ICD-11

ICD-11 is the eleventh revision of the International Classification of Diseases, maintained by the World Health Organization (WHO). It provides an international framework for recording, coding, reporting, and comparing diseases, injuries, other health conditions, and causes of death. Adopted by the World Health Assembly in May 2019, it came into effect on 1 January 2022; implementation within individual countries follows national transition arrangements rather than a simultaneous worldwide changeover. (who.int)

The classification supports public health, clinical documentation, and the production of comparable mortality and morbidity statistics. Unlike its predecessor, ICD-10, ICD-11 was designed for digital use. Its architecture connects a multidimensional Foundation Component with statistical classifications, notably ICD-11 for Mortality and Morbidity Statistics (MMS), while allowing approved combinations of codes to represent additional clinical detail. Its scope encompasses classification structure, coding rules, digital tools, changes in clinical categories, and the practical requirements of implementation and continuing maintenance. (who.int)

Purpose, scope, and institutional authority

ICD-11 is the eleventh revision of the International Statistical Classification of Diseases and Related Health Problems, maintained by the World Health Organization (WHO). It serves both as a framework for standardized health information and as a statistical classification for recording, reporting, analysing and comparing mortality and morbidity. Its terminology connects clinical descriptions with categories that can be aggregated into comparable datasets. The same framework therefore supports documentation of individual health-care encounters and population-level statistics, although these applications require different levels of detail. WHO’s February 2022 fact sheet described the system at that time as providing approximately 17,000 diagnostic categories and more than 100,000 medical diagnostic index terms. (who.int)

International comparison requires more than translating disease names. Records produced in different languages, institutions or countries must be assigned to categories with consistent meanings and processed according to common reporting rules. Otherwise, differences in terminology or grouping can appear as differences in disease frequency or causes of death. ICD supplies a shared classification through which detailed clinical expressions can be grouped for comparisons between places and across time. This supports epidemiology and public health by providing a common basis for analysing patterns of illness and death, planning services and allocating resources. Its statistical purpose depends on consistent category use, not merely on attaching a code to a clinical phrase. (who.int)

Its scope extends beyond named diseases to injuries, poisonings, symptoms, signs, abnormal clinical findings, external causes of illness or injury, and factors influencing health status or contact with health services. A record can therefore describe a symptom without an established disease, the circumstances producing an injury, or an encounter for a reason other than illness. Classification is distinct from diagnosis: clinical assessment establishes and documents the patient’s condition, while coding represents the documented information within the classification. Although ICD includes definitions and diagnostic guidance, selecting a category does not itself constitute a clinical assessment, and the presence of an ICD code does not necessarily indicate a disease. (icdcdn.who.int)

WHO’s stewardship rests on its constitutional responsibility for establishing and revising international nomenclatures of diseases and causes of death; the World Health Assembly provides the corresponding intergovernmental authority. The Seventy-second World Health Assembly adopted ICD-11 in May 2019 through resolution WHA72.15, with effect from 1 January 2022. WHO maintains the international classification and its accompanying guidance, while national authorities organize its operational introduction into their reporting and health-information systems. International adoption and entry into effect are therefore distinct from implementation within a particular country or institution. (who.int)

Within the WHO Family of International Classifications, ICD addresses diseases, related health problems and causes of death alongside two complementary reference classifications. The International Classification of Functioning, Disability and Health (ICF) describes functioning and disability, while the International Classification of Health Interventions (ICHI) classifies interventions. These instruments represent different dimensions of health information: the condition recorded, its relationship to functioning, and the intervention undertaken. Their complementary scope permits health problems, functional status and services to be described without treating any one classification as a complete account of an individual’s health or care. (who.int)

Development, adoption, and entry into effect

The World Health Organization (WHO) formally began revising ICD-10 in 2007. The work combined an update of medical content with a redesign for digital health-information systems, rather than simply transferring an existing classification into electronic form. Clinicians, statisticians, classification specialists and information-technology experts contributed through advisory groups, consultations and collaborative development. Chapter-specific groups reviewed clinical content, while cross-cutting groups addressed information modelling, mortality, morbidity, quality and safety. WHO collaborating centres and the ICD-11 Revision Steering Group supported the process; proposed content was field-tested in 31 Member States across all WHO regions. (apps.who.int)

