Monitoring with Repeatedly Measured Markers: Application to longitudinal PSA measurements in PLCO
Principal Investigator
Name
Paramita Saha-Chaudhuri
Degrees
Ph.D.
Institution
McGill University
Position Title
Associate Professor
Email
About this CDAS Project
Study
PLCO
(Learn more about this study)
Project ID
PLCO-408
Initial CDAS Request Approval
Oct 12, 2018
Title
Monitoring with Repeatedly Measured Markers: Application to longitudinal PSA measurements in PLCO
Summary
This is a statistical methodologic project for publication in a statistical journal. We request the data as an example to apply our new methodology. We propose a new measure of time-dependent prediction accuracy, called the Overall Detection Rate (ODR), and use the ODR curve to assess incremental value of additional marker measurements. The advantage of ODR is that it fixes the false-positive rate over time, rather than fixing a marker value threshold over time. For markers like PSA that naturally increase with age, any reasonable PSA threshold will be exceeded by nearly everyone who is old enough. However, by fixing the false-positive rate (which means the PSA threshold will implicitly increase with age), we obtain a more reasonable measure of time-dependent prediction accuracy.
Aims
1. Study the statistical properties of the new ODR estimator by extensive simulations.
2. Develop a model for longitudinal PSA data over age from PLCO, and calculate and compare all metrics of time-dependent prediction accuracy, including ODR.
Collaborators
Hormuzd Katki: Division of Cancer Epidemiology and Genetics, NCI/NIH