Yingqi Zhao, of the Biostatistics Program & Public Health Sciences Division at Fred Hutchinson Cancer Research Center, will present:
Abstract: In this talk, I will discuss two new concepts in personalized health care. First, I will introduce a learning algorithm, termed as reinforced risk prediction, which sequentially updates as longitudinal data accumulates on each subject until sufficient information is available to make a high-quality prediction. This algorithm takes into account the cost of delaying the decision to a follow-up time when more information is available, and thus minimizes burden associated with delaying a prediction by collecting information only as needed. We applied the method to the electronic health records data on complex patients with type II diabetes. I will also present a patient-oriented measure of treatment benefit, termed as personalized adjusted chance a longer survival. Such a measure answers a patient’s question: ”For a patient like me, what is the net chance of surviving longer with treatment than without?” or, ”For a patient like me, what is the net chance of surviving at least 6 months longer with treatment than without?’’ And the measure is meaningful regardless of the proportional hazard assumption. We proposed a nonparametric estimator, along with its inference procedure, for the quantity. We demonstrated the proposed method using simulation studies and data examples on cancer clinical trials.
A social tea will be held at 3:00 p.m. in A434 Mayo. All are Welcome.