Abstract
The introduction of DSM-III in 1980 [1] and of ICD-10 in 1992 [2, [3], [4]] presented a major innovation in clinical psychiatry. It was a paradigmatic change from nosological to syndromatical diagnostics and classification. The previous classifications were based on prototypical descriptions of psychiatric illnesses with unknown but presumed aetiologies, with ill-defined boundaries and low reliability, impeding research and the quality of clinical psychiatry. The DSM-III and ICD-10 introduced purely descriptive, non-aetiological classifications with operationally defined categories based on diagnostic criteria, with proven reliability in major national and international trials [5, [6]]. DSM-III became an immediate success in research and clinical psychiatry, and the use spread all over the world in the decade following its introduction, with DSM-III-R taking over in 1987 [7], and DSM-IV in 1994 [8, [9]].
This success had an impact on the psychiatric revision of ICD-9 [10] in the same decade. The ICD-10, however, could not just copy the DSM-III because of the World Health Organization's (WHO) responsibility to provide a worldwide consensus classification applicable and acceptable in all the member states. In spite of efforts to bring ICD-10 and DSM-IV in line as far as possible, certain principal differences remained of which the avoidance of social dysfunction criteria in ICD-10 and the hierarchical order preventing comorbidity diagnoses in ICD-10 are the most conspicuous. Among the clinical criteria, DSM requires chronic course for schizophrenia, has stricter criteria for schizoaffective disorders, and allows schizophrenia-like psychotic symptoms in affective disorders, which count for the major differences between DSM-IV and ICD-10 in these areas. Otherwise, the differences are minor, mainly affecting moderate or mild, threshold-near cases, which show different distributions in the two systems.
As to the clinical utility, the two systems are quite comparable, following similar criteria-based diagnostic rules. Experiences with one of the systems will therefore apply also to the other system, provided that the rules have been followed as intended. Denmark was among the first four countries in which ICD-10 was taken into official use, from January 1994, instead of ICD-8 which had been used for the previous 25 years. The present reflections thus are based upon experiences with ICD-10 psychiatric diagnostic criteria for more than 4 years of daily clinical use. A complete national registration of psychiatric diagnoses obliges Danish psychiatrists to provide ICD-10 diagnoses for inpatients and out-patients from all psychiatric hospital departments and community centres in Denmark.
Implementation of ICD-10
Because of the major changes in concepts and diagnostic practice, the transition from ICD-8 [11] to ICD-10 had to be well prepared and carefully managed to ensure that the rules of the ‘new psychiatry’ would be well known and correctly used [12]. A Danish translation and adaption of the WHO ICD-10, chapter V was provided with Danish and Latin names for all the diagnostic categories. An edition including the diagnostic criteria was published in a handy spiral-bound pocketbook format, corresponding to the WHO Pocket Guide to the ICD-10 Classification of Mental and Behavioural Disorders, with Glossary and Diagnostic Criteria for Research [13]. The edition has become a veritable best-seller, used not only by practically all Danish psychiatrists, but also by psychiatric personnel, psychologists and social security workers. A number of introductory courses were given, and within the first year, about 80% of Danish psychiatrists in this way received re-education in psychiatric classification according to ICD-10. Over the following 3 years, follow-up courses were arranged each year, discussing diagnostic problems, presenting preliminary results and experiences, and providing case presentations for diagnostic evaluation. The last follow-up course in 1997 was extended over half a year by sending out batches of case summaries each month for evaluation to be discussed at major meetings at the end of the course. These follow-up courses have served to ensure the quality of diagnostic usage among Danish psychiatrists. To promote the proper use of the symptom criteria included in the ICD-10 diagnostic criteria, a Danish edition was published of the Short Version of the Schedules for Clinical Assessment in Neuropsychiatry (SCAN) [14] Present State Examination, for clinical use in conjunction with the ICD-10 [15], which has an even greater distribution than the ICD-10 edition. At the same time, courses in ICD-10 and Present State Examination have become an obligatory part of the training curriculum for psychiatric trainees, so that ICD-10 eventually has become an integrated part of Danish psychiatry. Furthermore, a Danish edition of the ICD-10 Version for Primary Health Care [16] has been provided for general practitioners, and introductory and training courses have been held for psychologists, social workers, mental health personnel, and general practitioners, even including a course for high-school teachers in psychology.
