Weight calibration to improve efficiency for estimating pure risks from the additive hazards model with the nested...

Area of Science:

• Biostatistics
• Epidemiology
• Survival Analysis

Background:

• Estimating pure risk from survival data is crucial in medical research.
• Nested case-control studies offer efficiency but may have limited covariate data.
• Semiparametric additive hazards models are used for risk prediction.

Purpose of the Study:

• To improve the efficiency of covariate-specific pure risk estimates.
• To address challenges with incomplete covariate data in nested case-control samples.
• To enhance existing methods for survival data analysis.

Main Methods:

• Utilized a semiparametric additive hazards model.
• Applied weight calibration using full cohort data to nested case-control samples.
• Developed variance formulas based on influence functions for novel estimates.

Main Results:

• Weight calibration significantly improved the precision of pure risk estimates.
• Demonstrated consistency of variance estimators and validity of asymptotic inference.
• The proposed method is effective for both time-varying and time-invariant covariate coefficients.

Conclusions:

• Weight calibration offers a more efficient approach for pure risk estimation in nested case-control studies.
• The developed methods provide reliable variance estimation and inference.
• The approach was validated using data from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Study (PLCO).