Back to landing page The person’s report SAMPLE · CLINICIAN The decision support a signed-in clinician sees, as a printed report. Same person and the same sitting as the M3 Dementia Review.
M3 Dementia Review | m3mentalhealth.org/dementia
Confidential · clinician copy

Decision support — in three steps

What both axes measured this sitting, the options and why, and what the next sitting decides. Information for the conversation with the person and the prescriber. It does not diagnose, it does not stage anyone, and it recommends starting, changing or stopping nothing.

Clinician only · not shown to the person or on their report
Name: Sample Report
Age: 72
Birthdate: 01/15/1954
Sitting: 3 of 3 · 8/17/2026
Prior sittings: 3/2/2026, 5/25/2026
Cadence: quarterly (13 weeks)
This sitting in one line. IQCODE 3.38 (Decline warranting evaluation) and IADL-C 1.36 (At or above the MCI boundary) both sit just over their published lines and have not moved across three sittings; the M3 total has fallen to 29 (Mild), below the line of 33, and pain to 3. Gateway questions: negative. Reading: Q1. Next sitting due November 16, 2026.
Four-Line Need Count 2 of 4 lines — cognition and function
Cognition and function: 2 of 2 · steady since 5/25/2026
IQCODE 3.38 · line 3.30 · 0.08 above over
IADL-C 1.36 · line 1.29 · 0.07 above over
Mood and pain: 0 of 2 · better since 5/25/2026
M3 29 · line 33 · 4 points below (about 12%) under
DVPRS pain 3 · line 5 · 2 points below under
Gateway questions (Q5, Q24–27): negative, no flag. Shown on its own, never part of the count.
A count, not a score: each instrument is read against its own published line, and the four are never added or averaged. Information for the conversation with the person — never a recommendation to start, change or stop a treatment. The prescriber decides.

