PATTERNS AND DISPARITIES IN FORMAL DEMENTIA DIAGNOSIS AND AWARENESS OF DIAGNOSIS

Abstract

Background: Despite disease prevalence, about half of older adults with dementia are undiagnosed. Both formal diagnosis and patient/ family awareness of the diagnosis may impact medical care, decision- making, and future planning. Our objective was to examine characteristics of persons undiagnosed or unaware of dementia. Methods: We linked retrospective Medicare claims to the baseline participant or proxy interview of the National Health and Aging Trends Study (NHATS), a nationally representative sample of Medicare beneficiaries age≥65 in the United States. We examined claims-based diagnosis, awareness of dementia (self or proxy report of physician diagnosis), and participant characteristics (demographic, social/behavioral, functional, cognitive, medical and healthcare utilization characteristics). Among community-dwelling participants, 13.6% (unweighted n=1,038) met validated NHATS criteria for probable dementia. We analyzed 585 individuals with probable dementia (weighted n=1,972,534) and 3 years of prior Medicare expenses, comparing characteristics of undiagnosed versus diagnosed persons and then unaware versus aware among those diagnosed. Results: In weighted analysis, 379,479 people (19.2%) had a claims diagnosis of dementia but did not report physician diagnosis (diagnosed but unaware); 814,551 (41.3%) had both claims and reported physician diagnosis (diagnosed and aware); 778,504 (39.5%) neither had a claims diagnosis nor reported diagnosis (undiagnosed). Males, younger participants, non-whites and less than high school educated were more likely to be undiagnosed. Median income was greatest in those diagnosed and aware (p < 0.05 for unadjusted between group differences). The association between race and diagnosis was no longer present after adjusting for socioeconomic status. The odds of being undiagnosed versus diagnosed were significantly higher among persons who went alone to doctor's visits, had less mean IADL impairments, or were able to manage medications. These same three behavioral and functional characteristics were associated with greater odds of being unaware versus aware among people diagnosed (p < 0.05 for all adjusted odds ratios). Hospitalization, ER visit, and mean ambulatory visits in the past year were not associated with either diagnosis or awareness. Conclusions: Dementia diagnosis and awareness may be improved by encouraging family/friends to attend doctor's visits and taking detailed functional histories. Findings suggest that improving diagnosis and awareness does not increase healthcare utilization.