Measuring disease burden in the older population using the slope-intercept method for population log-linear estimation (SIMPLE)

Abstract

Estimating disease burden in the older population can be problematic, due to a dearth of measurements that take into account population dynamics, small population sizes, and age-related disease distribution issues. Age itself explains a substantial amount of the variability in population disease rates. However, in many common techniques to account for age, such as age standardization and age categorization, age is treated as a nuisance parameter. In this paper, we present a method, the slope-intercept method for population log-linear estimation (SIMPLE), to assess disease burden in the Medicare population of the US. We demonstrate the utility and potential limitations of this straightforward and crude method in assessing age-related morbidity, mortality, and case-fatality on multiple geographic levels. We highlight several examples of when this measure is most applicable using examples abstracted from a comprehensive administrative database of hospitalizations in older adults. Traditional measurements of disease burden are compared with the measurements extracted from this modeling method for comparison purposes. We also present spatial and temporal associations between the two measurements the SIMPLE method produces.