ICD-10 versus ICD-11 – What Has Changed in the Classification

Why the WHO replaced ICD-10 with a new classification

The International Statistical Classification of Diseases and Related Health Problems (abbreviated as ICD, from its English name) is a catalogue of diagnoses developed by the World Health Organization, which organises the whole of medicine – from infectious diseases to mental disorders. The tenth version (ICD-10) was created at the turn of the 1980s and 1990s, and in psychiatry and clinical psychology it was used for decades. Over that time, knowledge of mental health advanced considerably, and the classification itself began to lag behind the state of research.

ICD-11 is the answer to that gap. It is not a cosmetic update, but a rebuilding of the logic by which disorders are described. There were several reasons. First, ICD-10 was created still in the era of paper documentation, whereas ICD-11 was designed as a fully digital system, integrated with electronic medical records and adapted for automated data processing. Second, the WHO wanted to base the description of mental disorders on newer scientific evidence and on the opinions of clinicians from various countries, and not only from a wealthy cultural sphere.

Third, the developers were keen to achieve greater usefulness in real practice. Research showed that some of the categories in ICD-10 were rarely used or difficult to diagnose accurately, and that the boundaries between certain entities were sometimes artificial. ICD-11 was meant to reduce these problems while preserving continuity – most of the familiar diagnoses remained, while their organisation, codes and criteria changed.

It is worth emphasising that ICD diagnostics serves two roles at once: it is a clinical tool and at the same time the basis of health statistics and billing. The new version therefore had to reconcile precision of description with simplicity of coding on a large scale.

Timeline and implementation status

Work on ICD-11 lasted many years and involved broad consultation, reviews by working groups, and field studies (tests of the criteria conducted on real clinical cases in many countries). The World Health Assembly formally adopted ICD-11 in 2019, and the classification officially came into force on 1 January 2022 as the binding standard for international reporting.

Coming into force at the WHO level does not, however, mean that the classification automatically becomes binding in every country. The implementation of ICD-11 is a process spread over years: individual states adapt the translations, adjust their information systems, train staff and amend the regulations governing documentation and billing. For this reason, in many countries both versions – ICD-10 and ICD-11 – coexist for a certain period.

For someone checking ICD-10 versus ICD-11, this has a practical consequence: the pace and the exact moment of the transition depend on the local regulations of a given country and health-care sector. Regardless of the implementation calendar, it is worth knowing the logic of ICD-11 already now, because it sets the direction of thinking about diagnosis. The specific dates on which it becomes binding in a given health-care system should be verified in current official sources, because the timetables differ and are sometimes updated.

The most important structural changes in the mental health chapter

The chapter devoted to mental, behavioural and neurodevelopmental disorders was reorganised in ICD-11. In ICD-10, mental disorders were contained in a single chapter designated by the letter F (hence the popular codes such as F32 or F20). In ICD-11 a new, alphanumeric coding system applies, in which codes begin with a digit and a letter, for example 6A20 for schizophrenia or 6A70 for a depressive episode.

The order and grouping of the categories also changed. In ICD-11, related disorders were gathered into more coherent families, guided by current knowledge of their shared features. There arose, among others, separate, clearly delineated groups for anxiety and fear-related disorders, for obsessive-compulsive and related disorders, and for disorders specifically associated with stress. In ICD-10, some of these entities were scattered or combined in a less intuitive way.

A comprehensive companion publication with clinical descriptions and diagnostic guidance also appeared, referred to in the professional literature by the abbreviation CDDR (from the English name denoting clinical descriptions and diagnostic requirements). It replaced the former ICD-10 guidance and provides the clinician with an elaborated, descriptive portrait of each disorder instead of a dry list of points.

An important philosophical change is the move away from rigid thresholds towards a prototypical description. In practice this means that ICD-11 more often describes the typical picture of a disorder and leaves it to the clinician to judge whether a given patient fits that picture, rather than requiring a specific number of symptoms to be ticked off.

From subtypes to the dimensional approach

One of the deepest differences between ICD-10 and ICD-11 is a change in the way of thinking about the severity and variants of disorders. ICD-10 readily made use of subtypes – closed variants of a single illness, to one of which the patient was to be assigned. An example was schizophrenia with its distinct forms, such as the paranoid, hebephrenic or catatonic.

The dimensional approach (describing a disorder by means of gradable properties, rather than box-like categories) assumes something different. Instead of asking which compartment the patient belongs to, the clinician describes how strongly the individual dimensions of the problem are expressed – for example the severity of symptoms, their type or the degree of impairment of functioning. Such a description better reflects reality, because mental disorders rarely have sharp boundaries.

ICD-11 introduces dimensionality in several areas. In psychoses, instead of subtypes of schizophrenia, symptom specifiers are used (additional designations describing which groups of symptoms predominate, for example positive, negative, depressive or cognitive symptoms). In personality disorders, an assessment of severity was introduced. In mood disorders, the description of the features of an episode was made more precise.

