Mental state examination and how to present it

• The mental state examination is a structured description of what a person is experiencing and showing right now: appearance, behaviour, speech, mood and affect, thought form and content, perception, cognition, insight and judgement, and risk.
• It is not a checklist of tests. You are observing from the moment you meet the person and recording what you see and hear, in the person's own words where possible.
• The mood is what the person says they feel, in their own words; the affect is what you see on the surface. Both are recorded separately.
• Every abnormal finding must be described in observed, non-judgmental language before it is named, because "flat affect" or "delusional" only mean something if the reader can see how you got there.
• Cognition is tested with attention, orientation, memory and simple bedside tasks, sometimes helped by a short screen such as the Montreal Cognitive Assessment (MoCA) ^[3] or the Mini-Mental State Examination (MMSE). Those screens are screening tools, not diagnoses.
• Risk is assessed at every stage and summarised at the end: suicidal and self-harm ideation, intent, plan, means and timeline; risk to others; self-neglect; and protective factors.
• Present the findings by leading with the most important abnormality, grouping the normal domains, stating the negatives that matter, and ending with a one-sentence summary, a differential and a plan.

By the end of this page you should be able to:

1. Introduce, consent and position a person for a mental state examination in a safe and respectful way.
2. Perform the examination in the standard order, from first observations to risk assessment.
3. Record normal and abnormal findings for each domain in objective, non-judgmental language. ^[2]
4. Distinguish mood from affect, and thought form from thought content.
5. Choose and interpret a brief cognitive screen, naming its limits.
6. Assess and document risk without overstating it or understating it.
7. Turn a list of mental state findings into a spoken and a written presentation that leads with the most important abnormality.
8. Recognise the presentations of normality, depression with psychomotor retardation, mania and psychosis.

The mental state examination is the psychiatry equivalent of the physical examination: a precise, at-this-moment description of how a person's mind is working. It answers three questions that a history alone cannot answer: what is the person experiencing right now, what can you observe, and how well do they understand their situation. NICE guideline NG222 on depression in adults puts recognition and assessment at the centre of care, and that emphasis applies to every mental health encounter: you cannot manage what you have not first described. ^[4]

The examination is usually carried out after the history, with the person seated and comfortable, in good light and in a quiet room. It rarely needs equipment beyond a pen and paper for cognitive testing. Most of it is done by talking and observing, and the written note is the part that usually takes longest.

Setting matters. Sit at the same level as the person, do not interview from behind a desk if you can avoid it, and explain before you begin that you will ask about thoughts, feelings and concentration, and that people sometimes find some questions personal. Confidentiality should be explained at the start, with its limits stated plainly: you will share information with the responsible clinician, and you must act if there is risk to the person or to others. If an interpreter is needed, use a trained interpreter rather than a relative, and record that you did.

A Royal College of Psychiatrists supported review notes that the Mental State Examination is a core component of psychiatric assessment and medical training, that its domains can be numbered from 6 to 18 in different versions, and that the de facto United Kingdom standard is the structure reproduced here. ^[1] Versions vary across countries and textbooks, so use one structure consistently and say which one you used.

• Introduction and consent. Give your name and role, and explain that you would like to talk and to ask a few questions about thoughts, feelings and memory. Obtain consent before you start, and accept a request to pause or stop without making it feel like a test of compliance.
• Chaperone. Offer a chaperone for any examination that could be felt as intrusive, and record the offer and the response. A mental state examination is usually non-intimate, but trauma-informed practice means offering one whenever a person is very distressed, very guarded or describing abuse.
• Hand hygiene and dignity. Clean your hands, and before you begin arrange seating, privacy and temperature so that the person is not cold or exposed. A person in a hospital gown will answer differently from one who is dressed and warm.
• Safety and exposure. Before you enter a room with a person who may be agitated, check the exits, remove objects that could be used to harm someone, and tell a colleague where you are. During the examination, look at the whole body as well as the face: self-harm scars, weight change, injection marks and clothing choices are all part of the mental state.
• Position. Sit opposite the person, with eye contact easy and not forced. Avoid standing over them. Put the notes to one side and write in short phrases so that you are not hiding behind the clipboard.
• Environment. Use a quiet, well-lit room. Turn off televisions and radios. For cognitive testing, minimise interruptions and make sure the person can hear you and see the pen and paper.
• Safety planning before you start. Check whether the person has any immediate risk in the room. If you are unsure, do not conduct the interview alone.

