A Bayesian bivariate spatial modeling framework for stage-specific cancer incidence and targeting of screening efforts.

Geographic differences in cancer stage at diagnosis may reflect both variation in underlying cancer risk and variation in the extent of early detection through screening or other diagnostic testing. Mapping both late-stage incidence (as an indicator of disease burden) and late-minus-early difference (as an indicator of insufficient diagnostic activity relative to the underlying disease burden) provides an empirically grounded way to identify small areas (neighborhoods) where enhanced early detection may yield the greatest benefits. We introduce a Bayesian bivariate spatial modeling framework for jointly analyzing early- and late-stage cancer incidence, allowing for both spatial correlation across areas and outcome correlation within areas to improve small-area estimation. Using population-based registry data aggregated at the small-area level, we estimate age-standardized incidence rates of early- and late-stage cancers across the study area and propose an inferential approach to identify priority sub-areas. We compare single-outcome and multivariate spatial formulations via Integrated Nested Laplace Approximation. We demonstrate the approach using Swedish register data on prostate cancer linked to population data at the small area level. The framework can be extended to other diseases with correlated outcomes or multiple severity levels, as well as to other settings.