The practical benefit is that two patients with the same diagnosis can be described more precisely, capturing the differences between them. A drawback can be the greater complexity of documentation, which one has to get used to. It is a compromise that the WHO deliberately chose, judging fidelity to the clinical picture to be more important than simplicity.

New diagnoses: complex post-traumatic stress disorder and prolonged grief

ICD-11 introduced diagnoses that ICD-10 did not contain, because since the latter was created science has described these conditions better. Two clear examples concern reactions to stress and trauma.

Complex post-traumatic stress disorder (code 6B41) is a diagnosis distinct from classic post-traumatic stress disorder (6B40). Classic post-traumatic stress disorder usually develops after a single or short-lived traumatic event. The complex form, by contrast, encompasses its core symptoms and, in addition, further difficulties: serious problems with emotion regulation, a persistently negative self-concept and persistent difficulties in relationships with others. Such a picture appears more often after prolonged, repeated trauma from which it is difficult to escape. Distinguishing this category makes it possible to describe more accurately patients who previously were hard to assign an adequate diagnosis to.

Prolonged, that is complicated, grief disorder (code 6B42) is the second new diagnosis in this family. It describes a situation in which the reaction after the loss of a loved one is exceptionally intense, persists far longer than is typical for a given cultural context and seriously disrupts everyday functioning. It must be emphasised that grief in itself is not a disorder – it is a natural reaction. The diagnosis concerns only the few, intense and chronic courses that require help.

Within the same family of stress-related disorders there is also adjustment disorder (6B43), which in ICD-11 was made more precise as a reaction to an identifiable stressor, with a characteristic excessive preoccupation with it.

New behaviour-related diagnoses: gaming disorder

Another new diagnosis in ICD-11 is gaming disorder (code 6C51) – a condition in which gaming, including digital and online games, slips out of control. Three features are key here: loss of control over gaming, giving it an increasing priority at the expense of other important areas of life, and continuing or escalating gaming despite clearly negative consequences.

It is very important, however, to understand this category correctly. Frequent or intensive gaming in itself is not a disorder. The diagnosis requires a persistent pattern and significant impairment of functioning – in personal, family or social life, in study or at work. Usually this pattern must persist for a longer time, and in the presence of clear and severe symptoms the required period may be shorter. The purpose of distinguishing this diagnosis is not to pathologise a hobby, but to make help available to people who are genuinely suffering.

Gaming disorder was placed in the same ICD-11 group as gambling disorder (6C50). Both describe a pattern of behaviour that becomes compulsive and harmful, even though it does not involve the intake of any substance. This reflects the broader view of addictions adopted in ICD-11, in which behavioural addictions are treated seriously, alongside substance addictions.

The rebuilding of personality disorders: the severity model

Personality disorders are an area in which the differences between ICD-10 and ICD-11 are especially clear. ICD-10 listed a set of distinct types of personality disorder – paranoid, schizoid, dissocial, emotionally unstable, histrionic, anankastic and others – treated as separate categories.

ICD-11 almost entirely abandoned this list. In its place it introduced a single diagnosis – personality disorder (code 6D10) – assessed according to severity. The clinician first decides whether a personality disorder is present at all, and then determines its degree: mild, moderate or severe. A category of personality difficulty was also provided for pictures that do not reach the threshold of a full diagnosis.

Once the severity has been established, the description can be supplemented with personality traits (code 6D11) – gradable dimensions giving a more detailed portrait. Among them are, for instance, negative affectivity, detachment, disinhibition, dissociality and anankastia. An optional borderline pattern specifier was also retained, because the concept is widely known and clinically useful.

The point of this change is that the former personality types in practice strongly overlapped, and patients rarely fitted one of them purely. The severity model directs attention to what matters most for prognosis and for planning help – how deeply the disorder affects functioning and relationships – rather than to the attempt to assign the patient to a single label.

Changes in mood disorders and schizophrenia

Mood disorders in ICD-11 retained the familiar entities, but they were arranged into a clearer family. Bipolar disorder (also called bipolar affective disorder) was distinguished as type I (code 6A60) and type II (code 6A61). In type I, full episodes of mania occur, that is, a state of pathologically elevated mood and drive. In type II, episodes of hypomania (a milder form of mania) occur together with depressive episodes, without full mania. ICD-10 did not distinguish these types so clearly, so this clarification makes it easier to describe the course accurately.

Depressive episodes and recurrences were described by means of diagnoses such as single episode depressive disorder (6A70) and recurrent depressive disorder (6A71). To describe a specific episode there are additional specifiers concerning its severity and features, for example the presence of anxiety symptoms or melancholic symptoms. Dysthymic disorder (6A72) was also retained, describing chronically low mood of lesser severity.

In schizophrenia (code 6A20) the most conspicuous change is the abandonment of the classic subtypes present in ICD-10. The paranoid, hebephrenic, catatonic and residual forms ceased to be separate categories. The reason was weak evidence that these subtypes are stable over time and clinically useful, and that they accurately predict the course or the response to treatment.

