Cognitive screening is a basic part of every neurological examination. The tools available — Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Frontal Assessment Battery (FAB), and a handful of others — are not substitutes for formal neuropsychological testing. They are short bedside instruments designed for one purpose: to detect cognitive impairment that might otherwise be missed in clinical encounters. A clinician who knows which instrument to use, what its limitations are, and how to interpret the score in context can change the management of a patient in five to ten minutes.

This page covers the major bedside cognitive screens, what each measures, where each performs well, and what to do when a score is abnormal. The unifying message is that a screening score is a number, not a diagnosis; it tells you that the patient deserves further evaluation, not what the diagnosis is.

What a Screening Tool Is Trying to Do

Cognitive screens are designed to be brief (5-15 minutes), administrable by a non-specialist, and sensitive to the most common dementias. They give a numerical score and an associated cutoff that defines “impaired.” The tradeoff is between sensitivity (catching real cognitive impairment) and specificity (not over-calling normal people as impaired), and different instruments optimize this differently. Cutoffs are influenced by age and education, and a strict numerical cutoff that ignores both will misclassify a substantial fraction of patients.

Three principles guide the interpretation of any screen:

1. Adjust for age and education. A 30/30 MoCA in an 85-year-old with eight years of schooling is impressive; a 27/30 in a 50-year-old with a doctorate is concerning. Most published cutoffs assume an average-educated adult; deviation from that requires interpretation.
2. Use the score as a prompt for further evaluation. An abnormal score initiates a workup; it does not name the disease.
3. The pattern of subscores often matters more than the total. A patient who loses 5 points on memory and 0 elsewhere on the MoCA has a different profile from one who loses 1 point on each of five domains. The shape of the impairment is diagnostic.

The major bedside screens at a glance:

The Mini-Mental State Examination (MMSE)

The MMSE was developed in 1975 and was for decades the standard cognitive screen. It is brief (about 7 minutes), familiar to clinicians, and has accumulated a large evidence base. It tests orientation (10 points), registration and recall of three words (3 + 3 points), serial 7s or spelling “world” backward (5 points), naming of two objects (2 points), repetition of a phrase (1 point), three-stage command (3 points), reading and following a written command (1 point), writing a sentence (1 point), and copying intersecting pentagons (1 point). Total: 30 points.

Conventional cutoffs are:

• 24-30: normal
• 18-23: mild impairment
• 0-17: severe impairment

The MMSE has several well-recognized limitations:

• Poor sensitivity for mild cognitive impairment and early dementia. A patient with early Alzheimer disease may score 28 or 29 of 30 while showing measurable impairment on more sensitive testing.
• Almost no executive function testing. The MMSE measures orientation, language, and memory well but has little to assess frontal/executive function. Patients with frontotemporal dementia, vascular dementia, or other predominantly executive syndromes can score in the normal range while clearly impaired.
• Ceiling effects in highly educated patients. A patient with a college education and early dementia may continue to score normally because the test is too easy.
• Floor effects in moderately advanced disease. A patient with an MMSE of 5 cannot be meaningfully tracked further.
• Copyright restrictions since 2001 have made the MMSE less freely available than alternatives.

Use the MMSE when continuity with prior testing matters (the patient has serial MMSEs for tracking), when the patient is severely impaired and the MoCA would be uninterpretable, and when you need a familiar number for documentation. For initial screening of suspected mild cognitive impairment, the MoCA is generally a better choice.

The Montreal Cognitive Assessment (MoCA)

The MoCA was developed to address the MMSE’s poor sensitivity for mild cognitive impairment. It takes about 10 minutes, is freely available, and includes substantial executive function testing. The 30-point test covers:

• Visuospatial/executive (5 points): a trail-making variant (alternating numbers and letters), copying a cube, drawing a clock.
• Naming (3 points): naming three line drawings of less common animals.
• Memory registration of five words (no points awarded; the points come at delayed recall).
• Attention (6 points): forward and backward digit spans, vigilance task (tapping at the letter A in a sequence), serial 7s.
• Language (3 points): sentence repetition, fluency (words beginning with F in one minute).
• Abstraction (2 points): similarities between two words.
• Delayed recall (5 points): the five words from earlier.
• Orientation (6 points): date, day, month, year, place, city.

