Learning outcomes
By the end of this lesson you should be able to: explain why psychiatry classifies, and compare the purposes and structure of ICD-11 and DSM-5; describe the strengths and weaknesses of categorical classification and the dimensional alternative; take a psychiatric history in a structured order; perform and record a mental status examination across all its domains, including thought form and thought content; distinguish insight, judgement and mental capacity, and describe how each is assessed and documented; and describe the screening instruments used routinely in practice.
Prerequisites
None. This is the entry point to clinical psychiatry. The behavioural sciences lessons (1-4) supply the vocabulary of motivation, emotion, conditioning and memory used throughout, but nothing in this lesson depends on any of them.
The big idea
Classification does three jobs in psychiatry, and it is worth keeping them apart. It is a communication device: 'F20 schizophrenia' conveys a great deal to a psychiatrist in another country. It is a research device: it defines the population a trial actually enrolled. And it is a management device: it maps onto treatment selection and to the legal frameworks of detention. The difficulty is that no psychiatric diagnosis has a confirmatory test, so a category is a bet about natural boundaries that biology has not yet confirmed, and that has practical consequences — a misdiagnosis is not merely a labelling error but a route to the wrong treatment.
ICD and DSM: the two classifications
| Feature | ICD-11 (and ICD-10) | DSM-5 (and DSM-5-TR) |
|---|---|---|
| Publisher | World Health Organization | American Psychiatric Association |
| Legal and service use | The coding system of member states' health statistics, of death certification and of most national services including the NHS and the Indian health system | Dominant in North American research, insurance and practice |
| Structure of classification | Chapters by disorder group; mental, behavioural and neurodevelopmental disorders occupy a dedicated chapter with F-codes | Also chapter-based, but with explicit criteria sets designed for operationalised research use |
| Diagnostic culture | Statement of clinical practice; often deliberately less detailed about criteria sets and duration thresholds, allowing clinical judgement | Explicit polythetic criteria (a specified number from a list, for a specified duration, plus exclusions) |
| Axes | DSM-IV used a multiaxial system (Axis I clinical, II personality, III medical, IV psychosocial, V functioning); ICD-10 also used a multiaxial scheme for child and adolescent psychiatry | DSM-5 removed the multiaxial system and merged most of it into the diagnostic code plus separate measures of functioning |
| Frequency of revision | Decadal, deliberately conservative | Revised in more frequent iterative releases in practice |
NoteThe practical rule: know ICD-11 descriptions in order to practise anywhere in the world, and know DSM-5 criteria sets in order to read the literature. Many trial results cannot be applied to ICD-defined patients, and the two systems do not draw every boundary in the same place — for example, schizophrenia in DSM-5 requires six months of continuous disturbance, whereas ICD-10 required one month.
Categorical versus dimensional is the central unresolved argument. Categorical systems give categories: present or absent, with a threshold. Dimensional systems give scores along continua, on which everyone sits somewhere. The categorical approach is easy to communicate, drives clinical decisions and research sampling, and matches how law and entitlement work. Its cost is high comorbidity and a population of sub-threshold cases who are disabled but undiagnosable. The dimensional approach captures that variation honestly, but thresholds are unavoidable in practice — a treatment decision requires a cut point — and dimensional systems are harder to communicate at handover.
Emil Kraepelin's late-nineteenth-century separation of dementia praecox from manic-depressive illness is the historical root of the current categories, and his method — grouping disorders by course and outcome rather than by presenting symptom — remains the thinking behind the schizophrenia-bipolar divide. Case recognition is helped by two of the criteria a good category should meet: reliability (would two clinicians agree?) and validity (does the category predict anything about cause, treatment or outcome?). Most psychiatric categories have acceptable reliability and modest validity.
The psychiatric history
The history is not a form to fill in. Its purpose is to establish what has happened, in what order, with what consequence, in the patient's own words, with enough collateral to check the account. A defensible order follows.
