Why do we continue to use standardized mortality ratios for small area comparisons

摘要: Public health practitioners are often faced with the necessity to compare mortality experience of different geographical areas. Indirect standardization, producing a 'standardized ratio' (SMR) is most commonly used technique for doing this. However, as we show, indirect standardization inappropriate such comparisons, SMRs areas have denominators. The fact that usually chosen this type comparison probably based on two beliefs: (1) direct yields only rate rather than more easily interpreted ratio or index; (2) cannot be carried out in many cases because sub-group specific rates groups compared not available or, at least some age classes, upon small numbers completely unreliable. In paper show simple index (the comparative figure) can calculated from directly standardized cases. Using overall electoral wards Sheffield between 1980 and 1987 demonstrate also advantage gained by smaller standard error SMR outweighed bias inherent its construction. We recommend when absolutely necessary, is, rare circumstance data calculation age- sex-specific subgroup study population.