National authorities participated before formal adoption. Following the ICD-11 Revision Conference in Tokyo on 12–14 October 2016, WHO sought Member State comments between January and October 2017 on the proposed content, structure, implementation requirements and data priorities. In April 2018, statistical experts met in Geneva to examine whether the chapter structure and categories remained suitable for mortality and morbidity statistics. These reviews addressed the need to accommodate more detailed clinical information while preserving the classification’s statistical functions. (apps.who.int)

On 18 June 2018, WHO released a version specifically intended to support implementation preparation. This was an advance release, not the date of formal international adoption or entry into effect: countries could use it to plan their transition, prepare translations and train health professionals. At the time, WHO reported receiving more than 10,000 revision proposals. Robert Jakob, then WHO’s Team Leader for Classifications, Terminologies and Standards, identified simplification of coding structures and electronic tools as a guiding principle of the revision. The release therefore provided both classification content and a practical basis for preparing national systems. (who.int)

The Seventy-second World Health Assembly formally adopted ICD-11 on 28 May 2019, through resolution WHA72.15, specifying 1 January 2022 as its effective date, subject to transitional arrangements. Adoption covered not only the classification’s categories but also associated mortality definitions, reporting requirements and rules for selecting the underlying cause of death and the main condition in morbidity coding. The resolution requested publication in WHO’s six official languages, digital support tools and assistance to Member States. It also explicitly required transitional arrangements lasting at least five years from 1 January 2022, and longer where necessary, so that countries could continue compiling and reporting statistics using earlier revisions. (cdn.who.int)

International adoption, entry into effect, publication of an updated release and operational implementation are consequently distinct events. For example, WHO published an updated ICD-11 release on 11 February 2022, incorporating refinements proposed since 2019 by early adopters, translators and scientific groups; this did not change the established effective date. Nor did 1 January 2022 require every hospital, mortality register or reimbursement system to switch simultaneously. WHO permits continued ICD-10 use for as long as necessary and recognizes differing technical, financial and organizational constraints. ICD-10 and ICD-11 may therefore coexist across countries and reporting systems during transition: international endorsement establishes the standard, whereas operational replacement depends on the readiness and implementation timetable of the systems producing the data. (who.int)

The Foundation Component and statistical linearizations

The Foundation Component is a multidimensional knowledge base of interconnected entities rather than a single list of reportable codes. Its entities represent diseases, disorders, injuries, symptoms, external causes and other health-related concepts, with associated titles, definitions, synonyms and relationships. These entities need not be mutually exclusive, and an entity can have more than one parent—a broader category under which it is organized. Tuberculous meningitis illustrates this multiple parenting: the same condition belongs to an infectious disease branch and a neurological disease branch. These are two relationships attached to one entity, not two separately defined diseases. The Foundation can therefore preserve relationships that a single hierarchical arrangement would require the classification to choose between. (icd.who.int)

A linearization is a purpose-specific classification derived from the Foundation. It selects and arranges content into mutually exclusive categories, with each entity assigned a single parent within that arrangement. ICD-11 for Mortality and Morbidity Statistics, abbreviated MMS, is the principal statistical linearization. Its hierarchical tabular list plays a role comparable to the tabular list of ICD-10, while retaining links to the underlying Foundation entities. The distinction separates the representation of medical knowledge from its organization for statistical reporting: multiple relationships can remain in the Foundation even though MMS assigns a category only one hierarchical location. WHO’s schema also provides a “Foundation child elsewhere” relationship for concepts whose Foundation relationship is not reproduced as a parent–child relationship in the linearization; the browser displays these references in grey. (icd.who.int)

Entity identifiers and classification codes serve different purposes. A Foundation entity has a persistent uniform resource identifier (URI), identifying the concept independently of its displayed title or position in a statistical hierarchy. Classification codes, by contrast, belong to categories in linearizations; the Foundation itself does not assign statistical codes. Linearizations also contain residual categories needed for reporting, which are not part of the Foundation. WHO’s application programming interface distinguishes an entity’s classification code from its “source” property, which points back to the Foundation URI. This allows software to retain the connection between a coded category and its underlying concept without treating the code and the concept identifier as interchangeable. (icd.who.int)

