WITHDRAWN: Structural Nested Models for the Effects of a Treatment on Mixed Failure-Time and Repeated Measures Outcomes

• Activity Number: 49
• Type: Invited
• Date/Time: Sunday, August 3, 2014 : 4:00 PM to 5:50 PM
• Sponsor: ENAR
• Author(s): Marshall Joffe
• Companies: University of Pennsylvania
• Keywords: causality ; joint models ; nephrology ; structural nested models

Abstract

Structural nested models are useful methods for parameterizing the effect of a repeated treatment, especially in the presence of confounding by variables affected by treatment. We consider here extension of these models to modeling jointly the effects of a treatment on mixed failure-time and repeated measures outcomes, in particular when the repeated measures outcome is not meaningful after failure. We discuss interpretation of the model parameters, and possible issues in estimation. We motivate our discussion with two examples from nephrology: assessing treatment effects on progression of renal disease when renal function cannot be measured after death or development of end-stage renal disease, and assessing the effect of erythropoietin on hematocrit and mortality in dialysis patients.