The MoCA awards an additional point to patients with 12 or fewer years of education to partially correct for the impact of low education on test performance.

Conventional ranges, used as a rough guide rather than fixed bands:

• 26-30: typically normal
• 18-25: range often associated with mild cognitive impairment
• 10-17: range often associated with moderate impairment
• 0-9: range often associated with severe impairment

The original MoCA validation study used a cutoff of less than 26, but later work has shown that this threshold can produce substantial false positives in older adults and in some lower-education or culturally diverse populations. MoCA scores should therefore be interpreted in clinical context — age, education, language, sensory deficits, baseline function, and the pattern of subscores — rather than used as an absolute diagnostic cutoff in isolation. The severity bands above are useful for orientation, not for fixed classification.

The MoCA is more sensitive than the MMSE for mild cognitive impairment, particularly in executive and visuospatial domains. It is the preferred screen for suspected early Alzheimer disease, frontotemporal dementia, vascular cognitive impairment, and Parkinson disease dementia. Its limitations are similar to the MMSE: ceiling effects in highly educated patients and floor effects in advanced disease.

MoCA Variants for Special Populations

Several adapted versions of the MoCA exist for patients who cannot complete the standard form:

• MoCA-Blind: omits items requiring vision (cube copy, clock drawing, naming of pictures, trail-making, visual abstraction). Retotaled to a 22-point scale. Designed for patients with severe visual impairment.
• MoCA-Basic: developed specifically for patients with low literacy or limited formal education, with simplified items and culturally neutral material. Useful where the standard MoCA over-calls impairment because of education effects.
• MoCA telephone and audio-visual versions: validated for remote administration; useful in telehealth follow-up and rural settings.

Choosing the right variant matters: applying the standard MoCA to a patient with severe low vision or no formal schooling will produce a score that reflects the test conditions more than cognition.

The Frontal Assessment Battery (FAB)

The FAB was designed specifically to detect executive dysfunction that the MMSE misses. It consists of six subtests, each scored 0-3:

1. Similarities (conceptualization): how are a banana and an orange alike?
2. Lexical fluency: words beginning with S in one minute.
3. Motor series (Luria three-step): the patient is shown a fist-edge-palm sequence and asked to continue it independently.
4. Conflicting instructions: the examiner taps once when the patient is to tap twice and vice versa.
5. Go/no-go: tap once when the examiner taps once; do not tap when the examiner taps twice.
6. Prehension behavior: examiner places hand palm-up below the patient’s hands; the abnormal response is to grasp them.

Total: 18 points. A score of 12 or below suggests frontal dysfunction. The FAB is sensitive to behavioral-variant frontotemporal dementia, progressive supranuclear palsy, corticobasal syndrome, and other frontal syndromes. It is a useful complement to the MoCA when a frontal syndrome is suspected.

Ultra-Brief Tools for the Time-Pressured Encounter

The MMSE and MoCA each take roughly ten minutes; the FAB another five. In primary care visits, busy emergency departments, and inpatient rounds, even ten minutes is often not available, and shorter screens have a real role.

Mini-Cog

The Mini-Cog takes about three minutes and has two components:

1. Three-word recall: the patient is given three unrelated words to remember.
2. Clock drawing: while the words are held in mind, the patient draws a clock face with all numbers and sets the hands to a specified time (commonly “ten past eleven”).
3. Three-word recall again: the patient is asked to repeat the three words.

Scoring is simple: each recalled word is 1 point (0-3), and a normal clock is 2 points (abnormal 0). A total of 0-2 suggests cognitive impairment; 3-5 is considered normal. The Mini-Cog is sensitive enough for routine dementia screening and is widely used as the bedside cognitive screen in primary care, Medicare Annual Wellness visits, and emergency department dementia screening.