The terminology attached to categories performs several complementary functions. The title is the preferred label, while the definition characterizes the entity and clarifies its scope. Inclusion terms give examples of diagnostic expressions assigned to the category; they may be synonyms or distinct conditions grouped together, and neither constitute an exhaustive list nor create additional subcategories. Exclusions identify terms classified elsewhere, providing cross-references that delimit category boundaries. Index terms support retrieval through expressions drawn from titles, inclusions and narrower terms, as well as from Foundation entities not separately represented in the linearization. Consequently, a condition can be searchable without having its own statistical category: the terminology used to identify a concept is more extensive than the set of categories used to tabulate it. (icd.who.int)

Chapter organization and coverage of health conditions

ICD-11 uses a “variable-axis” chapter structure rather than arranging every condition by a single principle. Body-system chapters, such as those for circulatory, respiratory and digestive diseases, coexist with chapters organized by disease process, reproductive or developmental period, and type of health event. This structure retains the broad grouping principles associated with the nineteenth-century statistician William Farr. Special-group chapters bring together conditions that would otherwise be dispersed across anatomical systems; their placement generally takes priority over body-system placement, with exclusions indicating the relevant boundaries. Thus, the location of a condition reflects statistical grouping requirements as well as anatomy. (icdcdn.who.int)

Chapter 01, “Certain infectious or parasitic diseases,” groups infections affecting different organs: tuberculosis of the respiratory system and tuberculosis of the nervous system both appear there. Influenza, previously located in the respiratory chapter of ICD-10, was also moved into this chapter. Chapter 02, “Neoplasms,” covers more than malignant tumours: its organization distinguishes tumour behaviour, including benign, in situ, malignant, uncertain and unknown behaviour, alongside anatomical site and morphological type. Life-stage organization is represented by Chapter 18, “Pregnancy, childbirth or the puerperium,” and Chapter 19, “Certain conditions originating in the perinatal period.” Chapter 22 instead groups injuries, poisonings and certain other consequences of external causes. These arrangements preserve cross-organ groupings for infections, tumours, maternal conditions and injuries. (cdn.who.int)

Chapter 21, “Symptoms, signs or clinical findings, not elsewhere classified,” provides a separate location for observations that do not necessarily establish an underlying disease. Its sections are organized by body system and distinguish symptoms and signs from clinical findings; examples include abnormal serum enzyme levels and abnormal results of circulatory-system function studies. A record at this level can therefore describe an observed abnormality without asserting a definitive diagnosis of its cause. ICD-11 also moved several manifestations previously dispersed among ICD-10 organ chapters into this chapter, including joint pain and hemiplegia. (icdcdn.who.int)

The injury and external-cause chapters answer different questions. Chapter 22 describes the resulting harm, whereas Chapter 23, “External causes of morbidity or mortality,” describes the event or circumstances producing it, including transport events and causes of healthcare-related harm. Chapter 24, “Factors influencing health status or contact with health services,” separates reasons for contact with services from factors influencing health status. Consequently, an ICD entry may represent an injury mechanism, a clinical finding or the context of a healthcare encounter rather than a disease entity. (cdn.who.int)

Three dedicated chapters distinguish ICD-11 structurally from ICD-10: Chapter 04, “Diseases of the immune system,” Chapter 07, “Sleep-wake disorders,” and Chapter 17, “Conditions related to sexual health.” The immune-system chapter gives independent placement to immunodeficiencies, systemic immune disorders, autoinflammatory disorders and allergic or hypersensitivity conditions, rather than retaining immune disorders within ICD-10’s combined blood-and-immune chapter. The sleep–wake chapter consolidates categories previously distributed across mental, neurological and other chapters. The sexual-health chapter similarly brings together groups formerly classified in separate parts of ICD-10. These are changes in the architecture of the classification, distinct from changes to the definitions or requirements of individual diagnoses. (who.int)

Codes, postcoordination, and coding conventions

ICD-11’s Mortality and Morbidity Statistics (MMS) codes are alphanumeric, with four characters before any decimal point and up to two after it. The second character is a letter and the third a digit; “I” and “O” are excluded to avoid confusion with “1” and “0”. Stem codes represent classification categories that can stand alone. Extension codes, beginning with “X”, qualify stem codes and cannot be used independently for statistical classification or placed first in a code cluster. They supply attributes such as anatomical detail, severity, or diagnostic timing. (icdcdn.who.int)