The standard order is the one most commonly taught in English-language psychiatry and reproduced in the Royal College review: appearance, behaviour, speech, mood and affect, thought form, thought content, perception, cognition, insight, judgement, with risk woven through and summarised at the end. ^[1] The reason for the order is practical rather than scientific: you begin with what you can see without asking a question, move to what the person says, and finish with the abstract and judgmental domains.

1. Appearance. Begin the moment you see the person. Note apparent age, sex and build; dress and its appropriateness for the weather and situation; grooming and hygiene; posture and gait; eye contact; facial expression; and any visible signs of self-harm, substance use or neglect. Record what you see rather than your interpretation: "clothes unwashed for several days, strong body odour, hair matted" is useful; "neglected" is a conclusion that needs the observation behind it.
2. Behaviour. Observe activity level and style. Is the person restless, pacing, unable to sit still, or slowed, with long pauses before moving? Is there psychomotor retardation, with slowed movements and a heavy, effortful quality, or psychomotor agitation, with fidgeting and hand-wringing? Note cooperation, warmth or coolness of rapport, mannerisms, rituals, disinhibition, tearfulness, laughter, and any aggression or threatened violence. Watch the hands: tremor, repetitive picking and hand-washing are easy to miss if you only watch the face.
3. Speech. Listen to the rate, volume, rhythm, amount and fluency of speech, and to the latency between your question and the answer. Speech may be pressured (difficult to interrupt, rapid, with an increased amount), impoverished (little speech, short answers, long pauses), or abnormal in form (stammering, mutism, dysarthria from a neurological cause). Record the person's actual words when they are unusual or repetitive.
4. Mood and affect. Ask an open question first: "How have you been feeling in yourself?" Record the mood in the person's own words, in quotation marks where they are striking, because "I feel like I'm in a black hole" tells the reader far more than "depressed". Then record the affect that you observe: its range (full, restricted, blunted or flat), its reactivity (does it move with the conversation), its intensity, its appropriateness to the topic, and its congruence with the stated mood. A flat affect with a normal mood is as informative as the reverse, and a labile affect that swings within minutes is a finding in itself.
5. Thought form. This is how the person thinks, not what they think about. Listen to the connections between ideas. Normal thought is goal-directed: it gets to the point. Abnormal forms include tangentiality (drifting off the topic and never returning), circumstantiality (excessive detail but eventually reaching the point), flight of ideas (rapid shifts between connected ideas, often with punning or clang associations, typical of mania), derailment or loosening of associations (shifts between unrelated ideas), thought blocking (a sudden, complete stop in mid-sentence), perseveration (returning to the same idea despite new questions), and word salad or neologisms (inventing words or joining unrelated words). Note whether the person can follow a conversation and whether they seem to lose the thread.
6. Thought content. This is what the person thinks about. Explore systematically rather than waiting for it to emerge: preoccupations and worries; obsessions and compulsions; delusions, which are fixed false beliefs held despite evidence and not consistent with the person's culture or education (persecutory, grandiose, referential, guilty, nihilistic, somatic and jealous types); and ideas of reference and passivity (the feeling that thoughts are being inserted, withdrawn or broadcast). Always ask directly about thoughts of self-harm and death: asking does not plant the idea, and not asking misses the most important risk. Record suicidal ideation, intent, plans, means, preparatory behaviour, protective factors and a timeline. Record thoughts of harming others, of abuse, and of neglect or self-neglect.
7. Perception. Ask about hallucinations in each modality: auditory (voices, including command hallucinations), visual, tactile, olfactory and gustatory. Ask about illusions (misinterpreting real stimuli) and about depersonalisation and derealisation (feeling detached from oneself or from the world). Hypnagogic and hypnopompic hallucinations occur when falling asleep or waking and are usually benign; they are distinguished by context and by the fact that the person recognises them as unreal. Third-person auditory hallucinations and running commentary are highly suggestive of psychosis.
8. Cognition. Start by establishing that the person is attentive enough to be tested: an inattentive person cannot cooperate fully, so attention is tested first, by asking them to immediately repeat three words. ^[2] Then test orientation to person, time and place; short-term memory by asking the person to recall those three objects after about 2 to 5 minutes; ^[2] long-term memory by a question about the past; attention and concentration by spelling a five-letter word forwards and backwards or by serial sevens; ^[2] language by naming objects and their parts, reading, writing a sentence and following a three-stage command; ^[2] visuospatial ability by asking the person to draw a clock, a house or two intersecting pentagons; ^[2] abstract reasoning by similarities between three or four objects or by interpreting a proverb; and judgement by a hypothetical situation such as finding a stamped letter on the pavement. ^[2] A five-minute screen of the whole examination is word naming in one category (for example animals) for one minute. ^[2]
9. Insight and judgement. Insight has several levels: does the person believe they have a problem at all, do they accept that it might be a mental health problem, do they see the need for treatment, and are they willing to accept help. Judgement is tested by the hypothetical question above and by how the person has handled recent real decisions. Record both, because a person can have full insight and very poor judgement, or no insight and surprisingly good judgement.
10. Risk assessment. Risk is not a separate stage to be added at the end; it is being assessed in every earlier stage. Summarise it explicitly: suicide risk (ideation, intent, plan, means, lethality, timeline, protective factors, previous attempts, the effect of alcohol or drugs on impulse control), deliberate self-harm, risk to others (violence, abuse, exploitation), self-neglect and safeguarding, and risk to dependent children or others in the person's care. Where the risk is acute, say so plainly, keep the person in sight until a plan is agreed, and make the referral before you leave the room. NICE guideline NG222 on depression in adults emphasises that recognition and assessment include a full assessment of risk, and that emphasis is worth carrying into every presentation. ^[4]