| Section | What it establishes | Practical notes |
|---|---|---|
| Referral and reason for contact | Why the patient is here now — self, family, police, other services | Quote the source: 'his sister brought him; he said there was nothing wrong' |
| Presenting complaint | The patient's own words, not your translation | Record verbatim, in quotation marks |
| History of presenting illness | Onset, duration, course, precipitants, severity, functional consequence, treatment already tried, previous episodes | Establish the time scale — frequency, change from baseline, what was happening in the patient's life |
| Past psychiatric history | Previous episodes, admissions, self-harm, previous diagnoses and medications | Prior attempts and their response guide the next treatment choice |
| Past medical history and medications | Medical illness, surgery, head injury, epilepsy, currently prescribed and over-the-counter drugs | Steroids, L-dopa, dopamine agonists, interferon and thyroid disease cause psychiatric presentations |
| Family history | Affective, psychotic, substance, suicide, neurodevelopmental, dementia | Draw a genogram; ask about suicide explicitly, not only about diagnoses |
| Personal and developmental history | Pregnancy and birth, developmental milestones, education, employment, relationships, sexual history, children | Childhood adversity is best asked about once the relationship is established and framed as relevant to current symptoms |
| Substance history | Alcohol, tobacco, cannabis, stimulants, opioids, benzodiazepines, new psychoactive substances | Quantify: units per day, days per week; AUDIT rather than impression |
| Forensic history | Convictions, violence, current charges or bail conditions | Affects risk formulation and admission decisions |
| Social circumstances | Housing, income, debt, immigration status, carers, social support, dependants | A housing dispute or an asylum decision is frequently the operative change in an exacerbation |
| Premorbid personality | The person's standing pattern of thinking, feeling and relating before this episode | Ask a relative or old friend, not only the patient |
Watch outA history taken through a family member, a telephone interpreter or an intermediary tool is not inferior, but it is different. Document the interpreter's number, the mode (in person, video, telephone), and the patient's preferred language. Never use the patient's child as an interpreter for a psychiatric history.
The mental status examination
The MSE is the state of the mind at the time of interview: a picture with a date, not a diagnosis. It is built from observation and directed questioning, in a fixed order so that nothing is omitted.
| Domain | What to observe or ask | Descriptive vocabulary |
|---|---|---|
| Appearance and behaviour | Grooming, hygiene, clothing, eye contact, posture, movements, rapport, cooperation, agitation or retardation | Kempt or neglected; self-neglecting; restless; retarded; guarded or open; engaged or withdrawn |
| Speech | Rate, rhythm, volume, quantity, dysarthria, dysphasia | Pressured, retarded, poverty of speech, blocking, dysarthric, incoherent |
| Mood | The patient's subjective report over weeks | Depressed, anxious, elated, irritable, empty, numb |
| Affect | The objective expression observed in the room: range, intensity, congruence with mood, reactivity | Preserved or constricted; blunted or flat; labile; incongruous; reactive |
| Thought form | How thoughts are linked and expressed: speed, coherence, goal-direction | Pressure of thought, flight of ideas, derailment, tangentiality, circumstantiality, loosening of associations, neologisms, clang associations, blocking, word salad |
| Thought content | What the patient is thinking about: beliefs, preoccupations, concerns | Delusional or non-delusional; overvalued ideas; obsessions and compulsions; suicidal ideation; hopelessness; guilt |
| Perception | Hallucinations, pseudohallucinations, illusions, depersonalisation and derealisation | Auditory third person, running commentary, visual, somatic, olfactory — with insight into reality status |
| Cognition | Orientation, attention, concentration, memory, language, praxis, visuospatial construction | Delirium is an attention and arousal failure first; test attention by months backwards or digit span backwards |
| Insight | Does the patient recognise the experiences as abnormal, attribute them to illness, and accept treatment? | Insight is graded — full, partial or absent — on each of those three components independently |
| Judgement | Ability to make reasonable decisions about own safety and conduct | Impaired judgement is separate from impaired capacity, and from cognitive impairment |
NoteInsight is not a yes-or-no. It has three components — recognising that something is happening, attributing it to illness, and accepting treatment — and they can dissociate. A patient may accept that voices are abnormal while insisting they are spiritual gifts, and may accept a tablet while denying any illness. Record each component separately; the third is the one that matters for admission.
Mental capacity is a legal concept, separate from insight and from cognition. A patient has capacity for a decision when they can understand the relevant information, retain it long enough to weigh it, use and weigh that information in the reasoning, and communicate the decision. It is decision-specific and time-specific: a patient may have capacity to refuse a blood test but not to refuse admission, and capacity may return next week. Where capacity is absent and there is no valid advance refusal, a decision is made in the patient's best interests or by an appointed substitute decision-maker.
