Send to

Choose Destination
Clin Drug Investig. 2010;30(7):419-37. doi: 10.2165/11533350-000000000-00000.

Oral cyclo-oxygenase 2 inhibitors versus other oral analgesics for acute soft tissue injury: systematic review and meta-analysis.

Author information

Adult Emergency Department, Auckland City Hospital, Auckland, New Zealand.



Acute soft tissue injuries are common and carry significant societal costs. Cyclo-oxygenase 2 (COX-2) inhibitors (coxibs), non-selective nonsteroidal anti-inflammatory drugs (NSAIDs) and other analgesics are used to treat acute soft tissue injuries, with ongoing debate about their analgesic efficacy, effects on tissue healing and adverse effects (AEs).


To systematically review the evidence comparing oral coxibs with other oral analgesics for acute soft tissue injuries, using the outcomes: pain, swelling, function and AEs.


The following databases were searched: MEDLINE, EMBASE, Cochrane CENTRAL, CINAHL, AMED, PEDro and SPORTDiscus. Further studies were sought through clinical trials registries, dissertations, correspondence with pharmaceutical companies and manual searches of relevant journals. There was no language restriction.


'Coxibs' were defined as drugs that inhibit COX-2 >5-fold more than COX-1; 'acute' was defined as injury occurring within 48 hours of enrollment; 'soft tissue injury' was defined as closed injuries to upper or lower limb soft tissues (ligaments, muscles or tendons).


Randomized controlled trials in humans comparing a coxib to a different class of oral analgesic agent for the treatment of acute soft tissue injuries for <30 days, and in which >or=80% of participants met the definition of acute soft tissue injury, were included. Studies were excluded if >20% of participants enrolled had back pain, cervical spine injury, repetitive strain injuries, delayed-onset muscle soreness, fractures, cartilage injury, penetrating wounds or primary inflammatory conditions (tendonitis, bursitis and arthritis). Nine out of 23 (39.1%) potentially relevant studies met the selection criteria.


A standard form was used to extract data. Included studies were screened by the authors for risk of bias using the Cochrane risk of bias tool and evidence was graded for quality using the GRADE tool.


Clinical heterogeneity was minimized by application of strict selection criteria. Statistical heterogeneity was assessed using the I2 statistic and meta-analysis was undertaken if appropriate. Weighted mean difference (WMD) was used to assess pain, relative risk (RR) to assess AEs, and Peto odds ratio (OR) to assess return to function.


The nine RCTs evaluated in the meta-analysis included 3060 patients. Coxibs were found to be equal to NSAIDs (day 7+, n = 1884, 100 mm visual analogue scale [VAS]), WMD = 0.18 mm (95% CI -1.76, 2.13), p = 0.85 and tramadol (day 7+, n = 706, 100 mm VAS), WMD = -6.6 mm (95% CI -9.63, -3.47) [single study, difference clinically insignificant] for treating pain after soft tissue injuries. Coxibs had fewer gastrointestinal AEs than NSAIDs, even with short-term use (RR 0.59 [95% CI 0.41, 0.85], p = 0.004) [low quality evidence]. Swelling was measured in two studies with no difference being found between groups, but the presentation of the data was not sufficient to allow further analysis. Coxibs were found to be unlikely to be different to NSAIDs in helping patients return to function (OR 1.0 [95% CI 0.77, 1.3], p = 0.99); however, a single study suggested they may improve time to return to function (moderate quality evidence) and may have fewer AEs than tramadol (very low quality evidence). The risk of serious AEs with both coxibs and NSAIDs in this setting was low (but incompletely defined).


More studies comparing coxibs with NSAIDs and other analgesics in the setting of acute soft tissue injuries are necessary. A different review methodology would be required to answer the question of cardiovascular risk associated with short-term use of coxibs and NSAIDs.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Springer
Loading ...
Support Center