Abstract:
|
In many clinical studies, repeated measures and survival outcomes are of primary endpoints. To study the relationship between longitudinal profiles and survival processes, or the impact of repeated measures on time to event, joint modeling is an approach that efficiently uses the information from both processes with unbiased estimates of parameters. The shared random effects in the joint model represents the relationship between the repeated measures and survival outcomes, and avoid the misspecification of covariates trajectories. Meanwhile, if the subjects have substantial heterogeneity, which are reflected as different response from treatment, the latent class concept can be well utilized and combined with the joint modeling to better characterize the population and identify different patterns of longitudinal and survival trajectories. It also helps to conclude the key covariates which are impactful on treatment effect. Simulation studies and an application of Terry Beirn Community Programs for Clinical Research on AIDS study (CPCRA) are provided to illustrate the performance of the latent class joint modeling. Asymptotic properties are briefly summarized in the presentation.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.