A simple computer program for generating person-time data in cohort studies involving time-related factors

Area of Science:

• Epidemiology
• Biostatistics
• Statistical Software

Background:

• Grouped data methods like standardized rate ratios and Poisson regression offer advantages for cohort study analysis.
• Historically, the application of these methods has been limited by the complexity of generating person-time data, especially with time-related factors.
• This complexity hindered the widespread adoption of powerful statistical techniques in epidemiological research.

Purpose of the Study:

• To present a straightforward Statistical Analysis System (SAS) program designed to generate person-time data.
• To enable the direct use of this generated data in GLIM for Poisson regression analysis.
• To simplify and enhance the analysis of cohort studies using advanced statistical methods.

Main Methods:

• Development of a simple SAS program for person-time data generation.
• The program outputs data compatible with GLIM software for Poisson regression.
• The methodology accommodates multiple time-related factors and stratification levels without restriction.

Main Results:

• The SAS program successfully generates the required person-time data for Poisson regression analysis.
• The program is flexible, allowing for various numbers of time-related factors and their levels.
• It can be readily adapted for multiple disease outcomes and latency period analyses.

Conclusions:

• The developed SAS program effectively overcomes the challenges of generating person-time data for cohort studies.
• This facilitates the broader application of Poisson regression and grouped data methods in epidemiological research.
• The program's flexibility and ease of use enhance the analysis of time-related factors, disease outcomes, and induction times.