Combining 24-2 and 10-2 visual field tests can improve power and lessen time to detect progression in clinical trials

Abstract

Purpose : To investigate the ability of 24-2 and 10-2 visual fields to detect glaucomatous progression in up to 2 years of follow-up when clustering of tests is performed at baseline and final visits.

Methods : 97 eyes (97 patients) with or suspected glaucoma were followed for at least 1 year with 24-2 and 10-2 tests on the same day. These patients had 4 tests at baseline (within 2 months) and every 4 months thereafter. Based upon published methodology,[1] data from these eyes (slopes and residuals categorized by MD severity) were used to simulate (N=10,000) a testing paradigm with increased test clustering at the tails of follow-up (4 tests at baseline and 4 at final visit). Progression was defined based upon mean deviation (MD) slopes at matched 95% specificity for the 24-2, 10-2, and combined 24-2 and 10-2 tests.

Results : The mean rate of MD progression with 24-2 and 10-2 tests in the study data was -0.39 dB (95% CI: -.53 to -0.24) and -0.43 dB (-0.57 to -0.29), respectively. In the simulated data, 24-2 progression was detected in 34.2% of eyes at 2 years. With 10-2, progression was detected in 35.0% of eyes at 2 years. Combined 24-2 and 10-2 detected progression in 43.6% of eyes at 2 years. Assuming an effect size of 30% and 100 eyes/patients per arm, a 2-year duration randomized clinical trial would have 61% power with 24-2 MD slopes as endpoint and 62% with 10-2. When combining both tests, power increased to 91% in 2 years under the same assumptions.

Conclusions : With increased clustering and more frequent visual field tests, a combination of 24-2 and 10-2 fields may increase power and decrease time to detect progression of glaucoma in short duration clinical trials.
[1] Wu Z, Medeiros FA. Sci Rep. 2018;8(1):4889.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.