Regional Differences in the Estimation of Influenza Burden in the Elderly: Does Choice of Population Denominator Matter?

Steven A. Cohen, Johns Hopkins Bloomberg School of Public Health
Imran Oomer, Tufts University
Elena Naumova, Tufts University

Proper estimation of a population denominator is critical to many longitudinal epidemiological studies, though many researchers often overlook this. The objective of this study is to determine the sensitivity of age-specific influenza-related hospitalization rates in the elderly to the choice of four population denominators: decennial census, linearly interpolated from decennial censuses, intercensal estimates, and the Centers for Medicare and Medicaid Services (CMS). We abstracted 14 million hospitalization claims from the Centers for Medicare and Medicaid Services for the period 1992-2004 by age and influenza season and applied each of the three denominators to the total hospitalizations by census region, state, and race. We found large discrepancies in the rates using the difference denominators, especially comparing Census 2000 to intercensal estimates. These discrepancies differed by region. Our findings underscore the need to account for population dynamics that may play a role in spatiotemporal disease distribution in longitudinal studies.