Risk assessment: a formulation, not a score
Risk assessment combines specific and generic factors into a judgement about a specific patient over a specific period, and it is used to structure care, not to predict acts of violence — prediction of rare events with structured tools is no more than chance in most populations. The two questions to answer are: what is the risk (suicide, self-harm, violence, neglect, exploitation, absconding), and to whom is there a risk.
| Domain | Static historical factors | Dynamic clinical factors |
|---|---|---|
| Suicide | Previous attempt (the strongest single factor), family history of suicide, male sex, older age, chronic pain or illness | Current depressive episode, hopelessness, insomnia, recent loss, intoxication, access to means, planned intent, isolation |
| Self-harm | Repetitive self-harm history, childhood adversity, personality disorder diagnosis | Current distress, substance use, relationship breakdown, discharge planning |
| Violence to others | Previous violence, male sex, young age, substance misuse, early conduct disorder | Current psychosis with persecutory delusions naming a person, command hallucinations, agitation, intoxication, access to weapons |
| Neglect and exploitation | Dementia, intellectual disability, isolation, prior victimisation | Self-neglect, failing discharge supports, financial control by another |
Watch outAsk directly and specifically. 'I need to ask whether you have had thoughts of harming yourself' is not a suggestion, and patients expect the question. Ask about ideation, then plan, then intent, then means, then protective factors, and record the answers in the patient's words. A question never asked is a risk never assessed, and a half-asked question produces an answer you cannot interpret.
Screening instruments
| Instrument | Target | Use and caveat |
|---|---|---|
| PHQ-2, then PHQ-9 | Depression | Two screening questions followed by nine severity items; 5, 10, 15 and 20 mark mild, moderate, moderately severe and severe. Item 9 flags suicidality and must be followed by direct assessment |
| GAD-7 | Anxiety | Seven items; a score of 10 or more is the usual cut-off, and it performs poorly at separating anxiety from depression |
| AUDIT | Hazardous and harmful alcohol use | Ten items; a score of 8 or more in men and 6 or more in women is the usual screen-positive; a brief version, AUDIT-C, is three items |
| CAGE and CAGE-AID | Alcohol, and with drug use | Four questions; two or more positive is a positive screen — quick for bedside use, insensitive to drinking pattern |
| HoNOS and HoNOSCA | Clinical severity in services; child and adolescent version | Twelve-rated-item clinician scale for tracking change; not diagnostic |
| PANSS and BPRS | Psychotic symptoms in research and follow-up | Interview-based severity measures; indispensable in trials and the longer-term monitoring of patients, not for diagnosis |
| Y-BOCS | Obsessive-compulsive symptom severity | Ten items; the standard for severity tracking and for defining response in treatment studies |
Common mistakes
Watch outRecording a thought disorder as 'confused'. Thought form is assessed from the patient's speech, judged against the background of their usual speech. A patient who is slow because of depression, retarded by antipsychotic side effects, dysarthric from a stroke, or recoiling from the question is not showing formal thought disorder, and each of those has a different treatment.
Watch outUsing a symptom as a noun for a person. Write 'a patient experiencing auditory hallucinations', not 'a hallucinator' or 'an auditory hallucinator'. Categorical shorthand does real damage: it predicts willingness to engage, to employ, and to avoid discrediting the person — and it is precisely the linguistic habit that stigma research measures.
Watch outWriting 'insight impaired' as a conclusion. It is a description of three separable components, and it can change in weeks with treatment. A patient who refused admission on Tuesday may be asking for help on Thursday, and the note must make clear which components were impaired when.
Check your understanding
A 78-year-old is admitted overnight with inattention, fluctuating consciousness, poor memory and visual misidentifications. How should the mental status examination be interpreted?
Which four abilities define mental capacity for a specific decision?
Why should a screening questionnaire never be recorded as a psychiatric diagnosis?
Summary and next
Classification exists to communicate, to select treatment and to research, and it is realised in two systems — the WHO's ICD-11 and the APA's DSM-5 — with different purposes and different boundaries. The psychiatric assessment combines a structured history with a mental status examination whose domains are separately recorded, and it closes with risk formulated from static and dynamic factors. Screening instruments quantify severity; they do not make diagnoses.
Next: aetiology and the biopsychosocial model. You will see how genes, brains, minds and social worlds combine in ways that no single level of explanation captures — and why that structure is the justification for a psychiatric formulation rather than a bare diagnosis.
Multiple choice
Why does psychiatry need two major classifications, ICD-11 and DSM-5, rather than one?
Which of the following is most characteristic of formal thought disorder rather than dysarthria?
A patient is asked what day it is, to count backwards from twenty, and to name three objects after five minutes. She fails the second task only. What does this pattern indicate?
Which statement about insight, judgement and capacity is correct?
Which pair is the best-known brief alcohol screening tools, one using four questions and the other ten?