Eleven RCTs (1,493 patients) were included in the review. Nine trials had a prospective parallel group design and two had a prospective crossover design. Six trials scored 3 or more on the Jadad scale and five trials scored 1 or 2. Only one trial was open label.
Efficacy, peak intraocular pressure (eight trials): Compared to carbonic anhydrase inhibitors, brimonidine significantly reduced total peak intraocular pressure (WMD 0.99, 95% CI 0.45 to 1.53; significant heterogeneity Ι²=73%) when used as adjunctive therapy to beta blockers (WMD 0.85, 95% CI 0.42 to 1.29) or prostaglandin analogues (WMD 1.04, 95% CI 0.08 to 2.0) and when only adequate quality trials were compared (WMD 1.41, 95% CI 1.14 to 1.68). Significant associations were also found between brimonidine and dorzolamide but not between brimonidine and brinzolamide.
Efficacy, trough intraocular pressure (seven trials): There was no evidence of a significant difference between brimonidine and carbonic anhydrase inhibitors in reducing trough intraocular pressure (WMD 0.20, 95% CI -0.34 to 0.75; significant heterogeneity Ι²=63%). Subgroup analyses indicated that brimonidine used as adjunctive therapy to beta blockers and compared to dorzolamide significantly reduced trough intraocular pressure but brimonidine was less effective when compared with brinzolamide (results reported in the paper). There was no evidence of a significant difference between interventions when used as adjunctive therapy to prostaglandin analogues or when only trials of adequate quality were compared.
Efficacy, diurnal mean intraocular pressure (nine trials): Compared to carbonic anhydrase inhibitors, brimonidine significantly reduced diurnal mean intraocular pressure (WMD 0.62, 95% CI 0.07 to 1.18; significant heterogeneity Ι²=78%) and when brimonidine was compared with dorzolamide (WMD 0.83, 95% CI 0.36 to 1.29). There was no evidence of significant differences in the other subgroup comparisons.
Safety: Compared to carbonic anhydrase inhibitors, brimonidine was associated with significantly fewer episodes of ocular burning (RR 0.26, 95% CI 0.17 to 0.40; no significant heterogeneity; four trials) and dysgeusia (RR 0.20, 95% CI 0.11 to 0.37; no significant heterogeneity; five trials). There was no evidence of significant differences in rates of other adverse events between groups.
There was no evidence of publication bias from inspection of funnel plot symmetry.