Precoordination packages a combination of concepts into one category; postcoordination expresses additional detail through permitted combinations of codes. WHO’s browser illustrates this distinction with invasive breast carcinoma and infiltrating duct carcinoma: because their combination already has the precoordinated code 2C61.0, invasive ductal carcinoma of breast, the system returns that single code rather than two linked codes. Postcoordination instead allows details such as laterality, anatomical site, severity, or histopathology to be recorded without requiring a separate category for every combination. The available dimensions, called axes, depend on the selected category: histopathology is applicable to many neoplasms, not indiscriminately throughout the classification. (icd.who.int)

A postcoordinated group is a code cluster. A slash (/) links stem codes; an ampersand (&) attaches extension codes to the stem they qualify. Thus, the schematic expression stem1&extension1/stem2&extension2 preserves which qualifier belongs to which stem. Cluster boundaries and internal relationships must survive data transmission and analysis. (icdcdn.who.int) The browser restricts selections to applicable values: for an aneurysmal bone cyst, anatomical searches are limited to relevant bone structures. Most axes accept one value, so selecting “moderate” severity replaces “mild”; certain axes, including specific anatomy, permit multiple values. (icd.who.int)

Category-specific instructions determine whether additional coding is mandatory or optional. “Code also” requires the indicated additional information when known; “Use additional code, if desired” permits supplementary detail. For example, diabetic cataract requires coding the diabetes type: the reference guide gives the morbidity cluster 9B10.21/5A10 for diabetic cataract with type 1 diabetes mellitus. Unrelated hypertension is recorded separately, not inserted into that cluster. (icdcdn.who.int) The browser displays only applicable postcoordination axes and flags required additions; code combinations therefore follow defined relationships rather than unrestricted concatenation. (icd.who.int)

Mortality and morbidity statistics

Mortality coding begins with the medical certificate of cause of death, rather than a list of diagnoses taken without context from the medical record. On the international death certificate, Part I records the causal sequence: the condition immediately preceding death appears first, with antecedent conditions below it. Part II records other significant conditions that contributed to death but were not part of that sequence. The underlying cause is the disease or injury that initiated the fatal sequence, or the circumstances of the accident or violence producing the fatal injury. It is therefore distinct from the immediate cause and from unrelated contributing conditions. The condition properly entered on the lowest completed line of Part I and the underlying cause ultimately selected for statistical tabulation are related concepts, but are not interchangeable: selection requires application of ICD rules. (ncbi.nlm.nih.gov)

For routine cause-of-death tables, one underlying cause is selected for each death. ICD-11 provides an ordered procedure for identifying the starting point of the reported sequence and then checking whether special instructions require modification. These instructions address, among other matters, causal relationships, specificity, injuries and medical procedures. Multiple-cause coding preserves information about other reported conditions, permitting analysis beyond the single underlying cause. Thus, statistics based on deaths due to a disease answer a different question from statistics counting every certificate on which that disease appears. (icdcdn.who.int)

Morbidity coding describes conditions associated with an episode of care. For hospital-based care, the main condition is the reason for admission established at the end of the episode, after investigation—not necessarily the presenting symptom. Additional conditions are those present at admission or developing during care that affect patient management; earlier conditions with no bearing on the episode are excluded. The responsible practitioner identifies and documents these conditions, and the coder applies the classification to that documentation. Inconsistent records should be returned for clarification where possible; specified reselection rules apply when clarification is unavailable. WHO’s reference guide illustrates this with admission for staphylococcal meningitis followed by pneumonia during hospitalization: meningitis remains the main condition, while pneumonia is an additional condition. (icdcdn.who.int)

The resulting data support cause-of-death analysis, disease surveillance, cancer registration and patient-safety monitoring. WHO’s release announcement of 11 February 2022 identified these uses alongside antimicrobial-resistance monitoring and reimbursement, and highlighted support for recording cancer grade and stage. Codes for antimicrobial resistance were aligned with the Global Antimicrobial Resistance and Use Surveillance System, while patient-safety coding was designed to document harm associated with health care. These applications use different selections and combinations of the classified information rather than treating every coded record as equivalent. (who.int)

For reimbursement, ICD-11 supplies diagnostic information to administrative systems; it is not itself a payment schedule. Case-mix systems, including diagnosis-related groups, use grouping algorithms that combine diagnosis codes with procedure codes and other variables to allocate cases to groups. Payment and resource-allocation rules belong to those systems and their national implementations, rather than to the disease classification alone. (icdcdn.who.int)