Because of the major changes in the concepts, it was expected that the implementation of ICD-10 would be met with some hesitation or reluctance. On the contrary, the ICD-10 classification was well received, particularly by the younger generations among whom it was met almost with enthusiasm. Here it seemed to meet a need for a more precise and systematic classification, replacing the previous more ambiguous and diffuse, mainly psychodynamically inspired categories, following various trends and fashion-like ideas like borderline states or psychological crises, which inspire the psychiatrist to undisciplined thinking with interpretative shortcuts to diagnostic categories of low reliability. It was felt that ICD-10 gave a lift of quality to clinical psychiatry, reinforcing the identity of the psychiatric discipline as belonging to the medical sciences, counteracting recent developments to place psychiatry under the administration of the social authorities.
The clinical use of ICD-10
The National Psychiatric Register in Denmark allows an evaluation of the usage of ICD-10. The registration rules require at least one diagnosis for every inpatient admission and outpatient treatment period, the main diagnosis of the disorder for which treatment has been provided. If this disorder was in fact due to a basic psychiatric disorder like alcohol dependence in a patient with schizophrenia, two main diagnoses may be registrated, one for the disorder for which treatment was given to be used by health authorities for medical statistics, and the other for the basic disorder, in this case schizophrenia, for the psychiatric register for epidemiological research evaluation. Furthermore, subsidiary diagnoses may be given for other mental conditions of secondary significance, or for somatic disorders. After 2 years of use, a psychiatric register evaluation was performed including all the ICD-10 diagnoses given for in- and outpatient treatment courses [Munk-Jørgensen P: personal communication]. The majority only received one diagnosis. Among the main diagnoses, only a limited number of categories had been used. On an ICD-10 code three character level (Fxx, letter F and two digits) only about one-quarter of possible categories had been used for more than 1% of the patients, a quarter was hardly used at all, and the remaining categories only rarely. The categories used most frequently were F20 schizophrenia used for 17% of the patients, followed by F10 alcohol use disorders for 11%, and F60 personality disorders, mainly borderline personality disorder, for 10%. F43 stress-related disorders, mainly adjustment disorders, was used for 8%, F31 bipolar disorders for 7%, and F32 and F33 single and recurrent depressions for about 6% each. F22 persistent delusional disorders, F23 acute and transient psychotic disorders, and F41 other anxiety disorders, were each used in a low percentage of the patients. For the registration, four character diagnoses (Fxx.x) are requested, and for these, the most frequent were F20.0 paranoid schizophrenia, F10.2 alcohol dependence, F60.3, borderline personality disorder, F43.2 adjustment disorders, whereas the affective disorders were divided among various types and severities. Subdivisions using five characters (Fxx.xx) are provided for a number of categories but were hardly used at all.
The German ICD-10 coordination and reference centre in Lübeck has made an evaluation of the usage of ICD-10 diagnostic categories [Dilling H: personal communication]. It is based on the results from eight university hospitals in Germany and 11 psychiatric hospitals and departments worldwide, the majority of the patients being hospital inpatients (90%). The results were quite similar. The majority (about 85%) provided only one diagnosis. On the three character level (Fxx), the 10 most often used categories represented 70% of all main diagnoses, in order of frequencies: F20 schizophrenia, mainly paranoid schizophrenia, F10 alcohol use disorders, mainly alcohol dependence and alcohol withdrawal, F32 single depressive episode, F31 bipolar disorder, F43 stress-related disorders, mainly adjustment disorders, F33 recurrent depression, F60 personality disorders, mainly borderline personality disorder, F23 acute and transient psychotic disorders, F25 schizoaffective disorders, and F19, multiple substance use disorders. On the four-character level (Fxx.x), the 10 most often used categories accounted for no less than 40% of the main diagnoses. About one-quarter of possible categories were used in less than 0.1% or not at all.
Both the Danish and the German transcultural survey of the use of ICD-10 diagnoses thus reflect an aspect of the clinical utility of the diagnostic categories. The use seems to be limited to categories with clinical relevance for treatment and prognostic outcome. The use varied with the type of patient with higher percentages for psychotic disorders among hospital inpatients, and of neurotic and stress-related disorders and personality disorders among outpatients. In primary health care, the most frequent disorders encountered are depressive, anxiety and somatoform disorders as evidenced in the transcultural study by Üstun et al. [17]. The evaluations mentioned may, however, also reflect levels of insufficient training in reliable use of the diagnostic criteria indicated by the disproportionally high percentages of adjustment disorders and borderline personality disorders among hospital inpatients, probably due to reluctance to give up the previously popular categories of crises and borderline personalities.
