3128 Copula-Based Regression Models for Correlated Bivariate Binary Outcomes: Application to Ophthalmologic Data Structures

Abstract

OBJECTIVES/SPECIFIC AIMS: To account for association between the pair of binary outcomes, we adopt the Clayton and Frank copulas to indirectly specify their joint distributions. METHODS/STUDY POPULATION: We propose a regression model for the joint modelling of correlated bivariate outcomes using copulas. RESULTS/ANTICIPATED RESULTS: develop full maximum likelihood inference.