Development of Prefecture-specific Life Tables for Estimating Net Survival Rates

Abstract Calculating and comparing survival rate indicators across different population groups is essential for identifying disparities in treatment outcomes and understanding their underlying causes. Continuous monitoring in this regard is also crucial. In Japan, comparisons of net survival rates are often conducted at the prefectural level. The calculation of net survival rates requires life tables for the general population that share similar characteristics with the cancer patient cohort. However, in current practice, a common national life table is used across all prefectures in Japan, which may introduce bias in estimating true prefecture-specific net survival rates. To address this issue, the present study developed prefecture-specific life tables using a mathematical modeling approach that assumes a Poisson regression model for mortality counts. This method was designed to overcome the instability of estimates caused by limited data in certain regions. Our evaluation showed that the life tables constructed using this model not only accurately reflected average life expectancy across prefectures, but also reduced variability compared to conventional methods. These findings suggest that the proposed approach can be effectively applied to calculate net survival rates for smaller geographic units or populations stratified by socioeconomic status. It thus holds promise for enabling more flexible and context-sensitive cancer control strategies. Key messages • Using a Poisson model, we developed stable, prefecture-specific life tables that improve the accuracy of net survival estimates over a single national life table. • This method supports region-specific and fair cancer control by enabling survival analysis that reflects local and socioeconomic differences.