Reply to: “Comment on: Differences in Setting of Initial Dementia Diagnosis Among Fee-for-Service Medicare Beneficiaries”

We appreciate the comments offered by Wu et al. [1] on our paper “Differences in Setting of Initial Dementia Diagnosis among Fee-For-Service Medicare Beneficiaries” [2]. We agree that claims data do not allow for measurement of dementia stage at time of diagnosis. This is one of the reasons why we chose to examine differences in survival, adjusting for patient clinical characteristics, based on setting of diagnosis in this descriptive paper. While hospitalization and nursing home admission are independent risk factors for mortality, our finding of shorter survival among individuals diagnosed with dementia in these settings, versus in the community, also likely reflects individuals being diagnosed at later dementia stages.

We note the comments regarding the utility of biomarkers, neuroimaging, and structured cognitive testing in diagnosing dementia sub-types. We agree—however, the purpose of our study was to understand how individuals in the general population are being diagnosed under real-world conditions since most individuals are not diagnosed in research or specialty settings where these modalities are more often used. The Alzheimer's Association has recently revised its criteria for the diagnosis and staging of Alzheimer's disease to define Alzheimer's disease biologically rather than based on symptom presentation, and encourage the use of biomarkers to identify neuropathological change among asymptomatic individuals [3]. Our findings, in combination with the past literature showing that fewer than one-third of Medicare beneficiaries receive structured cognitive assessment as part of their annual wellness care [4], suggest that our existing health care systems are far behind where they need to be to accommodate this shift in diagnostic approach.

Finally, Wu et al. [1] note the important contribution of socioeconomic factors to dementia diagnosis. Again, we agree—and demonstrate this relationship empirically in our paper [2] (Table 3) with our finding that individuals living in communities with higher social deprivation are significantly more likely to be diagnosed with dementia in a hospital, and less likely to be diagnosed in the community.

Preparation of the manuscript and critical revision of the manuscript for important intellectual content: all authors.

The authors declare no conflicts of interest.