Abstract
This communication sets forth and illustratés a method to isolate local changes in cancer mortality rates from national and regional components. The method is suggested as a means of identifying noteworthy places for case control and other local studies from large data bases with temporal and geographical dimensions.