Any discussion of the classification of psychiatric disorders should begin with the frank admission that any definitive classification of disease must be based on aetiology. Until we know the causes of the various mental illnesses, we must adopt a pragmatic approach to classification that will best enable us to care for our patients, to communicate with other health professionals and to carry out high-quality research.

In physical medicine, syndromes existed long before the aetiology of these illnesses were known. Some of these syndromes have subsequently been shown to be true disease entities because they have one essential cause. Thus, smallpox and measles were carefully described and differentiated by the Arabian physician Rhazes in the tenth century. With each new step in the progress of medicine, such as auscultation, microscopy, immunology, electrophysiology and so on, some syndromes have been found to be true disease entities, while others have been split into discrete entities, and others still jettisoned. For example, diabetes mellitus has been shown to be a syndrome that can have several different aetiologies. On that basis, the modern approach to classification has been to establish syndromes in order to facilitate research and to assist us in extending our knowledge of them so that, ultimately, specific diseases can be identified. We must not forget that syndromes may or may not be true disease entities, and some will argue that the multifactorial aetiology of psychiatric disorder, related to both constitutional and environmental vulnerability, as well as to precipitants, may make the goal of identifying psychiatric syndromes as discrete diseases an elusive ideal.

Syndromes and Diseases

A syndrome is a constellation of symptoms that are unique as a group. It may of course contain some symptoms that occur in other syndromes also, but it is the particular combination of symptoms that makes the syndrome specific. In psychiatry, as in other branches of medicine, many syndromes began with one specific and striking symptom. In the nineteenth century, stupor, furore and hallucinosis were syndromes based on one prominent symptom.

Later, the recognition that certain other signs and symptoms co-occurred simultaneously led to the establishment of syndromes. Korsakoff’s syndrome illustrates the progression from symptom to syndrome to disease. Initially, confabulation and impressionability among alcoholics were recognised by Korsakoff as significant symptoms. Later, the presence of disorientation for time and place, euphoria, difficulty in registration, confabulation and ‘tram-line’ thinking were identified as key features of this syndrome. Finally, the discovery that in the alcoholic amnestic syndrome there was always severe damage to the mammillary bodies confirmed that Korsakoff’s psychosis (syndrome) is a true disease with a neuropathological basis.

Sometimes, the symptoms of the syndrome seem to have a meaningful coherence. For example, in mania the cheerfulness, the over-activity, the pressure of speech and the flight of ideas can all be recognised as arising from the elevated mood. The fact that we can empathise with and understand our patients’ symptoms by taking account of the context in which they have arisen has led to the distinction between those symptoms that are primary, that is, are the immediate result of the disease process, and secondary symptoms, which are a psychological elaboration of, or reaction to, primary symptoms. The term ‘primary’ is also used to describe symptoms that are not derived from any other psychological event.

Early Distinctions

The first major classification of mental illness was based on the distinction between disorders arising from disease of the brain and those with no such obvious basis, that is, functional versus organic states. These terms are still used, but as knowledge of the neurobiological processes associated with psychiatric disorders has increased and led to greater nuance, their original meaning has been lost. Schizophrenia and manic depression are typical examples of functional disorders, but the increasing evidence of the role of genetics and of neuropathological abnormalities shows that there is at least some organic basis for these disorders. Indeed, the category of ‘organic mental syndromes and disorders’ was renamed as ‘delirium, dementia and amnestic and other cognitive disorders’ in the Diagnostic and Statistical Manual of Mental Disorders (DSM−IV) (American Psychiatric Association, 1994), so that the recognition of the role of abnormal brain functioning is not confined to dementia and delirium only. In their literal meaning, these categories of classification (i.e., organic versus functional) are absurd, yet they continue to be used through tradition.

