Group Sequential Design and Monitoring of Clustered Data in Randomized Eye Trials.

Group sequential designs are the most commonly used methods for randomized clinical trials with interim monitoring plans, which control the type I error rate and protect statistical power. However, to the best of our knowledge, group sequential methods accounting for the inter-eye correlation under a unified design in randomized eye trials, do not exist in the literature. Failing to account for inter-eye correlations may lead to unnecessary study continuation or early stopping. We propose a unified group sequential design for vision research that accommodates all possible design options while accounting for inter-eye correlations when both eyes of a patient are included in the study for at least some patients. We derive closed-form formulae for the sample size/power calculations for fixed-sample designs and the information fractions during interim monitoring for group sequential designs while accounting for the inter-eye correlation for continuous endpoints. Simulation studies and an application to the Age-Related Eye Disease Study are provided. An R package iTrial implementing the proposed methodology is publicly available.