Problems and difficulties
One of the major difficulties in the Danish setting proved to be a difficulty of thinking on different logical levels. One level is the purely phenomenological and descriptive of syndromatical diagnostics for ICD-10 categories and the other, the understanding and interpretation of symptoms nosologically in accordance with prevailing explanatory hypotheses for the choice of therapy. The difficulty manifested itself in two ways. For the older generation of psychiatrists, the habitual way of nosological thinking created difficulties in keeping the syndromatical diagnostic criteria separate from the nosological interpretation resulting in diagnostic short-cuts, relying more on what they felt must be the case, than observed diagnostic symptoms. The younger generation, on the other hand, took the syndromatical thinking to their hearts to a degree that they almost forgot to apply nosological thinking in order to provide a comprehensive evaluation of aetiological factors leading to the formulation of a treatment plan, for which additional information is invariably necessary. This may result in a mechanical application of diagnostic criteria with a temptation to treatment shortcuts (e.g. ordering antidepressants for all F32 and F33 depressive disorders). While the elder generation of psychiatrists are focused on applying the correct individual treatment for what may prove to be a wrong ICD-10 diagnosis, the younger generation of psychiatrists are focused on applying correct and reliable ICD-10 diagnoses, for which they may apply stereotyped and poorly adapted treatment programs.
Another major difficulty has proved to be the lack of the possibility of providing tentative or provisional diagnoses. For some diagnostic categories it is only possible to give a full diagnostic evaluation after some time of observation. For example the diagnosis of dementia requires that the symptoms should have been present for at least 6 months. For F23, acute and transient psychotic disorders, the diagnosis can only be applied if the disorders remit within the time limits of the criterion for their maximum duration. This does not harmonise well with the obligation in clinical psychiatry to provide a diagnostic evaluation and a treatment plan within the first week of admission. Of course it is often possible to apply unspecified categories which, however, is unsatisfactory when the symptoms clearly point to a more specific diagnostic category.
Among the various diagnostic groups, several problems have appeared affecting the clinical utility of the diagnostic classification. A number of problems regarding specific categories will be mentioned in the following.
For the F0 group, the general criteria for F06, other mental disorders due to brain damage and dysfunction, requires a presumed relationship between the development and excercabation of the underlying disease or damage and the mental symptoms, which, however, sometimes may appear after many years (e.g. in psychotic disorders related to an epileptic disorder). Furthermore, the criteria require ‘recovery or significant improvement in the mental disorder following removal or improvement of the underlying presumed cause’, which cannot be met in many cerebral disorders of chronic nature.
For F1, substance abuse disorders, psychotic disorders due to substance use may have a duration that exceeds the allowed 6 months, which then will change the diagnosis to another category (e.g. 10.50, schizophrenia-like disorder due to alcohol use, to a F20 schizophrenic disorder). At the same time, it is difficult to differentiate between drug use psychotic disorders, and genuine psychotic disorders precipitated by drug use. If, for example, a schizophrenia-like disorder occurred during or within 2 weeks of cannabis use and persisted for more than 48 h, leaving out the possibility of an intoxication psychotic disorder, then we have the diagnosis of psychotic schizophrenia-like disorder due to cannabis use, but it could as well be a cannabis-precipitated schizophrenic disorder. Provided that no further amounts of the drug are taken, for how long shall we observe the disorder before a decision can be made? This is a diagnostic problem frequently met in a society with quite extensive drug abuse among mentally disordered persons. The criteria are not explicit, but we have chosen a 1 month period of drug-free observation to decide between the two possibilities. It is difficult enough to keep a drug-dependent patient free of drugs for 1 month, even in a locked ward, to say nothing of a 6 months' period after which the ICD-10 criteria will change the diagnosis to schizophrenia. The same problem applies to F1x.75, late onset psychotic disorder, which has the same criteria except for the late onset.
Among the F2 disorders, problems are created by the sharply defined duration criteria necessary to avoid overlapping between the categories. For a patient meeting the diagnostic criteria for the schizophrenia syndrome, except for duration, this means that the initial tentative diagnosis will be of F23.2, acute schizophrenia-like psychotic disorder, but if the symptoms remain for more than 1 month, the diagnosis will change to schizophrenia. If proper treatment ameliorates the symptoms within a month, the diagnosis will be acute schizophrenia-like psychotic disorder. After discharge, the patient may stop treatment and relapse, be readmitted and again have a remission within 1 month. This pattern may possibly repeat itself several times, and he will thus not get a diagnosis of schizophrenia unless one of the episodes exceeds the duration of 1 month. The patient may even alternate between the diagnostic categories of schizophrenia and acute schizophrenia-like psychotic disorder, depending on the length of the symptomatic episodes. In these cases, the division appears arbitrary and unsatisfactory. The very same considerations apply to the 3 months' duration limit between F22 persistent delusional disorder, and F23.3 acute predominantly delusional psychotic disorder.