In their place, ICD-11 introduces description by means of symptom specifiers, which make it possible to mark which groups of symptoms currently predominate. Catatonia (code 6A40) – a syndrome of motor and behavioural disturbances – was distinguished as a separate category, which can accompany various disorders, and not only schizophrenia. This is a more flexible view, truer to practice.

Other significant shifts of categories

Beyond the most talked-about changes, ICD-11 carried out many quieter shifts that are also worth knowing when checking ICD-10 versus ICD-11. Obsessive-compulsive disorder (code 6B20) ceased to be treated as a variant of the anxiety disorders and became the centre of a separate family of obsessive-compulsive and related disorders.

Entities previously scattered or not distinguished joined this new family. They include hypochondriasis, called in ICD-11 health anxiety (code 6B23), and hoarding disorder, that is pathological hoarding (code 6B24) – a persistent difficulty with parting from possessions, leading to the cluttering of living space. Previously, hoarding had no distinct, unambiguous diagnosis of its own.

The treatment of somatic symptoms was also changed. ICD-11 introduced bodily distress disorder (code 6C20), which focuses on the real, burdensome experiencing of bodily symptoms and on an excessive focus of attention on them, regardless of whether they have a medical explanation. This is a departure from the former logic, which placed the emphasis on the absence of a medical explanation for the symptoms.

The dissociative disorders were also reorganised, including dissociative amnesia (code 6B61) and dissociative neurological symptom disorder (6B60). In addition, ICD-11 consistently introduces categories of secondary disorders – such as secondary psychotic syndrome (code 6E61) – which describe mental pictures that are the direct consequence of another physical illness. This makes it possible to separate clearly a primary disorder from a state arising from a bodily illness.

What the change means for the clinician in practice

For the clinician, the transition from ICD-10 to ICD-11 is not only new codes, but a somewhat different way of thinking about diagnosis. Instead of striving for a single closed label, ICD-11 encourages a multi-layered description: the main diagnosis, its severity and specifiers refining the picture. This requires more reflection when documenting, but it yields a description truer to the patient's actual state.

A practical advantage is a better reflection of ambiguous situations. A patient who formerly fitted one category only with difficulty can, in ICD-11, be described more accurately – for example, by indicating the severity of a personality disorder and the accompanying traits, instead of choosing one of the overlapping types.

The transitional period remains a challenge. Where both versions coexist, the clinician must be conversant with both and able to translate a diagnosis from one to the other. It helps here to be aware that most entities have their counterparts, although these are not always an exact one-to-one rendering.

It must also be remembered that every classification has its limits. Neither ICD-10 nor ICD-11 replaces a full clinical assessment, a history and the context of the patient's life. The classification is a tool for organising and communicating, and not a diagnostic automaton. Educational materials, including this guide, serve to broaden knowledge and are not a tool for making a diagnosis on one's own.

The consequences of migration and coding

The transition to ICD-11 has tangible consequences for information systems and administrative processes. The most obvious is the change of the codes themselves: the former designations with the letter F give way to a new alphanumeric system. Documentation software, statistical reporting and diagnostic dictionaries all have to be updated.

Mapping between versions, that is, assigning an ICD-10 code to its corresponding ICD-11 code, is often feasible, but requires caution. Where the structure of categories has changed, one old diagnosis may correspond to several new ones, or vice versa. This concerns especially personality disorders, where the former types were replaced by the severity model, and schizophrenia, where the subtypes disappeared. In such places an automatic rewriting of the code is not enough, and a clinical decision is needed.

To preserve the continuity of data, it is important that historical diagnoses remain legible also after the migration. The data gathered under ICD-10 do not disappear – one still has to be able to interpret them, comparing them carefully with the new logic of ICD-11. For this reason, during the transitional period it is good practice to document according to which version a given diagnosis was made.

From the perspective of tools supporting ICD diagnostics, a benefit of ICD-11 is its digital design. The classification was conceived as a system that can be searched, is internally linked and is easier to integrate with electronic documentation, which in the longer term simplifies work, once the adjustment period is over.

A summary of the most important differences

Setting ICD-10 against ICD-11 in brief: the coding system changed, the families of disorders were reorganised, and prototypical description in many places replaced rigid numerical thresholds. The dimensional approach entered the psychoses, personality disorders and mood disorders, placing the emphasis on severity and specifiers instead of closed subtypes.

New diagnoses appeared, responding to real clinical needs: complex post-traumatic stress disorder, prolonged grief disorder and gaming disorder. Personality disorders were rebuilt into a severity model, bipolar disorder was split into type I and type II, and in schizophrenia the classic subtypes were abandoned in favour of a symptom-based description.

These ICD-10 and ICD-11 differences are united by a common direction: closer to the patient's actual picture, further from artificial boundaries. To see what the individual changes look like in practice, it is worth moving on to the pages describing specific ICD-11 disorders linked to this guide.

Frequently asked questions

Related ICD-11 disorders