Eleven RCTs were included in the review; the total number of participants was not reported. Sample sizes in the included studies ranged from 22 to 347. Only 5 studies were included in the meta-analysis.
Five trials were high quality. Nine trials did not report methods of randomisation. Four trials were not double-blinded; four described themselves as double-blinded but did not report methods of blinding. Six trials did not report withdrawals or losses to follow-up for each group; 2 trials had large losses to follow-up. None of the trials reported that their methods of allocation were concealed, however, two authors responded to correspondence and indicated that their method of allocation was concealed.
Pain scale score at 24 hours.
Five of the 11 included RCTs reported adequate data to be pooled (three of which were high quality). Three used verbal pain scales and two used a VAS. The standardised mean difference suggested a statistically significant reduction in pain favouring treatment with topical NSAIDs (-0.52, 95% CI: -0.91, -0.13).
When only the 3 high-quality trials were pooled (n=299), the standardised mean difference was -0.22 (95% CI: -0.45, 0.00).
When the 3 RCTs (n=459) that used verbal pain scales were pooled, the overall weighted mean difference suggested a statistically significant reduction in pain favouring treatment with topical NSAIDs (-1.30, 95% CI: -1.56, -1.03, P<0.00001). There was no evidence of significant heterogeneity (Q=1.57, P=0.46).
Sensitivity analyses for placebo-controlled trials, language and country of origin did not change the statistical significance of the overall estimate.
Adverse events.
Eight RCTs reported data on adverse events. There were insufficient data to conduct a formal meta-analysis. Two trials reported that no adverse events occurred, while four reported that some patients experienced transient stinging with instillation of the eye drops. In one trial there were three recurrent corneal erosions: two in the control group and one in the NSAID group. In one trial a patient in the control group developed a corneal abscess and three patients in the NSAID group developed urticaria; since patients in this trial also received gentamicin, the cause of the urticaria was unclear.
The authors stated that few of the trials were suitable for analysis using a funnel plot, therefore the results were not reported.