Selected clinical changes and diagnostic guidance

ICD-11 introduced new categories within mental, behavioural and neurodevelopmental disorders, including complex post-traumatic stress disorder, prolonged grief disorder and gaming disorder. Changes also concern how conditions are described: dimensional approaches, notably for personality disorders, characterize variation along continua rather than relying exclusively on separate categorical types. The accompanying clinical guidance adopts a lifespan approach, addressing differences in presentation during childhood, adolescence and older adulthood, and includes culture-related guidance on how symptoms and their interpretation may vary between settings. These features distinguish the requirements for clinical diagnosis from the shorter descriptions needed for statistical classification. (who.int)

Gaming disorder encompasses online and offline gaming and is defined by impaired control over gaming, increasing priority given to gaming over other interests and daily activities, and continuation or escalation despite negative consequences. Diagnosis requires significant impairment in personal, family, social, educational, occupational or other important functioning; the pattern would normally have been evident for at least 12 months. Ordinary participation in video games, enthusiasm for gaming or time spent playing does not by itself establish this disorder: the defining requirements concern the behavioural pattern and its functional consequences. WHO’s September 2018 explanation emphasized that available studies suggested the disorder affected only a small proportion of people who played digital or video games. (who.int)

Gender incongruence was moved from the mental and behavioural disorders chapter of ICD-10 to ICD-11’s chapter on conditions related to sexual health. The revised categories distinguish gender incongruence of adolescence or adulthood from gender incongruence of childhood, replacing earlier categories including “transsexualism” and “gender identity disorder of children”. WHO explains the relocation as reflecting that transgender and gender-diverse identities are not conditions of mental ill-health. Retaining categories elsewhere in ICD is intended to support access to relevant health services and financial coverage, rather than to classify transgender identity itself as a mental disorder. Gender-variant behaviour and preferences alone are not sufficient grounds for assigning these diagnoses. (who.int)

On 8 March 2024, the World Health Organization published the 852-page Clinical Descriptions and Diagnostic Requirements for ICD-11 Mental, Behavioural and Neurodevelopmental Disorders (CDDR). Its intended users include mental health specialists and qualified non-specialist practitioners, such as primary care physicians who assign diagnoses, as well as professionals and trainees who need to understand these conditions. The CDDR supplies clinical diagnostic guidance complementary to the ICD-11 Mortality and Morbidity Statistics linearization (MMS). The two are therefore related resources with different purposes: the CDDR supports identification and differentiation of disorders in clinical settings, while MMS provides the classification structure for standardized statistical recording and reporting. (who.int)

Development and testing involved hundreds of experts and thousands of clinicians internationally. An example of the underlying field-testing programme was a 2018 study led by Geoffrey M. Reed and colleagues, in which 339 clinicians assessed the clinical utility of proposed ICD-11 diagnostic guidelines using 1,806 patients across 28 mental health settings in 13 countries. The study examined their usability in actual clinical encounters, including whether descriptions fitted patients’ presentations and helped practitioners apply diagnostic categories. Such field studies formed part of the development of the clinical guidance, rather than constituting a separate revision of the statistical classification. (who.int)

Supplementary classification of traditional medicine conditions

The supplementary chapter on traditional medicine conditions provides an optional classification for recording traditional medicine diagnoses alongside conventional ICD categories. Its designated role is dual coding of morbidity data: the conventional classification records the disease, symptom, or other reason for care, while a supplementary code records the traditional medicine diagnostic description. WHO specifies that coding should also include a category from chapters 1–24. The supplementary chapter is not intended for mortality reporting, and its codes must not be used to report causes of death. Its applications include morbidity reporting, reimbursement, patient-safety monitoring, and research. (who.int)

Module I covers diagnostic concepts originating in ancient China and subsequently used in China, Japan, the Republic of Korea, and other countries. Its scope therefore overlaps with traditional Chinese medicine but is not restricted to practice within China. The module distinguishes traditional medicine disorders from patterns, rather than treating all its categories as direct equivalents of conventional disease diagnoses. The ICD-11 Reference Guide illustrates dual coding with asthma: a conventional asthma diagnosis may be recorded alongside “wheezing disorder (TM1),” coded SF81, or “turbid phlegm accumulation in the lung pattern (TM1),” coded SF86. These examples preserve both diagnostic descriptions within the record instead of replacing the conventional diagnosis with a traditional one. (who.int)

