Dementia Diagnosis Underreporting and Care Engagement and Planning Among Older Adults.

Importance: Inadequate disclosure and stigma often leave patients with a documented dementia diagnosis unaware of their condition. Resultant patient underreporting creates a major barrier to the benefits of detection; however, the extent of underreporting, its associated factors, and care implications remain understudied. Objective: To compare underreporting of dementia diagnosis with that of other diagnosed conditions among persons with probable dementia (PWPD), identify associated patient and care delivery factors, and examine postdiagnosis care engagement and planning. Design, Setting, and Participants: This population-based cohort study used Medicare-linked survey data from 1998 to 2020. US adults aged 65 years or older in the Health and Retirement Study (HRS) with both HRS-defined probable dementia and a claims-based clinical diagnosis of dementia were included. Main Outcomes and Measures: Underreporting of dementia diagnosis, defined as not reporting dementia or memory-related disease despite ever being classified as having dementia in the HRS and having a dementia diagnosis in Medicare claims, estimated using generalized estimating equation regressions and odds ratios (ORs). Results: Among 6158 person-waves with a claims-based dementia diagnosis from 3278 PWPD (3134 [50.9%] aged 65-84 years; 4282 [69.5%] female; 1176 [19.1%] non-Hispanic Black, 617 [10.0%] Hispanic, and 4256 [69.1%] non-Hispanic White), the adjusted proportion not reporting their diagnosis was 42% (95% CI, 40%-43%) among all respondents and 67% (95% CI, 64%-69%) among self-respondents. This underreporting far exceeded that of other diagnosed conditions (overall mean, 31%; arthritis: 17%; 95% CI, 15%-18%; hypertension: 23%; 95% CI, 21%-24%; diabetes: 39%; 95% CI, 37%-41%; depression: 46%; 95% CI, 43%-48%). Dementia underreporting was highest before HRS-defined onset (82%; 95% CI, 76%-87%) and remained high after onset. Patient and care delivery factors were associated with underreporting, such as living alone (OR, 1.65; 95% CI, 1.27-2.13) and Medicare Advantage enrollment (OR, 0.68; 95% CI, 0.52-0.89). Patients who underreported were less likely to have a postdiagnosis problem-based visit (OR, 0.70; 95% CI, 0.55-0.90), receive an influenza vaccination (OR, 0.63; 95% CI, 0.51-0.77), or have a witnessed will or trust (OR, 0.70; 95% CI, 0.54-0.90) within 12 months of diagnosis. Conclusions and Relevance: In this cohort study of PWPD, underreporting of dementia diagnosis far exceeded that of other conditions and was associated with reduced postdiagnosis care engagement, underscoring the need to improve diagnostic disclosure to realize the benefits of early detection.