1–3The three steps, as the clinician sees them on screen

Decision support — in three steps

Built from the same engines that produce the printed report: the M3 findings on the mental-health axis, and the crossing engine on the cognitive axis. Information for a conversation with a prescriber — never a recommendation to start, change or stop any medicine, and never a diagnosis. The prescriber decides.
STEP 1 The clinical overview
What both instruments measured this sitting, and what moved.
Cognitive change — IQCODE
3.38
Decline warranting evaluation. Sixteen items answered by an informant about change over ten years. 3.00 is “not much change”; this platform bands at 3.30 and 3.60, the two thresholds the scoring table uses.
Mental health — M3 total
29 of 108
Mild. Twenty-seven self-rated items across depression, anxiety, bipolar spectrum and PTSD. 33 or higher is the published threshold for discussing with a clinician.
Depression
10
Mild
Anxiety
13
Mild
PTSD
5
Mild
Bipolar
3
Low
Pain — DVPRS
3
The DVPRS intensity item, 0–10, current pain. The four supplemental items are listed separately and are not averaged into this.
Gateway questions
Negative
No gateway item is positive this sitting.
Q1 · BOTH STABLE
Nothing is changing. This is the report working — your baseline just got one quarter deeper.
STEP 2 The option, and why
Read as information for the prescriber’s conversation. Nothing here is a recommendation to start, change or stop any medicine, and nothing here is a diagnosis.
Mental-health axis — findings to work up
M3 total 29 — under the line of 33
No mental-health total that needs attention on its own at this sitting.
Depression mild (10) beside an IQCODE of 3.38
Depression is the commonest treatable impostor of decline, and the two often occur together. Treating it and re-measuring is how the record tells one from the other; finding it does not rule the other out.
Treatable causes to work up before, or alongside, the cognitive question. On the dementia path the mental-health medication ladder is not used (M3, 20 September 2026). Information for the conversation — the prescriber decides.
Cognitive axis — what is worth asking about
The early window — worth asking about while it is open
The cognitive axis sits at the published threshold where decline is worth evaluating, documented across dated sittings. Two FDA-approved anti-amyloid treatments are approved for early Alzheimer's disease: lecanemab (Leqembi), including the weekly at-home IQLIK subcutaneous injection, and donanemab (Kisunla). Both require confirmed amyloid, early-stage disease, and a schedule of MRI monitoring for ARIA — temporary brain swelling or micro-bleeding. Whether any of it applies is a question for the physician; the record is what makes the question askable.
The medicines this record puts on the table
No medicine here has been chosen for you. Each one is what the drug is good at, beside what it still cannot do and what it carries. The prescriber decides.
Leqembi · Eisai and Biogen · lecanemab-irmb
Its strength. The 2026 approval makes it the first anti-amyloid therapy that can be given at home across the whole course — a weekly injection of about fifteen seconds with an autoinjector, after a provider supervises the first two doses. The alternative is an infusion every two weeks at a facility. For a family without an infusion centre nearby, that is the difference between possible and not.
What it cannot do, and carries. Slows decline; it does not stop or reverse it. Requires early-stage disease and confirmed amyloid. Carries a boxed warning for ARIA — amyloid-related imaging abnormalities, brain swelling or micro-bleeding — and a schedule of MRI monitoring. APOE ε4 status affects that risk.
Kisunla · Eli Lilly · donanemab-azbt
Its strength. It is the one with a defined end. The FDA dosing instructions state that prescribers can consider stopping treatment once amyloid plaques are removed to minimal levels on PET imaging — so it can be a finite course rather than an indefinite one.
What it cannot do, and carries. Boxed warning for ARIA, which the label states “can be fatal.” People who carry two copies of APOE ε4 are at higher risk, and the label points to testing. Requires confirmed amyloid and MRI monitoring; not for moderate or severe stages.
Zunveyl · Alpha Cognition · benzgalantamine
Its strength. It is the old class made tolerable. Cholinesterase inhibitors help, and people stop taking them because of nausea and vomiting. Bypassing the gastrointestinal nervous system reduces that overstimulation and improves bioavailability — in trials only 2% of treated patients had adverse events from treatment, with no insomnia observed. A medicine someone actually keeps taking is worth more than one they abandon in week three.
What it cannot do, and carries. Symptomatic support. It helps function; it does not change the course of the disease. And cholinesterase inhibitors as a class generally do not help frontotemporal dementia and may worsen it.
The printed report carries three further reference pages: the arc (what comes when, and which medicine window is open), the newer medicines with what each is good at and what it carries, and the care options condition by condition from the book. None of them selects anything for you.
STEP 3 The next step, on the quarterly cycle
Sitting 3 is complete. This is what the next one decides.
Next sitting due
November 16, 2026
13 weeks — both instruments together, one dated record.
Sittings on record
3
Enough for a trend line on both axes.
What would count as movement next quarter
  • The cognitive axis moves if the IQCODE average reaches 3.68 — a rise of 0.30, the published change threshold — or if it crosses into a higher band.
  • The mental-health axis moves if the M3 total or any one of the four category scores moves up a band. Not a point or two — a band.
  • Nothing moving is the good outcome, and it is the one this record is built to be able to show. A flat line across four sittings is evidence, not an empty page.
Between now and then
  • Print this and bring it. It is written to be read across a desk, and dated observations are the thing a clinic cannot generate for itself.
  • Keep the medication list current. Several of the flags on this report fire off it, and a medicine started between sittings is exactly what the next comparison needs to know about.
  • The physician documentation page is included at the back of the printed report. It is written for the conversation about whether further evaluation is warranted.
  • If anything changes sharply, do not wait for the quarter. A sudden change over days or a few weeks is not what a quarterly instrument is for — that is a call to the clinician, and if there is any thought of self-harm, 988.

›The documentation block

Printed because this person is 55 or older and the cognitive axis is at or above 3.30. It is the dated, informant-answered record a clinician needs to order a blood biomarker test against medical necessity, or to refer.

Documentation Block — for the physician

Prepared by the family from two validated instruments. This page documents; it does not request or order anything.
Sample Report, age 72. 3 dated sittings on record; crossed the 3.30 community threshold on March 2, 2026. Most recent IQCODE average 3.38 on August 17, 2026.
Structured informant and self-report instruments document observed cognitive decline meeting the signs-and-symptoms criterion of the FDA-cleared plasma biomarker indication (adults 55+). Instruments and thresholds per Jorm 1994 and Gaynes 2010.
# Sitting date M3 total (0–108) IQCODE avg IQCODE band Quadrant IQCODE respondent
#1 March 2, 2026 45 (Moderate) 3.38 Decline warranting evaluation baseline A family member
#2 May 25, 2026 37 (Moderate) 3.38 Decline warranting evaluation Q1 A family member
#3 August 17, 2026 29 (Mild) 3.38 Decline warranting evaluation Q1 A family member
Clinic cognitive score, if one has been given: — not recorded —
Recorded as reported by the family. This platform does not administer or score any clinic cognitive instrument.
IQCODE bands: ≤3.00 stable · 3.01–3.29 watch · 3.30–3.59 published community threshold (Jorm 1994; Quinn et al., Cochrane 2014) · ≥3.60 published secondary-care threshold. M3 Checklist validated in Gaynes et al., Annals of Family Medicine 2010;8(2):160–169. Screening aids, not diagnostics.

