Summary of findings 2. Non‐selective NSAIDs versus selective cyclo‐oxygenase‐2 inhibitors.
| NSAIDs compared to cyclo‐oxygenase (COX)‐2 inhibitors (COXIBs) for acute gout | ||||||
| Patient or population: acute gout Setting: hospital Intervention: NSAIDs Comparison: cyclo‐oxygenase (COX)‐2 inhibitors (COXIBs) | ||||||
| Outcomes | Relative effect (95% CI) | Anticipated absolute effects* (95% CI) | Certainty of evidence (GRADE) | What happens | ||
| Without NSAIDs | With NSAIDs | Difference | ||||
| Pain: mean change from baseline on 5‐point (0 to 4, 0 no pain) Likert scale Follow‐up: 7 days №. of participants: 1044 (6 RCTs) | ‐ | Mean change in pain from baseline without NSAIDs was 1.99 points | Mean change in pain from baseline with NSAIDs was 2.02 points | MD 0.03 points higher (0.07 lower to 0.14 higher) | ⊕⊕⊕⊝ MODERATEa | NSAIDs probably result in little to no difference in mean change from baseline pain compared to COXIBs |
| Inflammation (swelling): mean change from baseline on 4‐point (0 to 3, 0 no swelling) Likert scale Follow‐up: 7 days №. of participants: 1044 (6 RCTs) | ‐ | Mean change in inflammation from baseline without NSAIDs was 1.92 points | Mean change in inflammation from baseline with NSAIDs was 2 points | MD 0.08 points higher (0.07 lower to 0.22 higher) | ⊕⊕⊕⊝ MODERATEa | NSAIDs probably result in little to no difference in inflammation (swelling). The 95% confidence intervals included both an increase and a decrease in swelling |
| Function: mean change in pain with movement from baseline on 4‐point (0 to 3, 0 best function) Likert scale Follow‐up: 7 days №. of participants: 91 (1 RCT) | ‐ | Mean change in function from baseline without NSAIDs was 0 | Mean change in function from baseline with NSAIDs was 0.04 | MD 0.04 higher (0.17 lower to 0.25 higher) | ⊕⊕⊝⊝ LOWa,b | NSAIDs may result in little to no difference in function. The 95% confidence intervals included both an improvement and a reduction in function |
| Participants' global assessment of treatment success on 5‐point (0 very good) Likert scale Follow‐up: 7 days №. of participants: 730 (4 RCTs) | ‐ | Mean participant's global assessment of treatment success without NSAIDs was 1.36 points | Mean participant's global assessment of treatment success with NSAIDs was 1.44 points | MD 0.08 points higher (0.04 lower to 0.21 higher) | ⊕⊕⊕⊝ MODERATEa | NSAIDs probably do not increase participants' global assessment of treatment success. The 95% confidence intervals include both treatment success and no success |
| Health‐related quality of life on SF‐36 (mental health component) Scale from 0 to 100 (0 worst quality of life) Follow‐up: 7 days №. of participants: 222 (1 RCT) | ‐ | Mean health‐related quality of life without NSAIDs was 51.11 points | Mean health‐related quality of life with NSAIDs was 50.93 points | MD 0.18 point lower (6.7 lower to 6.34 higher) | ⊕⊕⊝⊝ LOWa,b | NSAIDs may result in little to no difference in health‐related quality of life. The 95% confidence intervals include both improvement and no improvement in quality of life |
| Withdrawals due to adverse events Follow‐up: 7 days №. of participants: 1243 (6 RCTs) | RR 2.34 (1.33 to 4.14) | 2.9% | 6.7% (3.8 to 11.9) | 3.9% more (1 more to 9 more) | ⊕⊕⊕⊝ MODERATEa | NSAIDs probably increase withdrawals due to adverse events. Absolute change 4% more (1% more to 9% more). NNTH 26 (NNTH 11 to 105)c |
| Total adverse events Follow‐up: 7 days №. of participants: 1232 (6 RCTs) | RR 1.94 (1.36 to 2.78) | 23.1% | 44.8% (31.4 to 64.2) | 21.7% more (8.3 more to 41.1 more) | ⊕⊕⊕⊝ MODERATEa | NSAIDs probably increase adverse events. Absolute change 22% more (8% more to 41% more). NNTH 5 (NNTH 3 to 13)c |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; NNTH: number needed to treat for an additional harmful outcome; NSAID: non‐steroidal anti‐inflammatory drug; RCT: randomised controlled trial; RR: risk ratio; SF‐36: 36‐Item Short Form questionnaire. | ||||||
| GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect | ||||||
aXu 2016 was at high risk of bias regarding blinding of participants and personnel; Rubin 2004 and Willburger 2007 had unclear risk of bias for these domains. In three trials (Schumacher 2002; Schumacher 2012; Willburger 2007), the role of funding by the pharmaceutical industry might have biased the results. Only 1 out of 6 trials was at low risk of bias (Li 2013).
bNumber needed to treat for an additional beneficial outcome (NNTB) or number needed to treat for an additional harmful outcome (NNTH) = n/a when result is not statistically significant. NNT for dichotomous data was calculated using Cates NNT calculator (http://www.nntonline.net/ebm/visualrx/help.asp).
cEvidence from one single trial with a small number of participants (45 in each arm), downgraded for imprecision.