For F25 schizoaffective disorder, the difficulty is the unproven validity of the diagnosis. The diagnosis is trans-sectional applying only to the actual disorder, not taking the course into consideration, particularly the nature of previous episodes, which may be clearly affective, bipolar or depressive, or clearly schizophrenic. The criteria require that both the schizophrenic and the affective symptoms must be prominent. Where are we to categorise disorders in which one of the symptom groups are clearly less prominent, although they otherwise both fulfil the diagnostic criteria? If the schizophrenia-like symptoms have appeared before the affective, they can be categorised under schizophrenia, but cases in which the affective symptoms appeared first are difficult to categorise. Affective disorders do not allow schizophrenia-like symptoms at all, so the only possibilities seem to be F28 (other psychotic disorders) or F38.8 (other mood disorders).
For F3 disorders the depressive syndrome may create some problems, particularly for cases of mild severity. It furthermore creates difficulties that the ICD-10 diagnostic guidelines and research criteria are not completely identical. According to the research criteria a depressive episode of mild severity requires at least four out of 10 symptoms, two of which must belong to the first three ‘nuclear’ symptoms. It is possible to combine the required four symptoms in many ways into syndromes, which may be quite dissimilar except for the presence of depressed mood or anhedonic mood in the form of loss of interest or pleasure. Furthermore, the symptoms are dimensional, appearing somewhere on a line between normality and severe pathology, and it may be a matter of temper or level of sensitivity to symptoms that will decide whether a depressive symptom is diagnosed as present or not, which again will affect the reliability of the threshold of a depressive episode. Moderate or severe episodes are defined by the number of symptoms alone, whereas the diagnostic guidelines also take intensity of symptoms into consideration. Even if there is a correlation between number of symptoms and severity, it is not uncommon to meet patients who are able to describe the presence of nine or 10 depressive symptoms, and at the same time, only is mildly affected and may be able to manage work or other social obligations. In contrast, a severely depressed patient with psychomotor retardation may present with only a few depressive symptoms, one of which may even be psychotic. Evaluation of severity of depression is therefore perhaps better left to clinical judgement like in DSM-IV.
It is further possible to subdivide mild and moderate depressive disorders as to the presence of a somatic or melancholiform syndrome, which, however, is rarely used. Finally, a severe depressive episode with psychotic symptoms may be subdivided as to the mood congruency of the psychotic symptoms. This is hardly used at all in the clinical use of ICD-10, where schizophrenia-like psychotic symptoms are excluded in affective disorders, in contrast to DSM-IV, which for mood disorders allows schizophrenic first rank symptoms and bizarre delusions, most often of mood-incongruent character.
Another difference between DSM-IV and ICD-10 concerns the threshold for depressive disorder. DSM-IV requires five out of nine depressive symptoms, of which at least one is either depressed mood or diminished interest or pleasure. The word ‘major’ in the title of DSM-IV in some countries can cause the threshold to appear stricter than in ICD-10, because ‘major’ has a connotation of something of greater significance or higher severity, if one does not realise that the DSM-IV major depression is subdivided into mild, moderate and severe degrees.
Similar threshold reflections may also to some degree apply to a hypomanic episode, particularly when past episodes for the diagnosis of bipolar disorder, differentially from recurring depressive episodes are reviewed.
For F4 disorders, the major problem is the question of comorbid syndromes, which in ICD-10 are precluded by the hierarchical principle, in the presence of, for example, a depressive or a psychotic disorder. Important information about coexisting syndromes of panic or phobic anxiety, obsessive-compulsive disorder or somatoform disorders are thus lost in the diagnosis, but all the same, have to be taken into consideration in the treatment plan. Furthermore, it is not easy to decide whether coexisting syndromes are exactly simultaneous or the F4 syndromes in fact preceeded a depressive episode of mild severity. This may even appear to nosological thinking to be secondary to the F4 syndrome, which will make a diagnosis of only a depressive disorder less meaningful. The possibility also exists that F4 syndromes are totally independent of possible F2 and F3 disorders in which case the F4 syndrome could be used as subsidiary diagnosis. The Danish adaptation of ICD-10 here has provided additional diagnostic rules for coexisting disorders without complete simultaneity. In the case of mutually dependent syndromes, the primary, first appearing syndrome is rated as the main diagnosis, and the secondary syndrome as a subsidiary diagnosis. In a patient with social phobia for 10 years and mild depression for the last 1 year, the main diagnosis thus will be social phobia with a subsidiary diagnosis of mild depressive disorder. In the case of independent diagnoses, the most severe or significant disorder will become the main diagnosis.
