Life tables are seldom derived at the local level, despite a shift toward health service planning to that level. Life tables were calculated by sex for the 42 public health units in Ontario, using 1988 to 1992 mortality files. Traditional methods of life table construction were compared and validated. Data quality, particularly geographical coding of death certificates, poses the greatest difficulty in deriving accurate life tables for comparison between areas. Migration will affect estimates, but it is best considered during the interpretation of results. Except for the final age interval, methods of modelling life tables have little impact on final life expectancy estimates. It is feasible to calculate local level life tables with data and tools that are readily available. The results highlight the importance of examining such life tables, as variations within a province in life expectancy at birth may be as important as the differences between provinces.
View full text