The impact of the under-reporting of vital events upon epidemiological and demographic measures of the Manitoba Registered Indian population : an exercise in data quality

Abstract

In order for the various levels of government, the biomedical research community, and Aboriginal leadership to more carefully assess the needs of the Canadian Aboriginal population they must have an accurate picture of its demographic and epidemiological characteristics. Researchers of Aboriginal health have often used various data sources without a full appreciation of the flaws inherent in the data. This thesis examines the effect of the under-reporting of vital events upon one such data source, namely the Indian Register, and subsequent ramifications for the epidemiological and demographic analysis of the Manitoba Status Indian population. The study compares the magnitude of the problem for the aggregate of six bands from 1979 through 1983 with further differentiation into sex, residential and regional categories. Each of these populations was adjusted for the late- and under-reporting of vital events in order to obtain a coresponding set of population data for comparison purposes. The principal methodologies employed include direct and indirect standardization of mortality rates, life table analysis of mortality, and analysis of fertility and reproduction. These analyses reveal a preponderance of both birth and death reporting problems associated with the off-reserve populations although all populations were affected to some degree. Demographic and epidemiological calculations for all populations were affected to an extent depending upon the magnitude of the reporting problems and the age strata in which they were concentrated. Mortality rates tended to be inflated as a result of reporting problems.