Nine RCTs were included (n=1,090). Summary quality scores ranged from 12 to 18. Treatment allocation was concealed in four studies, four studies blinded participants, investigators and examiners and an additional two studies blinded only examiners. Five studies performed an intention-to-treat analysis. Withdrawals were reported in eight studies and ranged from 3% to 16%.
The relative risk reductions ranged from 21% for trough IOP with timolol to 34% for peak IOP for latanoprost. Latanoprost had the greatest reduction in IOP for all measurements (31% to 34%; five to seven RCTs), followed by travoprost (28% to 32%; one to two RCTs) and bimatoprost (26% to 28%; one to two RCTs) with timolol (21% to 24%; five RCTs) reported the lower relative reductions. However, differences were fairly small and had overlapping confidence intervals.
There was no evidence of publication bias in any of the measurements (p>0.67).