A modification of the redistribution to the right algorithm using disease markers

Abstract

Disease markers are time-dependent covariates which describe progression towards development of disease. Traditional methods in survival analysis do not make use of available data on these markers to recover additional information from censored individuals. Using a heuristic modification of the redistribution to the right algorithm (Efron, 1967), a new approach for recovering information for censored individuals using disease markers is proposed. Additionally, the statistical properties of the proposed method are examined. There are two possible advantages to this modification: (i) bias reduction when censoring is informative, and (ii) an increase in efficiency in the case of truly noninformative censoring.