Abstract
Abstract: :
Purpose: To compare the effect in glaucoma patients of travoprost and latanoprost on IOP and conjuctival hyperemia Methods:80 Glaucoma patients with both eyes on the same meds and IOP values within 1.5 mm Hg were randomly assigned topical travoprost 0.004% to one eye and latanoprost 0.005% to the other, in addition to the usual non–prostaglandin glaucoma medications.After one month, a masked comparison was made of the IOP and conjunctival hyperemia score (0 to +4), using a two tailed paired t test. Results: Baseline intraocular pressure and conjunctival hyperemia values were not statistically significantly different between the two eyes. After one month the mean ± S.D. of IOP in the travoprost treated eye was 0.52 ± 1.88 mm Hg lower that that of the latanoprost treated eye, p< .05, and the conjunctival hyperemia score was 0.11 ± .40 greater than the latanoprost treated eye, p<.05. Conclusions: Travoprost had a slight, but statistically significantly, greater IOP lowering effect than latanoprost. Travoprost caused slight, but statistically significantly, more conjunctival hyperemia. The two eyes are essentially "conjoined twins", subject to the same genetic and environmental influences and in this study were compared simultaneously, by the same masked examiner, using the same tonometer. Therefore, the variance of the measurements of the differences between the two eyes was about 2/3 that of a paired t test of relatively unrelated pairs. The statistical power of such a paired comparison of the two eyes can be shown to be about 6 fold greater than a non–paired comparison of two groups of subjects as almost all previous prostaglandin studies have done. The design of the study also eliminated the regression to the mean artifact inherent in most "switch studies".
Keywords: clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • intraocular pressure • clinical (human) or epidemiologic studies: outcomes/complications