7 Old age psychiatry7.3 Specialist assessment of older people
7.3 Specialist assessment of mental health in older people
7 min read3 figures
Written by Resident Doctor Committee · Published
Preparing and taking the history
Sensory aids come first: check hearing aids and glasses, and book a professional interpreter, because relatives should not routinely interpret sensitive discussions.
A home visit shows function, environmental risks and support needs that a clinic misses.
Onset and trajectory narrow the cause, since hours, weeks and years suggest different explanations.
The history covers cognition, mood, psychotic phenomena and their link to sensory loss, function, sleep, appetite, pain, continence, medicines, alcohol and physical illness. Do not assume a symptom is just ageing.
Collateral information
An informant gives the baseline and the change, but listening to a carer does not authorise you to disclose the patient's information back to them.
Disagreement between accounts neither proves confabulation nor makes the informant right; it may reflect relationship strain, little chance to observe, or abuse.
If no collateral is available, record that and how it lowers your confidence.
Cognitive assessment
Agnosia is failure to recognise a stimulus perceived through an intact sensory channel.
Scores are affected by education, language, culture, sensory loss and fatigue; a slow answer may mean the person cannot hear you.
A cut-off is not a diagnosis: read a score against the tool's validation population and the patient's background.
Instrument | Content and scoring | Best use |
|---|---|---|
Mini-Mental State Examination (MMSE) | Licensed; no executive items; ceiling effect in educated people, and floor effects | Widely used general screen |
Montreal Cognitive Assessment (MoCA) | About ten minutes, out of 30; five-word delayed recall, trail making, clock drawing, abstraction, fluency | Mild cognitive impairment, where it is far more sensitive than the MMSE |
Addenbrooke's Cognitive Examination III | Includes executive and fluency tasks | Behavioural variant frontotemporal dementia that a near-perfect MMSE misses |
Frontal Assessment Battery | Six tasks out of 18: similarities, lexical fluency, motor sequencing, conflicting instructions, go/no-go, prehension behaviour | Frontal dysfunction with preserved memory |
Rowland Universal Dementia Assessment Scale | Six items out of 30: memory, body orientation, praxis, cube drawing, judgement, animal fluency | Multicultural populations: developed in Sydney, easy through an interpreter, depends little on schooling or literacy |
Six-item Cognitive Impairment Test, General Practitioner Assessment of Cognition | Brief | Initial assessment |
Abbreviated Mental Test Score | Brief | Bedside screening; it does not identify the cause of impairment |
Behavioural variant frontotemporal dementia shows apathy, disinhibition, eating change and poor planning, often with memory spared.
CAMDEX-R is the comprehensive standardised instrument for the differential diagnosis of dementia in older people: a patient interview, an informant interview, a physical examination and the CAMCOG-R battery.
CAMCOG-R is the cognitive battery within CAMDEX-R, not a free-standing scale; it contains every MMSE item and adds praxis, abstract thinking and perception, with executive items in the revised form.
The Hachinski Ischaemic Score covers stepwise deterioration, emotional lability, nocturnal confusion, hypertension and stroke history; 7 or more suggests vascular dementia and 4 or less a primary degenerative cause.
Premorbid ability and executive tests
Test | What it measures | How it works |
|---|---|---|
National Adult Reading Test | Premorbid ability | The person reads aloud 50 irregularly spelt words; pronunciation depends on prior familiarity, relatively spared early in Alzheimer's disease |
Raven's Progressive Matrices | Current non-verbal fluid intelligence, with little cultural loading | The person picks the piece that completes a pattern |
Wisconsin Card Sorting Test | Concept formation and set shifting | Perseverative errors suggest frontal executive impairment |
Stroop test | Frontal executive function | Interference control between automatic word reading and colour naming |
Delirium instruments
Instrument | Content | Practical point |
|---|---|---|
4 'A' Test (4AT) | Two minutes; scores alertness, the four-item Abbreviated Mental Test, attention by reciting months backwards, and acute change or fluctuation | Untrained staff can use it, and a patient too drowsy to test still scores |
Delirium Rating Scale (DRS, DRS-R-98) | Broad symptom coverage with severity ratings; separates delirium from dementia, depression and schizophrenia | Needs a trained psychiatric rater |
Depression scales
Scale | Rated by | Content | Suited to |
|---|---|---|---|
Cornell Scale for Depression in Dementia | Clinician, interviewing the carer then the patient | 19 items over the previous week, including observed signs such as poor appetite and diurnal variation | Dementia that undermines self-report |
Geriatric Depression Scale, 15-item | Patient, self-completed yes or no | Omits most bodily symptoms; 5 or more prompts fuller assessment | Cognitively intact older adults |
Brief Assessment Schedule Depression Cards (BASDEC) | Patient sorts cards into true, false or don't know piles | 19 statements in large type, one per card | Older medical inpatients; deafness or poor writing, and needs no speech |
Montgomery-Åsberg Depression Rating Scale | Clinician | Ten items, light on somatic content: sadness, inner tension, loss of feeling, pessimism, suicidal thoughts | Detecting change with treatment |
The Cornell scale is chosen in dementia because it relies on the carer and observation, whereas the Geriatric Depression Scale needs intact cognition.