›Discussion of care options

The care options, condition by condition

From Ten Years Early, Chapter 11 — printed in full, newer options beside older ones, with what each still cannot do.
The assessment shows what needs helping; the label says what is approved to help it; and the same assessments, continued after the prescription, audit whether the match is working.
This table is reference, not a recommendation, and no row here has been chosen for you. This platform measures change; it cannot tell which form of dementia is present, and does not try. That is why the whole table is printed. Under each row is one line saying what your own quarterly record does and does not say about it — measurements, not conclusions.
Condition Newer options (2023–2026) Older, established options What medicine cannot yet do
Alzheimer’s — early stage Lecanemab (Leqembi, Eisai/Biogen, 2023; weekly at-home IQLIK injection approved as a starting dose 13 July 2026) and donanemab (Kisunla, Eli Lilly, 2024) — the first drugs that slow the disease itself. Zunveyl (benzgalantamine, Alpha Cognition, 2024), a re-engineered galantamine with far less gastrointestinal upset. Donepezil (1996), rivastigmine, galantamine, memantine (2003) — symptomatic support; they help function without changing the course. Slow ~25–35%, not stop or reverse; require early stage, confirmed amyloid, MRI monitoring.
Your record: IQCODE 3.38 — the published threshold at which decline is worth evaluating, documented across 3 dated sittings. That is the band in which this row's question is worth asking. Stage and confirmed amyloid are clinical determinations; this record documents symptoms, not biology.
Alzheimer’s — moderate and beyond Auvelity (Axsome, approved for this use 30 April 2026) — the first non-antipsychotic approved for agitation; brexpiprazole (Rexulti, 2023) — first approved antipsychotic for agitation, held in reserve. Memantine ± a cholinesterase inhibitor; the older practice of off-label sedating antipsychotics is now discouraged — boxed warnings apply. No disease-modifying option at this stage; the goal shifts to function, comfort, stability.
Your record: no agitation charted this quarter. This row is here for completeness, not because your record points to it.
Vascular dementia No new dementia-specific drug — the news is intensity: modern aggressive blood-pressure, statin, and stroke-prevention regimens. The established armor: antihypertensives, statins, antiplatelets; cholinesterase inhibitors sometimes used off-label for symptoms. Cannot undo existing injury; can genuinely prevent the next step down.
Your record cannot speak to this row. Vascular change is established by imaging and vascular history, and this platform collects neither. Blood pressure, cholesterol and stroke prevention are worth asking about regardless of what the cognitive axis is doing.
Lewy body dementia No new disease-specific approval; specialist-guided options for frightening hallucinations are studied and evolving. Cholinesterase inhibitors — older drugs, often unusually helpful here; melatonin for dream-enactment; levodopa for movement symptoms. Conventional antipsychotics can be dangerous — the diagnosis itself is the safety device.
Your record: no antipsychotic on your medication list this quarter. Worth knowing this row exists before one is ever started.
Frontotemporal dementia No approved disease-modifying drug; genetic-subtype trials are underway. Symptom-targeted older medicines (e.g., SSRIs for behavioral symptoms) plus structure and family support. Cholinesterase inhibitors generally do not help FTD and may worsen it.
Your record cannot distinguish this from the others. If a cholinesterase inhibitor is ever proposed, this row's caution is worth raising first.
Mixed dementia Anti-amyloid therapy possible for the Alzheimer’s component — if amyloid is confirmed and the stage is early. Each component’s established playbook, run simultaneously — vascular armor, symptomatic support. Requires knowing which disease authored which change — which only monitoring provides.
Your record: 3 quarterly sittings on two axes. What this row asks for — knowing which disease authored which change — is exactly what that line is built to show.
The two shaded cells are safety statements, not limitations: conventional antipsychotics can be dangerous in Lewy body dementia, and cholinesterase inhibitors generally do not help frontotemporal dementia and may worsen it. Both are worth raising by name if either form is ever suspected.
This is information for a conversation with the prescriber. It is not a recommendation to start, change, or stop any medicine, and it is not a diagnosis. Every decision here belongs to the prescriber.

§What this report is, and is not

Everything above was produced by the platform’s own engines from this person’s answers: the scoring, the crossing, the findings, the care options and the documentation block. Nothing was written for this sample by hand except the page furniture.

It does not diagnose or stage anyone; it detects no biomarker; it names no dose; and it recommends starting, changing or stopping nothing. On this path the mental-health medication ladder is not used at all: the mental-health axis is shown as findings to work up. A positive gateway answer, including any report of suicidal thoughts, appears on the person’s own screen and report with the 988 pathway, whatever the CDS settings. The prescriber decides.

M3 Dementia Review — clinician decision support · sample · 8/17/2026 · Instruments: IQCODE (Jorm 1994), IADL-C (Schmitter-Edgecombe 2014; boundaries Rahman 2025, shown for example only pending licence), M3 Checklist (Gaynes 2010), DVPRS (Polomano 2016). This is a screening record, not a diagnosis.