AD8 — Informant Tool

The AD8 is an eight-item informant questionnaire that takes about two to three minutes and asks a family member or close contact about changes over the past several years in:

1. Problems with judgment.
2. Reduced interest in hobbies or activities.
3. Repetition of questions, stories, or statements.
4. Difficulty learning to use a new tool, appliance, or gadget.
5. Forgetting the correct month or year.
6. Trouble handling complicated financial affairs.
7. Difficulty remembering appointments.
8. Consistent problems with thinking or memory.

Each “yes” is scored 1 point. Two or more suggests cognitive impairment and warrants further evaluation. The AD8 is particularly useful when the patient minimizes symptoms (common in early Alzheimer disease) or when a brief informant perspective adds critical context that bedside testing cannot capture. It complements rather than replaces a direct cognitive screen.

Choosing Between Brief Tools

• The Mini-Cog is best when only the patient is available and time is tight.
• The AD8 is best when a reliable informant is present and the question is functional change rather than current performance.
• Together, the Mini-Cog plus AD8 take about five minutes and provide both a direct cognitive sample and an informant report — a substantially stronger combined signal than either alone.

Other Useful Bedside Tools

Clock Drawing Test

“Draw a clock face. Put in all the numbers. Set the time to ten past eleven.” The clock drawing test is one of the simplest and most informative bedside cognitive tests. It engages executive function (planning), visuospatial function (placement of numbers), comprehension (the instruction), and memory (knowing what a clock looks like). Errors are diagnostically informative:

• Numbers crowded on the right side with the left half empty → left neglect (right parietal disease).
• Conceptual errors — numbers placed wrong, missing numbers, the hands pointing in random directions → executive or visuospatial dysfunction. Common in Alzheimer disease.
• The hands set to 10:10 instead of 11:10 → loss of abstract attention; the patient defaults to the visually familiar position of clock hands in advertisements.
• Perseverative numbers (the patient writes 12, 12, 12) → frontal dysfunction.
• Excellent overall performance in a patient with concerning history → the screen is not sensitive enough; consider formal neuropsychological testing.

Trail-Making Test (Parts A and B)

Trail A: connect numbered circles in order, as fast as possible (1 → 2 → 3 → … → 25). Trail B: connect alternating numbers and letters (1 → A → 2 → B → 3 → C → …). Part A tests visual scanning and motor speed; Part B adds set-shifting, the executive function that is impaired by frontal lobe dysfunction. The ratio of B to A times reveals selective executive impairment.

Verbal Fluency

“How many animals can you name in one minute?” “How many words beginning with F (or A, or S) can you name in one minute?” Animal fluency (semantic) and letter fluency (phonemic) probe different systems. Semantic fluency is more impaired in temporal lobe degeneration (Alzheimer, semantic dementia); phonemic fluency is more impaired in frontal lobe degeneration. Normal values are about 15-20 animals or letters per minute for educated adults.

Digit Span

The patient repeats a sequence of digits forward, then backward. Forward span of 7 ± 2 is the normal range; backward span of 5 ± 1. Reduced span suggests attention and working memory impairment, and is sensitive to delirium and frontal dysfunction.

Choosing the Right Screen for the Question

Interpreting Subscores Diagnostically

The pattern of subscores can suggest specific diagnoses:

