Standardizing Discrete-Time Hazard Ratios With a Disease Risk Score

Area of Science:

• Epidemiology
• Biostatistics
• Longitudinal Data Analysis

Background:

• The disease risk score (DRS) summarizes multiple covariates but has not been applied to survival or hazard function estimation.
• Standardizing exposure effects in cohort data often requires handling numerous confounding variables.

Purpose of the Study:

• To propose and evaluate a method for standardizing hazard ratios using the DRS in longitudinal cohort studies.
• To provide a tool for estimating the effect of a binary exposure on an outcome while controlling for a large set of covariates.

Main Methods:

• Developed a model-based standardization approach using the disease risk score (DRS).
• Applied the method to longitudinal analyses for binary exposure-outcome associations.
• Utilized simulation studies and an empirical example for validation.

Main Results:

• The proposed DRS method allows for the standardization of hazard ratios.
• This approach effectively manages a large number of covariates in survival analyses.
• Demonstrated utility in settings where exposure propensity score modeling is challenging.

Conclusions:

• The disease risk score offers a valuable method for standardizing hazard ratios in cohort studies.
• This approach facilitates robust estimation of exposure effects on survival.
• The method is particularly useful for complex epidemiological data with many covariates.