Organic Syndromes

The syndromes due to brain disorders can be classified into acute, subacute and chronic. In acute organic syndromes, the most common feature is alteration of consciousness, which can be dream-like, depressed or restricted. This gives rise to four subtypes: namely, delirium, subacute delirium, organic stupor or torpor, and the twilight state. Disorientation, incoherence of psychic life and some degree of anterograde amnesia are features of all of these acute organic states. In delirium, there is a dream-like change in consciousness so that the patient may also be unable to distinguish between mental images and perceptions, leading to hallucinations and illusions. Usually there is severe anxiety and agitation. When stupor or torpor is established, the patient responds poorly or not at all to stimuli and after recovery has no recollection of events during the episode. In subacute delirium, there is a general lowering of awareness and marked incoherence of psychic activity, so that the patient is bewildered and perplexed. Isolated hallucinations, illusions and delusions may occur and the level of awareness varies but is lower at night-time. The subacute delirious state can be regarded as a transitional state between delirium and organic stupor. In twilight states, consciousness is restricted such that the mind is dominated by a small group of ideas, attitudes and images. These patients may appear to be perplexed but often their behaviour is well ordered and they can carry out complex actions. Hallucinations are commonly present. In organic stupor (torpor), the level of consciousness is generally lowered and the patient responds poorly or not at all to stimuli. After recovery, the patient usually has amnesia for the events that occurred during the illness episode.

In addition, there are organic syndromes in which consciousness is not obviously disordered, for example organic hallucinosis due to alcohol abuse, which is characterised by hallucinations, most commonly auditory and occurring in clear consciousness, as distinct from the hallucinations of delirium tremens that occur in association with clouded consciousness. Amnestic disorders, of which Korsakoff’s syndrome is but one, also belong in this group of organic disorders, and are characterised primarily by the single symptom of memory impairment in a setting of clear consciousness and in the absence of other cognitive features of dementia.

The chronic organic states include the various dementias, generalised and focal, as well as the amnestic disorders. Included among the generalised dementias are Lewy body disease, Alzheimer’s disease and so on, while the best-known focal dementia is frontal lobe dementia (or syndrome). The latter is associated with a lack of drive, lack of foresight, inability to plan ahead and an indifference to the feelings of others, although there is no disorientation. Some patients may also demonstrate a happy-go-lucky carelessness and a facetious humour, termed Witzelsucht, whereas others are rigid in their thinking and have difficulty moving from one topic to the next. The most common cause is trauma to the brain such as occurs in road traffic accidents. The presence of frontal lobe damage may be assessed psychologically using the Wisconsin Card Sorting Test or the Stroop Test. Amnestic disorders are chronic organic disorders in which there is the single symptom of memory impairment; if other signs of cognitive impairment are present (such as disorientation or impaired attention), then the diagnosis is dementia. The major neuroanatomical structures involved are the thalamus, hippocampus, mammillary bodies and the amygdala. Amnesia is usually the result of bilateral damage, but some cases can occur with unilateral damage. Further, the left hemisphere appears to be more critical than the right in its genesis.

Functional Syndromes

Functional syndromes (or disorders), a term seldom used nowadays, refers to those syndromes in which there is no readily apparent coarse brain disease, although increasingly it is recognised that some finer variety of brain disease may exist, often at a cellular level.

For many years, it was customary to divide these functional mental illnesses into neuroses and psychoses. The person with neurosis was believed to have insight into his illness, with only part of the personality involved in the disorder, and to have intact reality testing. The individual with psychosis, by contrast, was believed to lack insight, had the whole of his personality distorted by the illness and constructed a false environment out of his distorted subjective experience. Yet, such differences are an oversimplification, since many individuals with neurotic conditions have no insight, and far from accepting their illness, may minimise or deny it totally, whereas people with schizophrenia may seek help willingly during or before episodes of relapse. Moreover, personality can be changed significantly by non-psychotic disorders such as depressive illness, while it may remain intact in some people with psychotic disorders, such as those with persistent delusional disorder.