For the F5 diagnoses, the ICD-10 classification and criteria for eating disorders have caused some problems, particularly among child and adolescence psychiatrists, who have found it difficult to apply the diagnoses of a purely anorectic or bulimic disorder to their young patients. They clearly prefer the DSM-IV criteria and categories with their subdivision of anorexia nervosa into a restricting and a binge-eating/purging type. The latter will in ICD-10 be referred to atypical anorexia.
For F6, the personality disorders present one of the areas which are most difficult to apply with any satisfactory degree of reliability. The general criteria require that the deviation should be pervasive and life-long with onset in childhood or adolescence. This may be difficult to establish by information obtained from the patient and usually requires additional information from a key-person, preferably using a special interview [18]. Furthermore, the deviation should be stable, and this calls for some hesitation in diagnosing personality disorders, which thus may have a labelling effect. Another major problem is that many psychotic or affective disorder patients may give the impression of disordered personality traits, which may tempt the clinician to apply a personality disorder diagnosis without considering that the traits may be limited to the psychotic or depressive episode. Such personality traits very often correspond to the psychodynamic concept of borderline disturbance, which is mistaken for the ICD-10 diagnosis of F60.31 emotionally unstable personality disorder of the borderline type. This tendency is apparent from the extraordinarily high percentage in which it appears as the main diagnosis in the Danish Psychiatric Register, accounting for 10% of the main diagnoses. This is perhaps the most conspicuous area in which syndromatical and nosological thinking clashes, mainly because of the great popularity of the borderline concept among psychodynamically orientated psychiatrists.
For gender identity disorders and disorders of sexual preference, powerful minority movements for special preferences have persuaded two succeeding Danish ministers of health to prohibit registration of transvestism and sadomasochism, not realising that happy members of these movements never will present as patients with clinically meaningful disorders for diagnostic registration, leaving those who experience distress with their special preferences to be diagnosed as unspecified disorders of gender identity or sexual preference.
The childhood and adolescence psychiatric disorders F7, F8 and F9, are two areas have presented with major difficulties. Diagnosing personality disorders in adolescence is hardly possible because personality disorders have to be manifest for some years after adolescence to be diagnosed. The psychodynamically orientated child and adolescence psychiatrists easily identify the popular borderline traits in the adolescents, and therefore want to use the diagnosis of F60.31 even if it is against the rules. Another major problem is the diagnostic categorisation of what is called disorders of attention, motor control and perception, the DAMPsyndrome, without hyperkinetic disturbances, which therefore cannot be categorised among the ICD-10 F90 hyperkinetic disorders to which it logically belongs. This is in contrast to DSM-IV which has a subtype of attention deficit/hyperactivity disorder, predominantly inattentive type.
Multiaxial presentation
Both classification systems have multiaxial presentations of which the ICD-10 presentations have only recently been published [19]. It has three axes, one for clinical diagnoses including somatic diagnoses. Axis II assesses social dysfunction in various areas and globally. Axis III assesses contextual factors influencing the development and course of psychiatric disorders, including a number of categories from the ICD-10 Z-chapter of factors influencing health status and contact with health services. It is doubtful whether the ICD-10 multiaxial presentation will be used to the same extent as the DSM-IV multiaxial assessment. Most psychiatrists in busy clinical settings find it too cumbersome to apply more than one or two diagnoses to each patient, as evidenced by the minority who used more than one diagnosis in the register evaluations. In the Danish adaptation of ICD-10, the multiaxial presentation has been anticipated by the possibility of using so-called associated diagnoses, particularly from the mentioned chapter Z to indicate factors of presumed aetiological significance. For a few conditions, associated diagnoses are required by the health authorities, particularly for suicidal attempts by taking overdoses of medicaments which must be specified in this way. Otherwise, the use of associated diagnoses is optional, and on the whole they are used very rarely. The utility of the ICD-10 axis II may await the results of the ongoing revision of the WHO International Classification of Impairments, Disabilities and Handicaps.
