Picture two people booked into the same memory clinic on the same Tuesday. The first is 68, noticed her word-finding slipping two years ago, has a spouse who keeps a diary of it, a clean vascular history and a story so textbook a resident could recite it back. The second is 74, has a tremor that came before the forgetting, two falls, a night of vivid confusion after surgery, three specialists who each own a piece of him and no one who owns the whole. Both are waiting for the same appointment slot, with the same scanner, the same neuropsychology battery, the same senior clinician. Only one of them actually needs all of it. That mismatch, repeated across thousands of clinic diaries, is the target of a Viewpoint published this year in JAMA Neurology.
The piece argues for a system-level redesign of how Alzheimer disease is diagnosed and treated: memory clinic services differentiated by purpose, so that patient needs and complexity are matched with the appropriate level of oversight. Put in plainer terms, not every case should be pushed down the same specialist pipeline, and the scarcest expertise should be spent where the diagnosis is genuinely hard. It is worth being precise about what kind of document this is, because the register matters. A Viewpoint is an argument. It reports no patients, no randomisation, no follow-up. Nobody has run a trial in which half the country's clinics are tiered and half are not, then counted who got diagnosed faster or better. What is on offer is a proposal, made in a peer-reviewed journal, about how care should be organised.
The pressure behind the proposal is real and measurable, and here the numbers are somebody else's. A retrospective analysis published in JAMA Neurology in 2025 looked at consecutive patients at the Washington University Memory Diagnostic Center, an outpatient specialty memory clinic in St Louis, and found that a single clinic initiated lecanemab in 234 patients over 14 months, between August 2023 and October 2024. The authors reported that the frequency of significant adverse events, including amyloid-related imaging abnormalities, was manageable, and framed their findings as something to inform conversations about risk. Read that as a workload statement as much as a safety one. Every one of those 234 people needed eligibility work, counselling, monitoring and a clinician's attention, inside a building that also has to see everyone else.
So what would tiering mean at the level of a kitchen table? For the woman with the textbook story, potentially a shorter, less elaborate route to an answer, run by a team that does not need to convene a case conference to recognise what it is looking at. For the man with the tremor, the falls and the delirium, potentially the opposite: a slot that is not being consumed by straightforward cases, with the time and the seniority his picture demands. The appeal of the idea is that it treats clinical complexity as the thing that should determine where you are seen, rather than the alphabetical accident of which referral letter arrived first.
The hazards are equally easy to name, and the Viewpoint's own summary does not resolve them. Somebody has to decide, early and often on thin information, which patient is simple. Atypical Alzheimer disease, Lewy body pathology, frontotemporal presentations in younger people and mixed disease all have a habit of looking straightforward at the front door and complicated three appointments in. A triage system is only as good as its ability to send someone back up a level without penalty or delay, and no published evidence yet shows how well such a system would do that in practice. Nor does the Viewpoint report what tiering does to waiting times, diagnostic accuracy or the experience of families, because it has not been tested.
For readers, the useful takeaway is not a decision to make. It is a vocabulary. When a clinic explains that your appointment is with a nurse-led service rather than a consultant, or that a scan will not be ordered yet, that is the same debate arriving in your letterbox. The questions worth putting to your own clinician are about the route, not the drugs: what happens if the picture turns out not to be typical, who reviews that, and how quickly.