• Alzheimer disease: prominent loss on delayed recall, with progressive involvement of orientation, naming, and visuospatial domains. Executive function is impaired later than memory.
• Vascular cognitive impairment: prominent executive dysfunction (slow processing, set-shifting deficits) with relatively preserved memory in the early stages. The pattern is sometimes called “subcortical” and is contrasted with the “cortical” pattern of Alzheimer disease.
• Dementia with Lewy bodies: prominent visuospatial impairment (poor clock drawing, poor cube copy) early, with relative sparing of memory. Attentional fluctuation is also characteristic.
• Frontotemporal dementia (behavioral variant): relatively preserved memory and orientation, with executive deficits, poor abstraction, perseverative responses, and impaired social judgment. Cognitive screens can be deceivingly normal early; the FAB is more sensitive than the MoCA.
• Primary progressive aphasia: language-specific deficits dominate. Memory, executive function, and visuospatial domains may be relatively preserved depending on the variant.
• Parkinson disease dementia: prominent executive dysfunction with bradyphrenia (slowed thinking), visuospatial impairment, and relative preservation of language and remote memory.
• Normal pressure hydrocephalus: subcortical pattern with prominent executive slowing, attentional deficits, and apathy, accompanied clinically by gait apraxia and incontinence.

What to Do When the Score Is Abnormal

An abnormal cognitive screen is the beginning of an evaluation, not the end. A reasonable next-step framework:

1. Rule out reversible causes: hypothyroidism, B12 deficiency, syphilis (in appropriate epidemiological contexts), medications, depression, sleep apnea, alcohol use. These are the standard reversible dementia workup items.
2. Image the brain: MRI when possible. Look for atrophy patterns, vascular disease, mass lesions, hydrocephalus.
3. Consider formal neuropsychological testing: especially when the pattern is unclear, when functional impact is being debated, or when distinguishing depression from dementia.
4. Consider specialist referral: cognitive/behavioral neurology or geriatric psychiatry when the diagnosis is unclear or the management is complex.
5. Discuss the findings with the patient and family: a screening result is best presented as one piece of information about brain function, with appropriate context about its limitations.

Delirium versus Dementia

The single most important distinction to draw before any cognitive screening is between delirium and dementia. The cognitive screen will reflect both, but the interpretation and management are entirely different.

The Confusion Assessment Method (CAM) is the bedside instrument that operationalizes the diagnosis of delirium: acute onset and fluctuating course, plus inattention, plus either disorganized thinking or altered level of consciousness. Any patient with an abnormal cognitive screen and acute change should be evaluated for delirium before further dementia workup, because delirium with a treatable cause is a medical emergency.

The 4AT is a four-item rapid bedside screen — alertness, AMT-4 (age, date of birth, place, current year), attention (months of year backward), and acute change or fluctuating course — that takes under two minutes and is widely used in inpatient and emergency settings as a delirium screen. A score of 4 or above suggests delirium; 1-3 suggests cognitive impairment without delirium; 0 is normal. Critically, the 4AT is a delirium screen, not a dementia screen: a positive 4AT prompts a workup for acute causes (infection, medication, metabolic, withdrawal, retention, hypoxia, structural), while a positive MoCA or Mini-Cog without acute change points toward dementia evaluation. The two questions answer different questions, and the tools are not interchangeable.

Always integrate informant history and functional status. A bedside score interpreted without information about the patient’s premorbid function, prior occupation, daily activities, recent changes, and the perspective of someone who knows them well is a number without context. Functional decline (managing finances, medications, appointments, driving, cooking, social engagements) often precedes objectively measurable cognitive decline and is what ultimately matters to the patient.