Reference JaspersJaspers (1962) regarded the person with neurosis as an individual who has an abnormal response to difficulties in which some specific defence mechanism has transformed their experiences. For example, in conversion and dissociative disorders (formerly hysteria), the mechanism of dissociation is used to transform the emotional experiences into physical symptoms. Since we can all use this mechanism, the differences between the neurotic person and the normal person is one of degree. Reference SchneiderSchneider (1959) has suggested that neuroses and personality disorders are variations of human existence that differ from the norm quantitatively rather than qualitatively. However, this view of the neuroses breaks down when obsessive–compulsive disorder is considered, since the symptoms are not variations of normal but differ qualitatively from normal behaviours.

Over time, the use of the terms ‘neurotic’ and ‘psychotic’ changed, and instead of describing symptoms, particularly symptom types such as hallucinations or delusions, in the psychotic person they were used to distinguish mild and severe disorders or to distinguish those symptoms that are ego-syntonic (i.e., creating no distress for the person or compatible with the individual’s self-concept or ego) from those that are ego-dystonic (i.e., causing distress and incompatible with the person’s self-concept). Some practitioners also used the word ‘neurotic’ as a term of opprobrium. Owing to the confusion that abounded in the various uses of these terms, DSM-IV excluded the term ‘neurosis’ totally from its nomenclature, and this has continued in DSM-5. The International Classification of Diseases (ICD-10) (World Health Organization, 1992) named a group of disorders ‘neurotic, stress-related and somatoform disorders’. However, ICD-11 does not use this term and the aforementioned group is now divided between ‘anxiety and fear-related disorder’, ‘disorders of bodily distress’ and ‘disorders specifically associated with stress’.

Personality Disorders and Psychogenic Reactions

The status of personality disorder vis-à-vis other psychiatric disorders was historically regarded differently in the English-speaking world compared with the rest of the world. In the English-speaking world, it was customary to separate the neuroses from personality disorders, but in the German-speaking countries, epitomised by Schneider, the neuroses were regarded as reactions of abnormal personalities to moderate or mild stress and of normal personalities to severe stress. This difference was reflected in the approach of DSM-IV in placing personality and other disorders (e.g., major depression) on separate axes, while ICD-10 did not. The DSM has now also removed the multiaxial approach in its entirety, including the removal of the assessment of functioning.

Psychogenic reactions constituted reversible prolonged psychological responses to trauma, the reactions being the consequence of the causative agent on the patient’s personality. Thus, acute anxiety and hysteria were considered to be varieties of psychogenic reactions provoked by stress and determined by personality and cultural factors. Sometimes, the stress was believed to cause psychotic reactions, termed symptomatic or psychogenic psychoses: for example, the person with a paranoid personality who, in light of ongoing marital difficulties, begins to suspect his wife’s fidelity, finally becoming deluded about this. The idea of delusional states that were not due to functional psychoses was treated with scepticism by English-speaking psychiatrists, but had adherents in Scandinavia, particularly in what were termed psychogenic psychoses. These have gained increasing acceptance and were included in ICD-10 and retained in ICD-11 as acute and transient psychotic disorders. In DSM-IV and 5, they are called brief psychotic disorder. In both, they are regarded as being associated with a stressor although this is not essential. They are classified in the group of disorders entitled ‘Schizophrenia and other primary psychotic disorders’. Other psychogenic reactions such as dissociation and conversion disorders are now renamed as dissociative disorders in both DSM-5 and ICD-11. The word ‘psychogenic’, like ‘neurotic’, has been eliminated from recent iterations of both classifications.

Modern Classifications

Two modern systems of classification are in use. The DSM is used mainly in the United States and is prepared by the American Psychiatric Association every few years. The International Classification of Diseases (ICD) is a World Health Organization document and covers all medical conditions; one chapter is devoted to mental and behavioural problems. International Classification of Diseases is in use throughout the world, although the DSM is often used in research, including drug trials, because each disorder is operationally defined and these criteria can be applied when attempting to obtain homogenous populations, as is required in drug trials for the treatment of certain conditions.

