Abstract:
|
Cancer surveillance research often begins with a 5-year tabulated Lexis diagram of cancer incidence derived from cancer registry and census data. This smoothing approach suffers from a significant limitation; important details useful in studying time trends may be lost in generating the summary rate. This study utilizes a Poisson model to describe the relationship between the number of new cases, the number of people at risk, and a smoothly varying incidence rate for the study of the incidence rate function. We propose a Bayesian approach to construct the posterior distribution of smoothed Lexis diagrams for the study of the effects of age, period, and cohort on incidence rates in terms of straight-forward graphical displays. These include the age-specific rates by year of birth, age-specific rates by year of diagnosis, year-specific rates by age of diagnosis, cohort-specific rates by age of diagnosis and annual percent change of incidence rate. Simulation studies indicate that this Bayesian approach outperforms standard Lexis diagram in terms of incidence rate estimation. We apply this method to compare the time trends in breast cancer incidence in Taiwan and those in the US.
|