The table below records the common findings, how you elicit or observe them, and what they usually mean. Individual signs are not diagnoses, and the same sign can mean different things in different contexts.

The presentation is a summary of the person, not a recital of every question you asked. The spoken and the written version use the same order and the same abbreviations, and each should be brief.

Spoken order:

1. Open with the most important finding, or with the absence of one. "This is a 29-year-old man with a three-week depressive episode, and the key findings are psychomotor retardation, nihilistic ruminations and active suicidal ideation." If there is no abnormality, say so in the first sentence, because that is usually the most reassuring and most important news.
2. Report domain by domain, in the examination order, in one or two short sentences for each domain. Group the normal domains together: "Speech, perception and cognition are normal to testing." Do not repeat every subtest; state the exception, not the whole normal.
3. State the negatives that matter. These are the findings that change the differential. For a depressed person, say whether psychotic features, suicidal intent and cognitive impairment are present or absent. For a person with psychosis, say whether mood symptoms, cognitive impairment and risk are present or absent. Silence on these points implies you did not look.
4. Give the summary sentence. In one sentence say what you think this is: a depressive episode with psychomotor retardation, a manic episode, a first episode of psychosis, a delirium, or no mental illness at all.
5. Give the differential in one sentence. Two or three plausible diagnoses, ranked by likelihood and by danger, not a textbook list. "The dangerous alternative is a steroid-induced mood change; the more likely is a first episode of bipolar disorder."
6. Close with the plan in one sentence, if you are presenting to a senior: what you would do now, including risk management and who you need to involve.
7. Say what you could not test and why. A sentence such as "Visuospatial function was not assessed because the person's English was limited, and an interpreter was used for the history but not for the memory testing" is more useful than a polished note that hides a gap.

Written order:

Use the same domains in the same order, abbreviated only after you have expanded each abbreviation once in the note. A useful shape is:

Appearance (dress, grooming, build); behaviour (activity, rapport, psychomotor); speech (rate, volume, form); mood/affect ("..." / objective); thought form; thought content; perception; cognition (orientation, attention, memory, screen, score); insight/judgement; risk (ideation, intent, plan, means, protective factors). Summary: ... Differential: ... Plan: ...

Write observations before conclusions. "Eye contact poor, looked at the floor throughout, three-second pause before answering most questions" is better than "socially withdrawn". Avoid judgement words such as "lazy", "difficult", "attention-seeking" or "manipulative"; they add nothing and close off better explanations. Avoid jargon without a description: "flat affect" should follow a line that shows what flat looked like.

