Grant: Operating Grant

Canada is internationally renowned for its large administrative databases and their value for describing the health of the nation's population. Databases that contain anonymized records of physician services are particularly important for studying chronic diseases, like diabetes, because they contain diagnostic information. However, the number of physicians being paid using the conventional fee-for-service method is decreasing, which reduces the amount of physician service information captured in administrative databases. This results in gaps in the completeness of information about the health of Canada's population and downward bias in estimates of the number of persons with chronic diseases. Our research will develop and compare different methods to estimate the completeness of physician databases in Canadian provinces and territories and apply them to study under-estimation for three chronic diseases (diabetes, hypertension, and bowel diseases) in four provinces, including Manitoba, Saskatchewan, Newfoundland and Labrador, and Nova Scotia. Computer simulations will also be used to compare the efficiency and performance of these methods. Recommendations will be developed about choosing a method for measuring the quality of physician databases. This research will help data managers and users across Canada focus on opportunities improve data quality. Ultimately, this research will lead to better information about the health of Canada's population.