Comparing Indigenous health status across regions: a numerical example of uncertainty

Abstract

Abstract Objective: To illustrate how regional variation in data quality could explain some or all of the apparent regional differences in the health status of Indigenous Australians. Methods: A series of simple hypothetical numerical examples is provided, with varying assumptions regarding the accuracy of identification of Indigenous deaths. Results: The apparent difference in Indigenous mortality in remote compared with urban areas is of a magnitude that could be explained by relatively modest regional differences in data quality. Conclusion and implications: Determinations of relative health status within the Indigenous population must take into account the impact of variability in data quality. Accepted: October 2002