Abstract
There is evidence health utilization increases after incident dementia, particularly toward the end of life. However, less is known about utilization in the years before dementia. Our study objectives were to compare outpatient emergency department (ED) and inpatient hospital utilization in the six years preceding incident dementia compared to a reference group without dementia. We obtained data on n=5,547 Beneficiaries from the Health and Retirement Study-Medicare linked sample, and defined dementia using a validated algorithm. Those with (n=1,241) and without (n=4,306) dementia were balanced on confounders using inverse probability weighting applied to longitudinal Generalized Estimating Equation models. We found persons with dementia had greater odds of ED (OR=1.46, 95% CI: 1.21, 1.77) and inpatient hospital (OR=1.35, 95% CI: 1.12, 1.63) usage in the years preceding dementia compared to those without dementia across a comparable timespan. This study provides evidence to suggest greater healthcare burden may exist before manifestation of dementia.