Pitfalls and Pearls

• The MMSE misses mild cognitive impairment. Use the MoCA for early screening.
• Both miss frontal disease. Add the FAB or specific frontal tests (clock drawing, fluency, Luria three-step) when frontal symptoms are suspected.
• Adjust for education and age. A 27/30 MoCA in a college-educated 60-year-old is concerning; a 23/30 in an 88-year-old with six years of schooling may be appropriate.
• The pattern of impairment matters more than the total. Look at which subscores are lost — memory, executive, visuospatial, language — to begin a differential.
• Acute change is delirium until proven otherwise. Always assess for inattention and fluctuating course. A patient with an abnormal MoCA who was reading the newspaper a week ago is delirious.
• Depression mimics dementia. The patient who answers “I don’t know” to many questions, who shows good attention but poor effort, who improves with prompting — these are clues to depression. Treat the depression and rescreen.
• The screen does not name the disease. An abnormal MoCA in a 75-year-old prompts a workup; it does not, by itself, diagnose Alzheimer disease.
• Always tell the patient what the score means. Patients are often told their MoCA score without explanation; many leave the visit believing they have been “diagnosed with dementia” by a number. The score is one piece of information; the diagnosis requires more.
• Document subscores, not just the total. “MoCA 22/30 with 5/30 lost on delayed recall and 2/5 lost on executive” is far more informative than “MoCA 22.”
• The original MoCA cutoff of less than 26 is not a universal truth. It produces substantial false positives in older adults and in lower-education or culturally diverse populations. Interpret in context.
• Use the right MoCA variant. MoCA-Blind for severe visual impairment; MoCA-Basic for low literacy or limited formal education.
• Mini-Cog plus AD8 takes about five minutes and pairs a direct cognitive sample with an informant report — a strong combined signal when time is short.
• 4AT is a delirium screen, not a dementia screen. Positive 4AT prompts a delirium workup (infection, medication, metabolic, withdrawal); positive MoCA without acute change prompts dementia evaluation.
• Informant history and functional decline are essential. A bedside score interpreted without context — premorbid function, daily activities, recent changes, the perspective of someone who knows the patient — is a number, not a diagnosis.

References

1. Campbell WW. DeJong’s The Neurologic Examination. 7th ed. Philadelphia: Lippincott Williams & Wilkins; 2013. Chapter 8.
2. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189-198.
3. Nasreddine ZS, Phillips NA, Bédirian V, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53(4):695-699.
4. Dubois B, Slachevsky A, Litvan I, Pillon B. The FAB: a Frontal Assessment Battery at bedside. Neurology. 2000;55(11):1621-1626.
5. Inouye SK, van Dyck CH, Alessi CA, et al. Clarifying confusion: the confusion assessment method. Ann Intern Med. 1990;113(12):941-948.
6. Tariq SH, Tumosa N, Chibnall JT, Perry MH 3rd, Morley JE. Comparison of the Saint Louis University mental status examination and the mini-mental state examination for detecting dementia and mild neurocognitive disorder. Am J Geriatr Psychiatry. 2006;14(11):900-910.
7. Petersen RC. Mild cognitive impairment. N Engl J Med. 2011;364(23):2227-2234.
8. Borson S, Scanlan J, Brush M, Vitaliano P, Dokmak A. The Mini-Cog: a cognitive ‘vital signs’ measure for dementia screening in multi-lingual elderly. Int J Geriatr Psychiatry. 2000;15(11):1021-1027.
9. Shulman KI. Clock-drawing: is it the ideal cognitive screening test? Int J Geriatr Psychiatry. 2000;15(6):548-561.
10. Galvin JE, Roe CM, Powlishta KK, et al. The AD8: a brief informant interview to detect dementia. Neurology. 2005;65(4):559-564.
11. Wittich W, Phillips N, Nasreddine ZS, Chertkow H. Sensitivity and specificity of the Montreal Cognitive Assessment modified for individuals who are visually impaired. J Vis Impair Blind. 2010;104(6):360-368.
12. Julayanont P, Tangwongchai S, Hemrungrojn S, et al. The Montreal Cognitive Assessment — Basic: a screening tool for mild cognitive impairment in illiterate and low-educated elderly adults. J Am Geriatr Soc. 2015;63(12):2550-2554.
13. Bellelli G, Morandi A, Davis DH, et al. Validation of the 4AT, a new instrument for rapid delirium screening: a study in 234 hospitalised older people. Age Ageing. 2014;43(4):496-502.
14. Carson N, Leach L, Murphy KJ. A re-examination of Montreal Cognitive Assessment (MoCA) cutoff scores. Int J Geriatr Psychiatry. 2018;33(2):379-388.