DSM-I, published by the American Psychiatric Association, first appeared in 1952, and since then it has evolved significantly, to the extent that DSM-5 includes large amounts of detail concerning each syndrome and, owing to its rigorous adherence to operational definitions for each disorder, is suitable for use in both clinical practice and research.

However, the DSM system is considerably less user-friendly than the ICD, since it is viewed as Procrustean by its critics. Interestingly, the billing codes for Medicare in the United States are mandated to follow the ICD system rather than their own DSM. The ICD, by contrast, is more clinically orientated and is not so rigid in its definitions, eschewing operational definitions in favour of general descriptions. It allows clinical judgement to inform diagnoses, but this freedom makes it unsuitable for research purposes, necessitating the development of separate research diagnostic criteria. Thus, different versions of ICD-10 now exist; these include the clinical version (World Health Organization, 1992), a version with diagnostic criteria for research (World Health Organization, 1993) (which resembles DSM in its use of detailed operational criteria) and a version for use in primary care (ICD-10−PC; World Health Organization, 1996), the latter consisting of definitions for twenty-five common conditions as well as a shorter version of six disorders for use by other primary care workers. Management guidelines incorporate information for the patient as well as details of medical, social and psychological interventions. Finally, assistance on when to refer for specialist treatment is provided. DSM-5 was published in 2013, and ICD-11 has been officially in use since 2022 (American Psychiatric Association, 2013). The historical timeline of the DSM and ICD systems are shown in Table 1.1.

Comparison of DSM-5 and ICD-11

The DSM-5 and ICD-11 are syndrome-based classifications. This means that they are based on commonly co-occurring symptoms and not on aetiology, psychobiology or prognosis. Hence, the removal of terms such as ‘psychogenic’, ‘neurotic’, ‘functional’ and ‘psychosomatic’, which have connotations for the cause of particular disorders. Apart from the few conditions classified under the stress-related rubric and substance misuse, both classifications are aetiology-free in thinking. It was envisaged by the authors of DSM-IV that by the time publication of DSM-5 was ready, our knowledge of the biological and genetic underpinnings of many psychiatric disorders would have increased to the extent that classification based on the underlying psychobiology would be possible.

Yet, as work commenced on DSM-5, Dr Gerard Kupfer, who chaired the task force charged with its development, commented:

Alas, the promise of a new approach to classification was not realised and both DSM-5 and ICD-11 continue to be based on predominant symptoms and syndromes, not disease entities.

An accompaniment to ICD-11 will be the publication of Clinical Guidelines and Diagnostic Requirements (or Guidelines) (CDDR or CDDG) for all the listed psychiatric disorders. This will include expanded clinical descriptions, differential diagnosis, boundaries with other disorders as well as cultural aspects of symptoms and their distinction from normal emotional responses. This may assist in preventing the over-diagnosis of mood and anxiety disorders.

Interview Schedules

In order to carry out epidemiological studies in which diagnoses are standardised, Diagnostic Interview Schedules (DIS) were developed to meet the criteria for the ICD and the DSM diagnoses.

The Structured Clinical Interview for DSM-5 (SCID-5) was developed from SCID-IV. There are now four versions: a clinical version (SCID-5-CV) (Reference 10First, Williams, Karg and SpitzerFirst et al., 2016a), a research version (SCID-5-R) (Reference First, Williams, Karg and SpitzerFirst et al., 2015a), a clinical trials version (SCID-5-CT) (Reference First, Williams, Karg and SpitzerFirst et al., 2015b) and a personality disorders version (SCID-5-PD) (Reference First, Williams, Benjamin and SpitzerFirst et al., 2015). There is also a screening version (SCID SPQ) (Reference First, Williams, Benjamin and SpitzerFirst et al., 2016) and one for personality disorder, termed an alternate model for personality disorders (SCID-AMPD). SCID-AMPD differs from SCID-5-PD as the former allows for a dimensional assessment of personality, while the latter is based on the traditional categorical model of personality disorder. This is a semi-structured interview since it allows some latitude in its administration to elaborate on questions.

