Researchers at Yale linked the Health and Retirement Study to Medicare claims covering 1998 to 2020, and compared what was written in the clinical record with what people said about themselves.
Among respondents with probable dementia and a claims-based diagnosis, 67 per cent did not report having been diagnosed.
That is more than double the underreporting rate for other conditions in the same population. For hypertension it was 23 per cent. For arthritis 17 per cent. For diabetes 39 per cent, and for depression 46 per cent. The gap is specific to this diagnosis rather than being a general feature of self-reported health.
The pattern of who underreports points at the mechanism. It was more common among people who lived alone, those who were dually eligible for Medicare and Medicaid, those with less education, and people diagnosed within the hospital system. It was less common among Medicare Advantage enrollees, people seen by a dementia specialist, and people diagnosed in an outpatient setting. Lead author Xi Chen’s reading is that those factors are about disclosure practices rather than patient preferences, and that many patients were either never clearly told or never fully understood, while some who were told found it too difficult to accept or to share. A neurological feature of the condition itself, anosognosia, reduces awareness of one’s own illness and accounts for part of it.
Underreporting was highest in the early stage, which is precisely when decision-making capacity is greatest, and it predicted lower engagement with care afterwards and a lower likelihood of setting up a will or trust.
Chen’s summary of the implication is the line worth keeping. Improving detection alone is insufficient. Ensuring patients understand the diagnosis is equally important.
That sentence belongs on a wall in almost every organisation.
Enormous effort goes into detection in business. The dashboard, the customer survey, the exit interview, the quarterly review, the mystery shopper report, the staff satisfaction score. All of it exists to find out something true. Far less effort goes into the second half, which is making sure the finding actually lands with the person who can act on it, in language they understand, at a moment when they still have room to respond.
The failure modes are identical to the clinical ones. The finding was communicated in a setting designed for throughput rather than comprehension. It was delivered to someone with nobody around them to help absorb it. It was phrased in the vocabulary of the people who produced it. Or it was perfectly clear and simply too unwelcome to be taken in.
A survey that identifies a problem nobody internalised has not identified anything. It has produced a document.
The Yale team’s phrasing, from postdoctoral associate Yuting Qian, is that a diagnosis does its work when it is clearly communicated and followed by real support.
The record was accurate. Two thirds of the people in it did not know what it said.
This is a sensitive subject, and anyone concerned about memory or cognition for themselves or a family member is best served by speaking with a doctor.
Source: Connecticut Public, “Many patients diagnosed with dementia don’t report having it, Yale study finds”