Function and staging
Basic activities are washing, dressing, toileting, eating and transferring, and the Barthel Index is the standard scale summarising them in older people.
Instrumental activities are shopping, cooking, communication, finances and medication, and difficulty here often appears first.
A dosette box helps only if the person can read labels, open packaging and use it safely; some need supervised administration.
The Clinical Dementia Rating stages dementia using informant-reported everyday function (community affairs, home and hobbies, personal care) as well as cognition.
The DLD (formerly the DMR) is a carer-rated dementia questionnaire for adults with intellectual disability, judged by change on repeat rating against the person's own baseline.
Driving
Dementia must be notified to the Driver and Vehicle Licensing Agency in Great Britain or the Driver and Vehicle Agency in Northern Ireland; with mild cognitive impairment, consider whether driving is affected.
If someone drives on against advice and poses a serious danger, follow General Medical Council disclosure guidance, telling the patient beforehand where practicable.
Physical examination and investigations
Examine for focal signs, parkinsonism, postural hypotension, frailty and sensory impairment, and review anticholinergic and sedative burden.
The routine blood panel offered to everyone with suspected dementia is a fixed set: blood count, renal function and electrolytes, calcium, glucose, liver and thyroid function, B12 and folate. Other tests need a specific indication.
Syphilis serology is added when the history carries a syphilis risk factor, because neurosyphilis is a treatable cause of cognitive and behavioural decline.
Fully reversible dementias are uncommon, but treating contributors can still improve function.
Imaging
The National Institute for Health and Care Excellence (NICE) recommends structural imaging to exclude reversible causes and identify the subtype, unless dementia is well established and the subtype clear.
Structural magnetic resonance imaging is the modality of choice for small vessel white matter lesions, lacunes and infarcts, with superior grey-white contrast on T2 and FLAIR sequences.
Subcortical vascular dementia presents with slowed thinking, apathy, small-stepped gait and urinary urgency.
FDG-PET or perfusion SPECT images function (regional glucose uptake or blood flow), not structure; NICE recommends it when the subtype is uncertain after structural imaging and frontotemporal dementia is suspected.
99mTc-HMPAO perfusion SPECT is used when FDG-PET is unavailable.
Targeted tests
Presentation | Likely cause | Test |
|---|---|---|
Drowsiness, nausea and confusion within weeks of starting an SSRI antidepressant | Hyponatraemia from inappropriate antidiuretic hormone secretion | Serum sodium on urea and electrolytes |
Head injury while anticoagulated, then weeks later headache, fluctuating drowsiness, cognitive change and a focal sign | Chronic subdural haematoma | Urgent non-contrast computed tomography of the head |
Long-term lithium with stones, bone pain, abdominal pain, constipation, low mood or confusion | Hyperparathyroidism with raised calcium | Serum calcium, then parathyroid hormone if high |
Risk, capacity and safeguarding
Ask directly about suicide, self-harm, self-neglect, falls, getting lost, exploitation and carer strain; older age and physical illness do not make suicidal thoughts less worth assessing.
Locks, sensors and tracking restrict liberty, so they need individual consent or legal authority, and must be proportionate and reviewed.
Capacity is assessed separately for each decision, and an unwise choice does not establish incapacity.
Formulation: delirium, dementia and depression
Feature | Delirium | Dementia | Depression |
|---|---|---|---|
Onset | Hours to days | Months to years | Often weeks |
Course over 24 hours | Fluctuating | Often stable; some subtypes fluctuate | Mood or performance may vary |
Attention | Often markedly impaired | Relatively preserved early | May be impaired |
Conscious level | May be altered | Usually clear | Usually clear |
Hallucinations | Often visual | Especially Lewy body disease | In severe psychotic depression |
A new change in someone with dementia may be superimposed delirium, and quiet drowsiness matters as much as agitation because hypoactive delirium is easily missed.
"Pseudodementia" misleads by implying the impairment is false or bound to reverse; "I don't know" answers are clues, not proof.
Follow cognition after treating mood, because depression and neurodegeneration can coexist.