The coding instructions allow use in hospital inpatient care, hospital outpatient care, and community settings. They direct coders to consult the patient record, identify the diagnoses relevant to the episode of care, and check the classification’s descriptions, inclusions, and exclusions before assigning codes. More than one traditional medicine disorder or pattern may be recorded when documented. In primary care, where a disorder or pattern may not yet be fully developed, the reason for the encounter may be more appropriate to record than a definitive main condition. The reference guide also identifies precedents in China, Japan, and Korea for using traditional medicine classifications in hospital reimbursement, insurance claims, and clinical costing. (icdcdn.who.int)

Module I is distinct from the Ayurveda-related expansion announced by the World Health Organization on 14 February 2025. WHO’s announcement of the 2025 ICD-11 update included a new module covering Ayurveda and related traditional medicine systems, including Siddha and Unani. This expanded the range of traditional diagnostic systems represented; it was not simply a renaming of the China-origin concepts in Module I. WHO described its purpose as enabling systematic tracking of traditional medicine services for research, reporting, and evidence-based policymaking. (who.int)

WHO explicitly distinguishes classification from endorsement: inclusion neither validates the scientific basis of a traditional medicine practice nor establishes the efficacy of an intervention. The categories describe conditions and encounters, not approved treatments. Standardized recording is intended to make service use, outcomes, costs, and adverse events more consistently measurable across institutions and countries. Safety and efficacy remain subjects for research and evaluation; possession of an ICD code is not evidence that a particular treatment is safe or effective. (who.int)

Digital tools, multilingual access, and interoperability

The ICD-11 Browser and Coding Tool serve complementary functions. The Browser displays the classification hierarchy alongside category descriptions, inclusions, exclusions, coding notes, and permitted postcoordination. The Coding Tool starts from a clinical expression entered by the user and supports finding and selecting the corresponding category. Its integrated browser allows a search result to be examined in its hierarchical context without leaving the tool. WHO’s reference guide specifies a three-step electronic workflow: enter the reported term, select the matching expression, and verify the result against the tabular classification, including instructions inherited from grouping and chapter levels. A search match therefore does not replace examination of the applicable coding rules. (icd.who.int)

The ICD application programming interface (API) provides programmatic access through an HTTP-based REST service. WHO’s Embedded Classification Tools add ready-made Browser and Coding Tool components to web applications through a JavaScript library powered by this API. These components can place classification search and selection within an electronic health record or another clinical application rather than requiring a separate website. Configurable callbacks allow the host software to process a user’s selection; settings specify the classification source, release, and language. WHO also distributes local API deployment options, including Docker containers, Windows services, and Linux services, allowing classification services to operate within local infrastructure. (icd.who.int)

Release and language are distinct integration parameters. For example, the embedded tool documentation uses 2025-01 as a classification release identifier and en, es, and fr for English, Spanish, and French. Leaving the release setting unspecified makes the component use the latest available release, whereas explicitly selecting a release fixes the classification content used by that integration. Recording both release and language alongside exported coding metadata supports reproducibility: the release identifies the classification consulted, while the language identifies the terminology displayed during selection. The software-library version is separate from these content settings; upgrading an embedded interface and changing the classification release are not the same operation. (icd.who.int)

Multilingual access involves terminology management rather than simply translating interface labels. On 14 February 2025, WHO reported that ICD-11 was available in 14 languages. Its translation platform provides an editing page for each entity, accommodates additional language-specific synonyms, and permits a synonym to be selected as the translated title. Authorized reviewers approve translations through defined stages. A translation memory suggests previously translated phrases, but translators can replace inappropriate suggestions. The platform also distinguishes an unfinished translation from a term deliberately recorded as having no translation in that language. These mechanisms support consistent conceptual coverage while retaining locally used clinical expressions. (who.int)

WHO’s February 2025 update identified interoperability with Orphanet, MedDRA, and other terminologies, together with FHIR API integration for health-information exchange. Such connections do not establish complete equivalence between the linked systems. Orphanet’s mapping documentation explicitly distinguishes exact matches from narrower-to-broader and broader-to-narrower relationships. It further records whether a rare-disease concept has its own ICD category, appears as an inclusion or index term, or has a code attributed by Orphanet. Consequently, an ORPHAcode linked to an ICD-11 category need not represent a separately coded ICD-11 disease. These relationship qualifiers make the meaning and granularity of a mapping explicit rather than treating every connection as interchangeable terminology. (who.int)

