REALISTIC POWER SIMULATIONS COMPARE POINT-BASED AND AREA-BASED DISEASE CLUSTER TESTS

One can roughly divide disease cluster tests into area-based (using re gional data) and point-based (using exact locations). We have compared the power of two area-based methods (Moran's I and I(pop), a new met hod) to that of two point-based methods (the Cuzick-Edwards test and G rimson's test), using three realistic simulations of disease (fox rabi es in England, childhood leukaemia in North Humberside, England, and L yme disease in Georgia). The naive belief that point-based methods sho uld be better is not supported: for the complex data simulated here, I (pop) and the Cuzick-Edwards test had higher power than Grimson's met hod or Moran's I. I(pop), capitalizes on high inter-region variabilit y, while Moran's I cannot.