|
Activity Number:
|
192
- Study of Health Outcomes Using Large Cohort Data
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #309759
|
|
|
Title:
|
Weight Calibration to Improve the Efficiency of Pure Risk Estimates from Case-Control Samples Nested in a Cohort
|
|
Author(s):
|
Yei Eun Shin* and Ruth Pfeiffer and Barry Graubard and Mitchell H. Gail
|
|
Companies:
|
National Cancer Institute and National Cancer Institute and National Cancer Institute and National Cancer Institute
|
|
Keywords:
|
Aalen additive hazards model;
Cox proportional hazards model;
Influence functions;
Nested case-control designs;
Pure absolute risk;
Weight calibration
|
|
Abstract:
|
Nested case-control designs measure expensive covariates on all cases and the controls sampled from a cohort. Such data can be used to estimate covariate-specific pure risk. Standard analyses only use the case-control samples, but information available on the entire cohort may be used to improve the efficiency of estimates. We use weight calibration that incorporates such cohort information. This approach has been used to improve estimates of relative risk under the Cox proportional model and excess risk under the Aalen additive model. We extended weight calibration to improve pure risk estimates by incorporating both covariate and time-on-study information from the entire cohort. Moreover, we derived variance formulas for weight-calibrated estimates that account for the two-phase case-control sampling from a cohort and random sampling from an infinite population. Simulations show improvement in precision from using weight calibration and confirm the validity of variance estimators. Examples are provided using data from the NIH-AARP Diet and Health Cohort Study to estimate the association of pancreatic cancer risk with waist circumference.
|
Authors who are presenting talks have a * after their name.