The following reports use invented patients and are illustrative. They are written to show the shape of a presentation, not as a substitute for assessing a real person.

Illustrative normal. "Mr James Okafor, 46, referred by his general practitioner because of headaches. He appears well and appropriately dressed, and is cooperative and relaxed throughout. Speech is normal in rate, volume and fluency, with no latency. He describes his mood as 'fine, just a bit worried about these headaches', and his affect is full, reactive and congruent. Thought form is goal-directed, and thought content is free of delusions, obsessions or suicidal ideas. He denies hallucinations and illusions, and there is no depersonalisation. He is oriented to person, time and place; recalls three of three words after five minutes; spells 'world' backwards correctly; and completes a clock drawing. He has full insight into why he has come, and there is no risk to himself or others. In summary, a normal mental state examination with no psychiatric abnormality."

Illustrative depression with psychomotor retardation. "This is Mrs Amira Hassan, 58, seen in the medical ward after a deliberate overdose three days ago. She is unkempt, sits very still with her hands in her lap, and answers slowly after long pauses; her movements are heavy and effortful. Speech is monotonous, quiet and sparse. She says her mood is 'empty and hopeless', and her affect is flat and congruent, with occasional tearfulness. Thought form is slow but goal-directed. Thought content includes persistent ideas of guilt and worthlessness, the belief that her family would be better off without her, and passive suicidal ideation with a wish to be dead, but no active plan and no intent to act today. She denies hallucinations and delusions. Orientation and attention are preserved; she recalls two of three objects after five minutes and spells 'world' backwards with one error, which may reflect poor concentration rather than an amnestic problem. She accepts that she is depressed and that she needs help, and recognises the overdose as a moment of desperation. Protective factors include her wish to care for her grandchildren and her strong faith. In summary, a severe depressive episode with psychomotor retardation and passive suicidal ideation, but no psychosis and no active plan. The alternatives I would keep in mind are a medication-related low mood and a bereavement reaction. I would keep her on close observation, refer to the liaison psychiatry team and involve the family with her consent."

Illustrative mania. "This is Mr Daniel Reyes, 31, brought to the emergency department by his partner because of three days of 'not sleeping and spending'. He is flamboyantly dressed, wearing sunglasses indoors, and paces the room, shaking hands repeatedly and unable to stay seated. Speech is pressured, rapid and difficult to interrupt, with frequent jumps between topics and puns. He describes his mood as 'fantastic, never better', and his affect is elevated, expansive and irritable when challenged, with rapid swings to anger. Thought form shows flight of ideas with loose associations and clang associations. Thought content includes grandiose beliefs that he can run three businesses at once and that he is about to sign a record deal, and he has been inviting strangers to invest money. He denies hallucinations. He is oriented and alert, and scores well on formal cognitive testing, although his judgement is plainly impaired: he has spent a large sum of money in two days. He has no insight that anything is wrong, and attributes his admission to his partner's jealousy. Risk includes financial loss, escalating irritability, and a history of one physical altercation during a previous episode. In summary, a manic episode with flight of ideas, grandiosity and impaired judgement, with no psychosis and high risk of harm to his finances and relationships. The differential is recurrent mania in the context of bipolar I disorder; the urgent alternative is a stimulant or steroid-induced state. He needs a senior review, a safe environment away from internet banking and driving, and a discussion about admission."

Illustrative psychosis. "This is Miss Sarah Chen, 22, found by her flatmates talking to empty space in the kitchen. She appears thin, dishevelled and suspicious, and avoids eye contact, scanning the corners of the room. Activity is mixed: periods of frozen stillness alternating with bursts of restless pacing. Speech is slow and sparse, with long pauses, and occasionally halts mid-sentence. She says her mood is 'frightened, all of the time', but her affect is blunted and often incongruent, with a faint smile while describing distress. Thought form is disorganised, with derailment between unrelated ideas. Thought content includes the fixed belief that her neighbours are recording her through the walls and broadcasting her thoughts to people on the street, and the fixed belief that her food is being poisoned, so she has stopped eating. She denies thoughts of harming anyone, and she has stopped eating because of her fear of poisoning, not because she wishes to die. She hears a male voice, sometimes two, commenting on what she is doing, including comments that she is 'dirty and should be punished'. She denies visual hallucinations. She is oriented to person and place but not to the exact date, and recalls one of three words after five minutes. She has no insight: she believes the voices are real people and that treatment would be dangerous. Risk includes refusal of food and fluids, high vulnerability, and a recent period when she was planning to flee the flat. In summary, a first episode of psychosis with persecutory and referential delusions, third-person auditory hallucinations, thought disorder and impaired insight, with associated self-neglect but no expressed suicidal or homicidal intent. The differential includes schizophrenia, a psychotic mood disorder and a substance-induced psychosis, which I would clarify with a drug screen and a careful history. She needs a senior psychiatric assessment today, a safe environment, and physical review, including hydration and nutrition."

