal colic, there is, therefore, a strong argument to give oral NSAIDs when patients can swallow.