Another schedule used to evaluate diagnoses based on the DSM criteria is the Composite International Diagnostic Interview (CIDI) (Reference Robins, Wing and WittchenRobins et al., 1989). It developed from the DIS (Reference Robins, Helzer, Orvaschel, Eaton and KesslerRobins et al., 1985) but unlike SCID is not a semi-structured interview. Instead, it is standardised and is suitable for use with lay interviewers. No clinical judgement is brought to bear in rating the symptoms since questions are asked in a rigid and prescribed manner. The questions are clearly stated to elicit symptoms, followed by questions about frequency, duration and severity. The only judgement the interviewer has to make is whether the respondent understood the question, and if not, it is repeated verbatim. Composite International Diagnostic Interview is available in computer format also and so can be self-administered. It was later explained to facilitate ICD-10 diagnoses. This resulted in the World Health Organization World Mental Health-CIDI (WHO WMH-CIDI) (Reference Kessler and UstunKessler and Ustun, 2004). It does not include personality disorders and only evaluates lifetime and twelve-month disorders. It has a screening section and can also be used in modular form for evaluating specific disorders. It has sections on functioning, services sought and family burden. CIDI-5 is currently being developed to coincide with the use of DSM-5.

As with Schedule for Clinical Assessment in Neuropsychiatry (SCAN) (see in the next paragraph), the symptoms are then entered into a computer algorithm for diagnosis according to ICD or DSM. The advantage of this approach is that it is cheaper than using semi-structured interviews, since lay people can be trained in its use. However, the absence of clinical judgement is an obvious disadvantage that has resulted in its validity being questioned. Some recent reviews question the prevalence for some psychiatric disorders obtained using standardised interviews such as CIDI and suggest that the high rates identified in some studies require revision downwards (Reference Regier, Kaelber and RaeRegier et al., 1998). These different approaches are discussed in detail by Reference Brugha, Bebbington and JenkinsBrugha et al. (1999) and Reference Wittchen, Ustun and KesslerWittchen et al. (1999).

In Europe, SCAN (Reference Wing, Babor and BrughaWing et al., 1990) has evolved from the older Present State Examination (PSE) (Reference Wing, Cooper and SartoriusWing et al., 1974). Schedule for Clinical Assessment in Neuropsychiatry itself is a set of instruments aimed at assessing and classifying psychopathology in adults. The four instruments include PSE-10 (the tenth edition of the PSE); the SCAN glossary, which defines the symptoms; the Item Group Checklist for symptoms that can be rated directly (e.g., from case notes); and the Clinical History Schedule. Schedule for Clinical Assessment in Neuropsychiatry provides diagnoses according to both ICD-10 and DSM-IV criteria. The interview itself is semi-structured, the aim being to encapsulate the clinical interview while minimising its vagaries. There are probe questions with standard wording to elucidate the psychopathological symptoms, defined in the glossary and accompanied by severity ratings. Where there is doubt, the interviewer can proceed to a free-style interview to clarify the feature further and may, if necessary, include the patient’s phraseology in questioning to enhance clarity. It is designed for use by psychiatrists or clinical psychologists, thereby utilising clinical interviewing skills in evaluating each symptom. The symptoms ratings are then entered into a computer algorithm and a computer diagnosis obtained according to either classification.

The role of the interviewer is to rate symptoms rather than make diagnoses. Schedule for Clinical Assessment in Neuropsychiatry can generate a current diagnosis, a lifetime diagnosis or a representative episode diagnosis. The use of mental health professionals in interviewing with SCAN makes this an expensive method but has the advantage of approximating the ‘gold standard’ diagnosis achieved by clinical interview. Schedule for Clinical Assessment in Neuropsychiatry pays little attention to personality disorder and it is only in the clinical history section that details of diagnoses that are not covered in PSE-10 are recorded, usually from other sources of information. It is unclear if there will be changes to SCID and SCAN so that epidemiological research can follow the publication of ICD-11.