|
Activity Number:
|
477
- Methods in Clinical Trials
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #311044
|
|
|
Title:
|
Patient Recruitment Using Electronic Health Records: A Two-Phase Sampling Framework
|
|
Author(s):
|
Guanghao Zhang* and Lauren J Beesley and Xu Shi
|
|
Companies:
|
University of Michigan and University of Michigan and University of Michigan
|
|
Keywords:
|
two-phase sampling;
electronic health records;
clinical trials;
patient recruitment;
optimization
|
|
Abstract:
|
Electronic health records (EHRs) are increasingly seen as a cost-effective resource for targeted patient recruitment into a research study. Suppose we want to conduct a study to estimate the mean of an expensive outcome Y in some target population. Inexpensive auxiliary covariates predictive of Y may often be available in the EHR, presenting an opportunity to recruit patients selectively efficiently. We propose an efficient two-phase sampling design that leverages available information on auxiliary covariates in the EHR cohort. A key challenge with using EHR data for multi-phase sampling is the lack of representativeness of the EHR cohort. Extending existing literature on two-phase sampling designs, we derive an optimal two-phase sampling method that improves efficiency over random sampling while accounting for potential selection bias in the EHR sample and the cost of collecting Y. We conduct simulation studies to assess the finite sample performance of our method, and we illustrate our method with an application to patient recruitment from the Michigan Genomics Initiative.
|
Authors who are presenting talks have a * after their name.