• Recording conclusions instead of observations. "Neglected" or "agitated" tells the reader nothing without the evidence; write the behaviour and let the conclusion follow.
• Treating mood and affect as the same thing. Mood is subjective and belongs in quotation marks; affect is what you observe. A person can say they are fine while looking devastated.
• Not asking about self-harm directly. Many trainees avoid the words "suicide" and "kill yourself" for fear of giving ideas. Asking never plants the idea, and avoiding it is the actual risk.
• Confusing thought form with thought content. Form is the rails; content is the train. A person can have normal content and very disordered form, which is the hallmark of formal thought disorder.
• Over-calling delusions. An unusual belief is not automatically a delusion. Test it: is it fixed, is it shared by the person's culture, and does counter-evidence change it?
• Under-testing cognition in younger or articulate people. Dementia, delirium and acquired brain injury are not confined to older adults, and a fluent talker can be profoundly disoriented.
• Forgetting that cognition tests are educational-culture dependent. Clock drawing, proverb interpretation and category fluency all depend on schooling and language. Interpret with the whole person, not against a single norm.
• Ignoring the negatives. A list of abnormalities with no statement of what was normal leaves the impression that the examination was incomplete. Say explicitly that perception, cognition or risk screening was normal.
• Leading with trivia. List the most important finding first. Examiners and colleagues remember the first sentence and the last.
• Using judgement words. "Difficult", "manipulative" and "attention-seeking" are not clinical findings and usually conceal a failure to understand what is happening.
• Assuming that a screen is a diagnosis. The MoCA and the MMSE are screening tools with their own limitations, discussed below. A score does not equal a disease.
• Risk assessing once and moving on. Risk changes during the interview and during treatment. Reassess and record the change.

• The mental state examination describes the person's present mental state, from appearance to risk, and is built on observation rather than on a battery of tests.
• It is not intimate in the physical sense, but it is deeply personal, and introduction, consent and privacy shape the answers you get.
• Appearance and behaviour are recorded from the moment you meet the person, in the words you would use to describe what you saw.
• Mood is what the person says; affect is what you see; both are recorded.
• Thought form is how ideas move; thought content is what the ideas are.
• Perception covers hallucinations and related phenomena in every modality, not only hearing voices.
• Cognition is tested by attention, orientation, memory, language, visuospatial ability and abstract reasoning, and may be aided by a short screen.
• The MMSE and MoCA are screening tools only: they quantify cognitive performance against a total, but their scores are influenced by education, language, hearing and culture, and neither test diagnoses a cause. Use them to document change over time and to trigger a fuller assessment, not as an endpoint.
• Risk is assessed throughout the examination, not just at the end, and is summarised as ideation, intent, plan, means, timeline, harm to others and protective factors.
• Insight and judgement are separate findings and should be recorded separately.
• The presentation leads with the most important finding, groups the normal domains, states the relevant negatives, and ends with a summary, a differential and a plan.
• In writing, observations come before conclusions, and judgment words are never used.

Plain-text related topics include: psychiatric disorders (depression, bipolar disorder, schizophrenia and other psychoses), anxiety disorders, dementia and delirium, self-harm and suicide risk assessment, frontal lobe and cognitive disorders, and the approach to the unconscious or confused patient. Earlier MedGrasp pages on neurological and physical examination, and on presenting a patient orally and in writing, take the same presentation structure used here. These MedGrasp topics are listed as plain text until the corresponding pages are verified.