Online Program

Return to main conference page

All Times ET

Thursday, February 3
Thu, Feb 3, 12:30 PM - 1:30 PM
Virtual
Poster Session 3

Consequences of Unmet Proportionality of Hazard Assumption and Analytic Choices When Using the Cox PH Model in the Obesity and Longevity Relationship (305349)

View Presentation View Presentation

*Luis Miguel Mestre, Indiana University 

Keywords: Cox PH Model, Assumptions, Reproducibility

The consequences of a violation of the Cox PH model's proportionality of hazard (PH) assumption in the context of research examining risk factors for mortality are uncertain. We hypothesized that scenarios in which the PH assumption is unmet might lead to a biased hazard ratio. To test this hypothesis, we conducted data analyses that determines the performance of the Cox PH model when PH assumption is unmet. Putative results will consist on a comparison of the Cox PH model when the PH assumption is unmet with other statistical models.