A0239
Title: Dynamic landmark prediction for mixture data
Authors: Layla Parast - RAND Corporation (United States)
Tanya Garcia - UNC Chapel Hill (United States) [presenting]
Abstract: In kin-cohort studies, clinicians are interested in providing their patients with the most current cumulative risk of death arising from a rare deleterious mutation. Estimating the cumulative risk is difficult when the genetic mutation status in patients is unknown and, instead, only estimated probabilities of a patient having the mutation are available. We estimate the cumulative risk using a novel nonparametric estimator that incorporates covariate information and dynamic landmark prediction. The contributions are three-fold. Our estimator better informs patients of their risk of death, as it yields improved prediction accuracy over existing estimators that ignore covariate information. The estimator is built within a dynamic landmark prediction framework whereby we can obtain personalized dynamic predictions over time. Compared to current standards, a simple transformation of our estimator provides more efficient estimates of marginal distribution functions in settings where patient-specific predictions are not the main goal. We show our estimator is unbiased and has substantial gains in predictive accuracy compared to approaches that ignore covariate information and landmarking. Our method is motivated by and illustrated using data from a Huntington disease study; results illustrate the development of survival prediction curves incorporating gender and familial genetic information, and the creation of personalized dynamic risk trajectories over time.
