What's New at Tokei: A More Complete Cognitive Screening
By Mai Shimada, MD, Emergency medicine-trained physician, Founder of Tokei Health
When we started Tokei, the screening was simple: three words to remember, a clock to draw. It was based on the Mini-Cog, a tool that primary care doctors have used for years as a quick way to screen for cognitive impairment in a 5-minute visit.
That tool is good — it is short, validated, and works. But it answers only one question: is there evidence of cognitive impairment right now? It leaves out three other questions that any thoughtful clinician would also ask:
- Are any of this person's medications causing or worsening their symptoms?
- Is something treatable, like sleep apnea, contributing?
- Have there been personality or mood changes that could be the first sign of something happening in the brain?
A few weeks ago I sat in on a session at the American Psychiatric Association annual meeting on how dementia diagnosis has changed. The argument of that session — and the new clinical practice guidelines it referenced — is that a complete cognitive evaluation has to ask all four of those questions, not just the first one. So we updated Tokei to do that.
What's new today
The screening now has four parts.
1. A fuller cognitive battery. The Mini-Cog has been replaced with a cognitive battery based on SLUMS, the Saint Louis University Mental Status exam — a public-domain instrument that covers more ground than the Mini-Cog. The battery tests:
- Orientation (day, year, state)
- Word registration and delayed recall (5 words)
- Math reasoning (a small word problem)
- Verbal fluency (typing as many animals as you can in 60 seconds)
- Working memory (saying number sequences backward)
- Visuospatial ability (the clock drawing — still here — plus identifying shapes)
- Story recall (a short story we ask you to remember)
Scores are on a 30-point scale with education-adjusted cutoffs. About 10 minutes total.
2. Medication review. You enter the prescription drugs, over-the-counter medications, and supplements you currently take. We check each one against a list of about 80 medications known to have anticholinergic effects on the brain — and we show you which of yours are on that list and how much each contributes.
If you take Benadryl, ZzzQuil, or Tylenol PM at night, a bladder medication, or an older antidepressant, you may be surprised. Anticholinergic burden is one of the most underappreciated, most reversible contributors to cognitive symptoms in older adults.
3. Sleep apnea screening. Eight quick yes/no questions — the STOP-Bang questionnaire — to estimate your risk of obstructive sleep apnea. Untreated sleep apnea is a common, treatable cause of fatigue, poor concentration, and what feels like memory problems. It is also a significant risk factor for long-term cognitive decline, especially in middle age.
4. Behavioral changes (the MBI-C). A 34-item checklist that screens for new, persistent changes in motivation, mood, impulse control, social behavior, or beliefs — symptoms that can sometimes precede memory complaints in dementia. You can complete it for yourself, or a family member can complete it about a loved one.
You'll see all four sections in your results, with the raw scores, a plain-language explanation, and a "discuss with your doctor" pointer where appropriate.
Why we made these specific additions
Two of these four parts — the medication review and the sleep apnea screen — exist because the most common reversible causes of cognitive symptoms are not captured by any cognitive test. A patient can score perfectly on the Mini-Cog and still be quietly cognitively impaired because of nightly Benadryl. A patient can score badly on the Mini-Cog and the cause can be a treatable sleep disorder. A good screening tool should at least raise these questions, not just measure the result.
The MBI-C exists because the field's understanding of how dementia begins is changing. We used to assume that memory problems came first, with behavior changing later as the disease progressed. The data now shows that for many people — especially those whose dementia involves frontotemporal pathology, Lewy bodies, or vascular disease — behavior changes come first, sometimes years before any measurable memory problem. The 2024 revised Alzheimer's diagnostic criteria treat persistent unexplained mood and motivation changes in older adults as a transitional stage of the disease. We should be screening for them. Now we do.
The SLUMS-based cognitive battery exists because the Mini-Cog is too brief to characterize what's happening when there is something happening. It can tell you "something is off" or "looks fine." A fuller battery can tell you which cognitive domains are affected — and that pattern is often the first clue about why. Memory-only problems look different from problems in executive function, which look different from language problems. Different patterns suggest different diseases.
What's still the same
- The clock drawing is still there. It is the test we're named after for a reason: it is one of the simplest, most informative tests in cognitive medicine, and our AI-assisted scoring of it has gotten better over the past year.
- You still get a personal results email.
- It's still free.
- It is still not a diagnosis. This is a screening tool — a way to organize information and to flag things worth bringing to a clinician. It does not replace a clinical evaluation.
