Mild cognitive impairment and dementia assessment

01Purpose and principlesWhat the assessment is for and the core concepts behind it.

Cognitive assessment should answer three separate questions: is there an acquired decline, does it impair independence, and what is the likely cause or subtype? Memory, executive function, attention, language, visuospatial ability and social cognition may be affected in different proportions. Alzheimer disease often starts with episodic memory impairment, vascular disease may emphasise executive speed and gait, Lewy body disease brings fluctuation and visuoperceptual symptoms, and frontotemporal disease may start with behaviour or language. Mixed pathology is common, particularly later in life.

Collateral history is not a vote against the patient. It provides examples of change, while the patient’s account establishes concerns, values and insight. Speak separately when appropriate and consented, and consider coercion or safeguarding. Functional decline should be separated from arthritis, visual loss, poverty, lack of opportunity or a partner who has always managed tasks. A decision-specific capacity assessment is required when choices are questioned; a diagnosis or low test score does not make someone globally incapable.

Specialist memory assessment integrates examination, cognitive profile, structural imaging and response to reversible-factor work. CSF biomarkers, amyloid or metabolic imaging and genomic tests are reserved for cases where diagnostic uncertainty, young onset or an atypical phenotype makes the answer clinically useful. Local memory-service criteria, laboratory panels, imaging access, capacity law and driving rules govern practice.

Key points

Dementia is acquired progressive cognitive decline severe enough to interfere with independent everyday function; mild cognitive impairment causes objective decline while basic independence is substantially retained.
MCI is a risk state, not inevitably early Alzheimer disease: some people progress, some remain stable and some improve when sleep, mood, medicines, sensory loss or illness is addressed.
The most diagnostic information is often the change from the person’s previous abilities—money, medication, travel, cooking, technology, work and judgement—not one cross-sectional memory score.
Obtain collateral history with consent and create a timeline of onset, progression, stepwise change, fluctuation, neurological symptoms, behaviour and daily function.
Exclude delirium first and assess depression, anxiety, alcohol, sleep apnoea, pain, hearing, vision, endocrine disease and anticholinergic or sedative medicines as contributors.
Use a validated brief cognitive instrument such as 6CIT, Mini-Cog, 10-CS or GPCOG, but do not rule out dementia solely because a score is normal.
Interpret testing through education, first language, literacy, culture, hearing, vision, motor and communication ability; choose an adapted tool or specialist assessment when necessary.
Initial blood tests commonly include full blood count, renal, liver, thyroid, calcium, glucose or HbA1c, vitamin B12 and folate, with infection tests selected from risk and presentation.
Structural CT or MRI helps exclude important pathology and support subtype, but atrophy or white-matter change cannot replace clinical functional diagnosis.
Assess driving, medication safety, falls, nutrition, fire, finances, vulnerability, capacity and carer strain from the first evaluation while avoiding blanket assumptions of incapacity.

02Indications, selection and cautionsWhen it is useful, when urgency changes and important limitations.

Mild cognitive impairment

Concern plus objective decline with preserved independence in most daily activities supports MCI after delirium and treatable contributors are considered, but longitudinal review is needed.

Dementia-level impairment

Progressive cognitive change that causes medication errors, unsafe cooking, lost travel, financial failure or dependence in instrumental activities supports major functional impact.

DeliriumRed flag

Hours-to-days onset, fluctuation, impaired attention and altered arousal in an acutely ill person indicates delirium even when chronic neurodegeneration increases vulnerability.

Depressive mimic or contributor

Low mood, anhedonia, slowed thinking, sleep change and prominent subjective concern may impair cognition, but depression and neurodegeneration can coexist and both require follow-up.

Atypical or young onsetRed flag

Presentation before 65, rapid progression, seizures, neurological signs, early language or behavioural syndrome or strong family history warrants expedited specialist and sometimes tertiary assessment.

03Method and interpretationA systematic approach to the test and its findings.

Investigation order

Read from the initial assessment onwards. Tests may run in parallel in urgent care; first-line, preferred, confirmatory, definitive and gold-standard labels appear only when the chapter explicitly states them.

01
Patient and informant functional historyFirst step
Why
Establish decline from baseline, functional consequences and the pattern that suggests a subtype.
Interpretation and limitations
Use concrete dated examples and distinguish inability from never having performed a task. Informant questionnaires can structure evidence but do not replace a sensitive interview.
02
Validated cognitive assessment
Why
Sample cognitive domains, document baseline and support referral and longitudinal comparison.
Interpretation and limitations
A score is affected by language, education, sensory and motor ability, anxiety and delirium. Normal performance does not overrule persuasive functional decline, and a low result does not itself diagnose dementia.
03
Physical and neurological examination
Why
Identify delirium, parkinsonism, stroke, gait disorder, focal signs, sensory loss and systemic disease.
Interpretation and limitations
Fluctuating attention or arousal changes urgency. Eye movements, tone, reflexes, gait and vision can reveal an atypical neurodegenerative or structural process requiring targeted investigation.
04
Baseline laboratory screen
Why
Find anaemia, metabolic, endocrine, nutritional and organ-function contributors and support safe treatment.
Interpretation and limitations
FBC, U&E, eGFR, liver profile, calcium, glucose or HbA1c, TSH, B12 and folate are common components. HIV, syphilis, inflammatory or other tests follow risk and phenotype rather than routine stigma-generating screening.
05
Structural CT or MRI brain
Why
Exclude a mass, subdural collection or hydrocephalus and help characterise vascular and neurodegenerative patterns.
Interpretation and limitations
MRI gives better vascular and regional detail when available and tolerated. Age-related atrophy or white-matter change is common; diagnosis requires concordance with cognition and function.
06
Specialist biomarkers
Why
Increase aetiological confidence when subtype remains uncertain and the result would alter counselling or treatment.
Interpretation and limitations
CSF amyloid and tau, amyloid imaging or selected functional imaging are commissioned through specialist pathways. A biomarker should answer a stated clinical question and still requires phenotypic interpretation.