Forensic psychiatry
In this area, the new ICD-10 diagnostic categories are no longer in harmony with the Danish legislation concerning mentally disordered patients. The legislation is based on previous concepts of psychotic states and insanity, and it is now necessary in court statements to specify whether the ICD-10 diagnostic category corresponds to the legal criteria for insanity or not (e.g. that a patient with a manic episode without psychotic symptoms must be considered insane in the understanding of the legislation, in cases concerning involuntary admission, enforced treatment or regarding the responsibility for criminal acts). A list of ICD-10 disorders corresponding to psychotic states and insanity in the previous classification has been provided for the Danish adaptation of ICD-10. Registration of legal consequences has similarly been supplied with codes under Z04.6 General Psychiatric Examination requested by Authority, with codes for subgroups like Forensic Psychiatric Observation, Sentence to Psychiatric Treatment, Sentence to Psychiatric Custody, or Other Legal Examination or Consequences.
Concluding comments
A correct diagnosis is a prerequisite for proper treatment. In spite of the problems mentioned above, both DSM-IV and ICD-10 and their diagnostic criteria have been a major improvement, providing accurate, well-defined, and highly reliable diagnoses in clinical psychiatry, ensuring the quality of clinical work. Explicit criteria also make psychiatric diagnoses intelligible to lay people including patients and relatives, removing the mystification of psychiatric diagnoses based on subjective impressions like ‘schizophrenia-feeling’. This has improved the communication with the patients and their relatives, and also with the social and legal authorities, which now have become able to check the premisses for the psychiatric diagnoses. Psychiatric classification is no longer just a coding convention for statistical use, but has become an integrated part of the conceptual framework of clinical psychiatry. This has clearly reinforced the identity of the discipline, and replaced psychiatry among the other specialities of medicine, giving back the proper meaning to the title ‘psychiatrist’ from the Greek ‘psyche-iatros’. The proper use of the classification and its criteria, however, requires extensive training in clinical examination and the use of the diagnostic criteria as part of the psychiatric education and training curriculum for clinical psychiatrists.
It must, however, be emphasised that a correct diagnosis is not a goal in itself, but only the first step in what is the primary goal of clinical psychiatry: caring for the mentally disordered patients. This requires additional information for evaluation of factors possibly determining or contributing to the development of the actual psychiatric disorder, in accord with currently accepted biopsychosocial concepts. Only then, it will be possible to achieve a complete clinical evaluation leading to suggestions for treatment personally adapted to the individual patient. The syndromatical diagnosis only focus on the surface manifestations of mental disorder obtained from clinical interview and observation by an empathic and respectful interviewer creating a confident report and collaboration with the patient for the diagnostic evaluation. It must never be done in a mechanistic ‘computer-like’ way resulting in a dehumanising reification approach towards the patient. Terms like ‘schizophrenic’ or ‘bipolar cases’ should be avoided to prevent reificating thinking. Instead, it should always be ‘patients with’ schizophrenia or other disorders. Each patient is a unique individual, and two cases are never identical. It is therefore important to make a careful and comprehensive psychopathological description of each patient, and not just apply the diagnostic criteria by a ‘top-down’ approach and be satisfied with one or two diagnostic ‘hits’. The interest for the clinical variability will be reduced, and it will lead to impoverishment of the psychiatric culture. The WHO has coordinated the development of a clinical instrument for a ‘bottom-up’ approach in form of an extended and supplemented edition of the Present State Examination in the SCAN published by WHO, adapted to both the ICD-10 and DSM-IV diagnostic classifications [20].
The WHO, by the revision of ICD-10, followed a major principal aim of promoting and creating a worldwide common language in the classification of mental and behavioural disorders. Unfortunately, this goal has only been incompletely achieved. We have today two major competing diagnostic classifications, ICD-10 and DSM-IV, with major differences in important areas, although they eventually will use the same coding system. It is important to realise that none of the classifications is true natural systems, but fictional in the way described by Jaspers as a classification of ‘a provisional ordering value—fulfilling its purpose when it is the comparatively most correct at the time being’ (quotation translated by AB) [21]. Which of the diagnostic categories from the two systems will prevail depends on their validity, confirmed by external factors like genetics, prognostic outcomes, and treatment responses. Extensive validating studies are needed and it is hoped they will be provided before the revision of the classification systems. This may call for a moratorium of 10 or 15 years to allow time for studies of the validity and clinical utility of the categories from the two systems. It is to be hoped that eventually it will be possible to unify the two systems into one psychiatric classification making the motto of the recent Tenth World Psychiatric Association Congress in 1996 come true: one world—one language.