Advisory Council January 2020 Meeting Presentation: Methods for Identifying Dementia

Helen Lamont, Ph.D.
Office of Disability, Aging, & Long-Term Care Policy
Office of the Assistant Secretary for Planning & Evaluation

Strategy 2.G: Ensure a timely and accurate diagnosis
Goal 5: Improve data to track progress
- Identify reliable sources of data for research and surveillance

Ask "Has a doctor ever told you that you have dementia?"
Assess executive functioning, memory, orientation
- Sometimes more in-depth clinical assessments
Administrative data- Medicare claims
- Already being collected
- Less vulnerable to recall issues
- Person-level
- Publicly available
- Inexpensive

Appropriate diagnosis of dementia is necessary to ensure that people receive proper treatment and support.
Many cases of dementia go undiagnosed or may be misidentified.
Some people with dementia are not told of their diagnosis, and so may not obtain appropriate treatments and supports.
- Healthy People 2020/2030 Objective: Improve awareness of dementia diagnosis
A recent study found that, among people with probable dementia, 39% were undiagnosed and 19% were unaware of their diagnosis (Amjad et al., 2018)
Understanding the prevalence and characteristics of older adults with dementia, and how those features vary by receipt and awareness of a diagnosis, can help guide the development of appropriate policies and programs to provide the necessary supports.

What proportion of older adults who are classified as probable dementia based on functional tests have a Medicare claim with a dementia diagnosis code within the last 3 years? (missed diagnosis)
What proportion of older adults who show indications of dementia on functional test report they have been diagnosed with dementia? (missed diagnosis or not told about diagnosis)
What proportion of older adults who have a Medicare claim with a dementia diagnosis code in the last 3 years report that they have been diagnosed with dementia? (not told about diagnosis)

2015 National Health and Aging Trends Study (NHATS): Annual survey of individuals aged 65 and older who are enrolled in Medicare (~8,000).
Proxy respondents were used when the sample person was unable to serve as a respondent (N=483).
About 40% of Medicare beneficiaries are in a Medicare Advantage plan. The claims information for MA were not available for these analyses. We only included people who had a full 3 years of claims data so these analyses include only 47.4% of the original sample. People in MA plans were less likely (6%) to be identified with dementia based on reported diagnosis or functional assessment than the entire 65+ population (8.5%)

Ask
- Reported diagnosis of dementia: Survey participant or their proxy reported that a physician has ever said they have dementia or Alzheimer's disease.
Assess
- Functional assessment: Assessed survey participant's cognitive abilities in three domains: cognition, orientation, and executive function and classified as having probable dementia, possible dementia, or no dementia (Kasper et al., 2013). Proxies completed the AD8. Used "probable dementia" for these analyses.
Admin data
- Claims-based diagnosis: Medicare claims data from 2012-2014 based on one or more diagnosis codes in inpatient, home health, skilled nursing facility, hospice, or Carrier claims. Methods consistent with CCW.

About 14.8% of community-dwelling older adults were identified as having dementia by one or more methods.
- 9.1% were identified based on functional assessment
- 7.1% were identified based on reported diagnosis
- 10.6% were identified based on Medicare claims
Only 4.3% were identified as having dementia by all three methods.

Approximately 9.1% met criteria for probable dementia based on functional assessment
- Of those 63.4% had a relevant Medicare claim for Alzheimer's disease or dementia in the last 3 years
- So...more than 1/3 of people with probable dementia did not have a dementia claim in the last 3 years
Implications: Large population of older adults may have dementia but have not been diagnosed

Approximately 9.1% met criteria for probable dementia based on functional assessment
- Of those 54.3% reported they were told by a physician they had Alzheimer's disease or dementia
- Almost half (45.7%) reported they were not told
Implications: Large population of older adults may have dementia but have not been diagnosed OR if they were diagnosed, the physician did not tell individual (or proxy) about this diagnosis

Of those with 3 years of Medicare claims, 10.6% had a claim indicating dementia
- Of these, 53.2% reported they had been told of diagnosis
- 46.8% were not told of a diagnosis
BUT...
- The vast majority (71.2%) of people who had a diagnosis in claims data, but who were not told they have dementia, did not show probable cognitive impairment on functional assessments
Implications: Erroneous dementia claims, inclusion of dementia claims for subjective complaints, doctors identify cognitive impairment early OR functional assessments are not sensitive enough

Current methods are imperfect
- "Has a doctor ever told you that you have dementia" may be affected by memory, proxy awareness, differential diagnosis
- Medicare claim indicating dementia may be error, result of minor subjective complaints, temporary conditions that resolve
- Functional assessments through survey like NHATS may not be sensitive enough
Future research opportunities:
- Longitudinal follow-up
- Comparison to other chronic conditions
We must do a better job getting people a diagnosis and improve data so we can track progress