04Clinical next stepsHow the result changes management or prompts escalation.

01First assessmentSeparate acute from progressive changeFirst stepThe person or someone close reports new memory or thinking difficulty.

11. Establish hours, weeks or years of change and screen attention, arousal, observations, acute illness, medication change and focal neurology before cognitive scoring.
22. If delirium is possible, investigate and treat causes urgently while documenting the pre-illness cognitive baseline from collateral history.
33. For progressive concerns, obtain patient and consented informant examples across cognition, behaviour and instrumental daily activities.
44. Assess mood, sleep, alcohol, hearing, vision and medication burden and agree immediate safety measures without waiting for a final subtype.

02Initial work-upTest cognition fairly and exclude contributorsChronic acquired cognitive decline remains plausible after acute illness assessment.

11. Select a validated cognitive tool suited to language, education and sensory or motor ability, recording any adaptation and why the score may mislead.
22. Perform physical and neurological examination and request the locally agreed blood screen, adding tests only for a defensible clinical question.
33. Review and cautiously reduce anticholinergic, sedative or other cognitive contributors where benefit–risk permits, treating depression and sensory loss in parallel.
44. Refer to the specialist memory pathway when dementia remains suspected, sending the timeline, functional examples, test context, results and risk concerns.

03Specialist formulationDistinguish MCI, dementia and subtypeInitial assessment supports objective decline or uncertainty remains clinically important.

11. Integrate extended cognitive profile, function, neurological signs and structural imaging, identifying pure, mixed or still-uncertain pathology.
22. Use biomarkers or tertiary review for young-onset, atypical or discordant cases only when the answer changes management or genetic counselling.
33. Communicate the diagnosis and uncertainty in accessible language, provide a named contact and assess driving, future planning and support needs.
44. For MCI, address modifiable contributors and arrange longitudinal review with clear triggers for earlier reassessment rather than promising stability or inevitable dementia.

05Procedure and medicine safetyRelevant preparation, treatment and contraindications.

Reduces potentially reversible cognitive, falls and delirium burden and clarifies the person’s baseline before a neurodegenerative diagnosis is finalised.

Deprescribing cognitive-burden medicines

Review anticholinergic, sedative, opioid and other contributing medicines one at a time; taper gradually where withdrawal is possible and use current drug-specific and local deprescribing guidance.

Do not abruptly stop benzodiazepines, gabapentinoids, antidepressants or necessary psychiatric treatment. Balance recurrence of the treated condition, involve the original prescriber and monitor after each change.

Avoids exposing people with a heterogeneous risk state to adverse effects without established symptomatic indication.

No routine cognitive enhancer for isolated MCI

Do not start donepezil, another cholinesterase inhibitor or memantine solely for an undifferentiated MCI label; treatment requires an indicated dementia subtype and specialist-informed plan.

MCI can represent early neurodegeneration, so absence of medication is not absence of care. Arrange risk-factor work, support and longitudinal reassessment, and use biomarkers only through appropriate services.

06Risks, monitoring and follow-upComplications, safety checks and further assessment.

Follow concrete instrumental activities, cognition and behaviour over time using the same adapted methods where possible rather than comparing raw scores from different tools.
Review delirium recurrence, sleep, mood, alcohol, hearing, vision, vascular risk and medicine burden because each can alter the apparent trajectory.
Ask regularly about medication management, cooking, falls, wandering, scams, finances, driving and vulnerability, applying the least restrictive effective support.
Assess carer wellbeing, information needs and willingness to continue tasks, with separate opportunities to disclose strain or safeguarding concerns.
Escalate unexpectedly rapid decline, seizures, focal signs, gait change or systemic symptoms for renewed neurological and medical investigation.

07Special situationsVariants, exceptions and circumstances that change the usual approach.

Function sets the threshold

The distinction between MCI and dementia depends substantially on interference with independent life, not a universal number on a cognitive screen.

Normal scores can conceal decline

A highly educated person may compensate on a brief screen despite major workplace change, while language or education can depress scores without neurodegeneration.

Delirium exposes vulnerability

An acute delirium may reveal previously unrecognised cognitive impairment, but definitive baseline assessment is usually more reliable after the acute syndrome improves.

Collateral needs consent and tact

Informant evidence is valuable but can contain conflict or coercion, so clinicians should seek permission, preserve the patient’s voice and assess safeguarding.

Capacity remains decision-specific

A person may lack capacity for one complex financial decision while retaining capacity for daily care choices and treatment preferences.

08Common pitfallsFrequent interpretation and management errors.

01
Diagnosing dementia from a single low cognitive score ignores delirium, language, education, hearing and function.
02
Reassuring from one normal brief screen can miss executive, visuospatial or high-premorbid-ability decline supported by collateral evidence.
03
Calling acute inattention a dementia progression delays treatment of infection, hypoxia, medication toxicity or another delirium cause.
04
Ordering specialist biomarkers before basic history, function, bloods and structural imaging uses advanced tests without a clear clinical question.
05
Assuming a diagnosis removes all capacity or driving rights replaces individual legal assessment with stigma.