Aging
Identifying Medicare provider characteristics associated with timely dementia diagnosis Bryan James* Joya Bhattacharyya Annie Chen Yi Chen Francine Grodstein Bryan D. James Ziwei Song Emma K. Stapp Ali Moghtaderi Melinda Power
Background: Late or missed diagnosis of dementia is common. Research in other areas of healthcare has shown that patient-provider concordance can influence patient-provider interactions above the influence of patient characteristics alone. Increased specialization and decreasing patient-provider relationships may also contribute, given clinical recognition of cognitive decline is needed for dementia diagnosis. We aim to identify provider characteristics associated with timely dementia diagnosis.
Methods: We used data from 5 harmonized cohorts from the Rush Alzheimer’s Disease Center (RADC), linked Medicare fee-for-service claims data, and the National Plan and Provider Enumeration System (NPPES) database. In this preliminary work, we examined whether provider gender, provider specialty, or patient-provider gender concordance were associated with timely dementia diagnosis among those with cohort-determined dementia using univariate logistic regression.
Results: In our preliminary analysis, we identified 516 participants with cohort-determined dementia; 385 received a timely diagnosis of dementia and 131 received a late diagnosis in the healthcare setting, determined via linkage to Medicare claims. We did not find any significant associations between provider characteristics and timely diagnosis: male provider gender (OR: 0.81, 95% CI: 0.54, 1.21); patient-provider gender concordance (OR: 1.07, 95% CI: 0.71, 1.61); or specialty of primary care (OR: 0.84, 95% CI: 0.56, 1.25), gerontology (OR: 0.76, 95% CI: 0.23, 2.51), cardiology (OR: 0.80, 95% CI: 0.37, 1.73), and radiology (OR: 1.11, 95% CI: 0.49, 2.52).
Discussion: The provider characteristics investigated in this preliminary analysis do not appear to be central to timely dementia diagnosis, although these findings may change when explored in a larger sample. Expanded analyses will include additional provider characteristics and adjustment for patient characteristics known to be associated with timely diagnosis of dementia.