National implementation and continuity of health data

National transition is an organizational programme rather than a single change of code lists. ICD-11’s entry into effect on 1 January 2022 did not require every national reporting system to switch on that date: the World Health Organization (WHO) permits continued use of ICD-10 while implementation proceeds. National planning therefore distinguishes the classification’s international status from its operational use in particular hospitals, mortality registers, or administrative systems. WHO describes implementation as a collaborative process involving clinical, administrative, research, and policy stakeholders, with technology and the level of coding detail matched to sustainable national resources. (who.int)

WHO’s implementation guide proposes national coordination through a task force or centre of excellence. Initial assessment identifies existing classifications, reporting obligations, decision-makers, workforce requirements, technological capacity, and costs. Preparatory work includes translation, training clinicians and coders, adapting information systems, and testing workflows before wider introduction. Software changes extend beyond displaying new codes to revising catalogues, data fields, and reporting arrangements. Pilot testing identifies local requirements and disruptions before expansion. WHO’s 2022 fact sheet estimated that introducing ICD-11 into a simple information system might take one to two years, whereas replacing an earlier revision in a sophisticated system might require four to five years; these were planning estimates, not uniform deadlines. WHO technical assistance includes translation-platform support and integration of coding tools into local software. (icd.who.int)

Mapping tables, or crosswalks, identify correspondences between ICD-10 and ICD-11, but do not make the revisions interchangeable. When categories split, merge, or change their boundaries, an old code may not identify a unique new category. In particular, a historical record cannot supply clinical detail that was never collected. WHO explicitly states that its mapping tables support comparison rather than automatic conversion of ICD-10 records into ICD-11 codes. A mapped historical dataset is therefore distinct from one coded directly from clinical documentation using ICD-11. (who.int)

Bridge coding, also called dual coding, addresses statistical continuity by coding the same deaths or episodes of care under both revisions. Comparing the results identifies redistribution between categories and supports comparability factors for longitudinal analysis. This helps distinguish classification-related discontinuities from changes in population health. WHO’s transition guide encourages retaining the existing system during an overlap of at least 18 months, allowing transition studies, detection of implementation problems, and continuity of reporting while the new system becomes operational. (icd.who.int)

Tunisia provides a documented example of hospital-level implementation rather than nationwide replacement. A transition and capacity-building plan developed in 2023 covered patient medical records in two public hospitals. During a WHO mission in November 2023, participants reviewed the existing ICD-9 system and developed a national transition plan, followed by a workshop establishing the pilot timetable and actions. Virtual meetings in 2023–2024 connected Tunisian Ministry of Health information-technology staff with Kuwait’s National Center for Health Information to support integration of the ICD-11 application programming interface. In December 2024, the Canadian Institute for Health Information provided three days of French-language training for more than 40 participants. WHO’s regional account reports implementation in the two hospitals within nine to twelve months; that reported institutional outcome does not establish nationwide replacement of Tunisia’s coding systems. (emro.who.int)

Maintenance, versioning, and classification quality

The World Health Organization maintains ICD-11 through an online proposal platform on which interested parties can suggest additions, corrections, and changes to terminology or classification structure. Submissions require supporting references and scientific evidence. Scientific and classification experts review proposals through the WHO Family of International Classifications (WHO-FIC) network, involving mortality and morbidity reference groups, the Medical and Scientific Advisory Committee, and the Classification and Statistics Advisory Committee. The platform supports discussion, feedback, and recording of outcomes; submission alone does not make a proposed change part of an official release. (who.int)

Maintenance distinguishes continuously evolving development content from published statistical classifications. WHO’s development site explicitly identifies its contents as unreleased work in progress. Official releases for mortality and morbidity are produced annually, while changes with substantial statistical consequences follow longer update cycles. The release archive identifies successive editions—for example, 2024-01, 2025-01, and 2026-01—and supplies separate change documentation for each. Consequently, an accepted proposal, its appearance in development content, and its inclusion in a published release are distinct stages. (icd.who.int)