What's coming
This is Phase A of a larger reorganization. Two things we're working on next:
- A way to upload your recent lab results (B12, thyroid, basic metabolic panel) so the screening can flag laboratory abnormalities that contribute to cognitive symptoms. These are part of the "Tier 1" workup recommended by the 2025 DETeCD-ADRD clinical practice guideline.
- Better follow-up resources — guidance on what to do with each kind of flag, written in plain language, including how to talk with your doctor about a medication change or a sleep study.
A note on privacy: nothing about your screening is sold or shared with advertisers. The current scope is non-clinical screening only; this is not a substitute for medical care.
If you've taken the screening before, you're welcome to take it again. If you're new — start here.
Frequently Asked Questions
- What changed in the Tokei screening?
- Until today, the Tokei screening was a brief memory-plus-clock test (the Mini-Cog). It now covers four areas: a fuller cognitive battery based on the public-domain SLUMS instrument; a medication review that flags common drugs with anticholinergic properties; a STOP-Bang sleep apnea screen; and the MBI-C, a checklist that screens for early behavioral changes that can precede dementia. The clock drawing is still part of the test.
- Why did you make it longer?
- Because diagnosing cognitive decline well requires more than a memory test. Two of the most common reversible causes of memory symptoms — anticholinergic medication burden and untreated sleep apnea — never show up on a memory test. And one of the most important newer concepts in dementia care, Mild Behavioral Impairment, is about behavior, not memory. We wanted Tokei's screening to reflect how a thoughtful clinician actually evaluates someone.
- Is this a diagnosis?
- No. This is a screening tool for informational purposes only. It's designed to give you a structured snapshot — and, more importantly, to flag things worth bringing to your doctor. A high anticholinergic burden score, a high STOP-Bang score, or behavioral changes flagged on the MBI-C should all be conversations with your clinician. None of it is a diagnosis.
- Why SLUMS instead of Mini-Cog or MoCA?
- Two reasons. First, SLUMS is in the public domain — Saint Louis University placed it there explicitly — so we can use it without licensing complications. MMSE is copyrighted by PAR Inc. MoCA requires certification for commercial use. Mini-Cog is free for clinical/research use but its commercial-use terms are unclear. Second, SLUMS covers more ground than Mini-Cog: it tests orientation, attention, math, semantic fluency, working memory, visuospatial skills, and recall, in addition to the clock.
- Is the SLUMS test the same as a clinician-administered SLUMS?
- No, and we want to be honest about that. SLUMS was originally designed to be given by a clinician in person. We adapted it to be self-administered with typing instead of speech. The questions and scoring follow the published SLUMS rubric, but we cannot claim the same sensitivity and specificity that the original validation studies reported. That's a limitation of every self-administered cognitive test, including ours.
- What is the anticholinergic burden score?
- Many common medications block acetylcholine, a brain chemical involved in memory and attention. The total amount of this blocking effect across all of someone's medications is their anticholinergic burden. Higher burden is associated with worse cognition, especially in older adults — and it is often reversible if a prescriber substitutes alternatives. Common culprits include over-the-counter sleep aids (diphenhydramine in Benadryl, ZzzQuil, and Tylenol PM; doxylamine in Unisom), bladder antispasmodics, and some older antidepressants. Our screen uses the revised Anticholinergic Drug Scale (Carnahan 2025) to score about 80 common medications.
- What is the MBI-C and why is it on the screen?
- The Mild Behavioral Impairment Checklist (MBI-C) screens for new, persistent personality, mood, motivation, social, and perceptual changes in adults over 50. The construct of Mild Behavioral Impairment was formalized in 2016, and the 2024 Alzheimer's Association revised diagnostic criteria for Alzheimer's now treat unexplained recent-onset behavioral changes as a transitional stage of the disease. Many people enter dementia through behavior, not memory — and that pathway is under-screened in most current tools. The MBI-C is free for clinical and research use.
- Will my old results still be valid?
- Old results are saved in your record but they were generated by the older Mini-Cog scoring model. If you took the screening before today and want a result on the new scale, you can take it again.
- How long does it take now?
- About 10–12 minutes if you go through it without distractions. The animal-naming task has a 60-second timer, and the MBI-C has 34 items — those are the longest parts.
Take the Clock Drawing Test
A quick, evidence-based screening tool you can take from home in a few minutes.