Release identification supports reproducibility when categories, labels, or instructions change. The ICD application programming interface exposes releaseId and releaseDate properties, with release identifiers generally incorporating a year and month. Its documentation distinguishes classification releases from versions of the API itself: requesting API version v2 does not specify which classification release is being retrieved. Historical Foundation content can be requested using a releaseId parameter; without that parameter, Foundation identifiers resolve to the latest release. For reproducible analysis, the relevant classification, release, language, and coding rules therefore need to remain identifiable alongside the coded dataset, rather than being represented merely by a reference to “ICD-11.” (icd.who.int)

Classification quality also depends on faithful application of coding conventions. Postcoordination must reflect documented relationships, not associations inferred by the coder. The reference guide illustrates this with cataract and type 2 diabetes: when a causal relationship has not been documented, the conditions are reported separately rather than combined as a causally related cluster, although laterality can still be added to the cataract code. Such distinctions affect whether institutions produce comparable representations of apparently similar clinical records. (icdcdn.who.int)

WHO describes data-quality assurance as a workflow extending from clinical documentation through reporting, verification, grouping, and analysis. Conditions recorded in a medical record or death certificate must be complete and accurate; unclear, incomplete, or illogical statements require queries back to the documenter. Category selection should retain the greatest specificity supported by the record, rather than introduce unsupported detail. Standardized codes permit language-independent reporting, but comparisons across languages, institutions, and countries remain conditional on consistent coding rules. Across releases and major revisions, documented classification changes must likewise be distinguished from changes in the underlying frequency of disease. (icdcdn.who.int)

References

1. ICD-11who.int
2. International Classification of Diseases (ICD)who.int
3. ICD-11 Reference Guideicdcdn.who.int
4. Classifications and terminologieswho.int
5. World Health Organization Family of International Classificationscdn.who.int
6. Eleventh revision of the International Classification of Diseases: Report by the Director-Generalapps.who.int
7. WHO releases new International Classification of Diseases (ICD 11)who.int
8. Eleventh revision of the International Classification of Diseases: Resolution WHA72.15cdn.who.int
9. WHO’s new International Classification of Diseases (ICD-11) comes into effectwho.int
10. ICD-11 Implementationwho.int
11. User Guideicd.who.int
12. ICD Schemaicd.who.int
13. API Documentation (version 2.x) - ICD-API Homepageicd.who.int
14. Content Model Reference Guide for ICD, ICF and ICHIicd.who.int
15. International Classification of Diseases, 11th Revision — Dr Robert Jakobcdn.who.int
16. Clinical descriptions and diagnostic requirements for ICD-11 mental, behavioural and neurodevelopmental disordersiris.who.int
17. Obtaining cause-of-death information for HIV/AIDS — Guidelines for HIV Mortality Measurementncbi.nlm.nih.gov
18. ICD-11 Reference Guideicdcdn.who.int
19. ICD-11 2022 releasewho.int
20. New manual released to support diagnosis of mental, behavioural and neurodevelopmental disorders added in ICD-11who.int
21. Addictive behaviours: Gaming disorderwho.int
22. Inclusion of “gaming disorder” in ICD-11who.int
23. Gender incongruence and transgender health in the ICDwho.int
24. What does the ICD-11 say about transgender health?cdn.who.int
25. Clinical descriptions and diagnostic requirements for ICD-11 mental, behavioural and neurodevelopmental disorders (CDDR)who.int
26. Traditional medicinewho.int
27. International Classification of Diseases (ICD)who.int
28. WHO releases 2025 update to the International Classification of Diseases (ICD-11)who.int
29. World Health Organization: traditional medicine reviewiris.who.int
30. Integrated browsericd.who.int
31. Home Page ICD APIicd.who.int
32. Embedded Classification Tools (ECT)icd.who.int
33. Embedded Coding Tool (version 1.8)icd.who.int
34. User Guide: Translation Page for Individual Entitiesicd.who.int
35. WHO releases 2025 update to the International Classification of Diseases (ICD-11)who.int
36. Mappings Browser User Guideorpha.net
37. ICD-11 Implementation or Transition Guideicd.who.int
38. WHO EMRO - Implementing ICD11 in the Eastern Mediterranean Regionemro.who.int
39. ICD-11 Reference Guide, 2025 releaseicdcdn.who.int
40. WHO-FIC Maintenance Platformicd.who.int
41. WHO-FIC Classificationsicd.who.int
42. ICD Schemaicd.who.int
43. Release Notes for ICD API version 2.0icd.who.int