The World Health Organization adopted ICD-11 as the international standard in 2022. Most countries that use ICD-coding for clinical or epidemiological purposes have transitioned or are in active transition. The United States, with its longstanding ICD-10-CM (Clinical Modification) variant tightly integrated with billing infrastructure, is on a slower path — but the transition is no longer hypothetical.
For US clinicians, the ICD-11 transition is something to start preparing for now, even if formal adoption remains 2–4 years out for billing purposes.
What's actually different in ICD-11
ICD-11 is not a cosmetic update. It's a structural rebuild of disease classification that addresses several long-standing limitations of ICD-10.
The structure is fundamentally different. ICD-10 is hierarchical — a code lives in a single chapter. ICD-11 uses a more flexible structure where conditions can be classified along multiple axes simultaneously. A condition has one primary code but can carry additional descriptive elements about anatomy, severity, etiology, temporal pattern, and other dimensions.
The terminology base is much larger. ICD-11 includes roughly 17,000+ codeable entities, compared to roughly 14,400 in ICD-10. More importantly, the underlying terminology layer (the Foundation Component) is even larger and provides the granularity that the codes themselves abstract over.
Several chapters are substantially restructured. Mental and behavioral disorders, sleep-wake disorders, and conditions related to sexual health have been particularly reorganized. The mental and behavioral chapter has been brought into closer alignment with DSM-5-TR while maintaining international applicability.
Traditional medicine and digital health have dedicated sections. ICD-11 includes structured codes for traditional Chinese medicine and other indigenous medical systems. It also has codes for specific digital health phenomena that ICD-10 didn't anticipate.
The official documentation is at the WHO ICD-11 portal, and ongoing implementation guidance for the US comes through CMS at cms.gov.
Where the US is in the transition
The US adoption picture is layered.
Public health and surveillance. US public health agencies, including the CDC, are already mapping mortality and morbidity data into ICD-11 frameworks for international harmonization. This work is ongoing.
Clinical care. Most US health systems have not yet transitioned clinical coding. Documentation, problem lists, and EHR coding remain ICD-10-CM-driven.
Billing and reimbursement. This is the slowest layer. The US billing infrastructure is deeply integrated with ICD-10-CM. CMS and the National Center for Health Statistics have been building the foundation for an eventual ICD-11-CM (a US clinical modification of ICD-11), but the timeline is years out.
For now, US clinical coding remains ICD-10-CM. But the timeline for parallel use, eventual transition, and the cutover for billing purposes is firming up, and physicians can usefully start thinking about the differences before they're mandated.
What clinicians should know now
Three practical points.
Documentation that's specific in ICD-10 will be even more useful in ICD-11. The granularity advantage of ICD-11 favors detailed clinical documentation. Generic "chest pain" remains acceptable. Specific "non-cardiac chest pain attributed to costochondritis with reproducible point tenderness" maps cleanly into multiple ICD-11 axes and translates to richer downstream coding. The disciplined documentation habits that produce well-coded ICD-10 charts and MDM sections will produce even richer ICD-11 coding.
The mental health and substance use restructuring is the most clinically visible change. Physicians who care for patients with mental health conditions or substance use disorders will see meaningfully different code structures in ICD-11. The reorganization aligns more closely with current clinical thinking about these conditions.
Some codes will not have clean ICD-10 → ICD-11 mappings. Direct one-to-one mappings exist for many codes, but ICD-11's multi-axis structure means some ICD-10 codes split into multiple ICD-11 components, and vice versa. Mapping tables are being maintained by both WHO and CMS; the ones that matter for US clinicians will eventually be authoritative through CMS guidance.
What the AI documentation tools are doing
Modern ambient AI scribes that suggest ICD codes are largely ICD-10-CM-trained today, reflecting current US billing reality. The forward-looking ones are building infrastructure to support both code systems and to handle the eventual translation.
The longer-term advantage of structured AI documentation in this transition is that the underlying clinical content — the documented findings, differentials, and management decisions — is what drives coding. If the documentation is clean and specific, mapping it to ICD-11 (or any future system) is mostly a coding-engine problem rather than a re-documentation problem. Physicians whose notes are vague will face a harder transition than physicians whose notes are specific.
The international perspective
For physicians who practice in or train in multiple countries, ICD-11 is already routine. Many European, Asian, and Latin American countries have adopted it for clinical use, mortality coding, or both. International peer-reviewed literature increasingly references ICD-11 codes alongside or in place of ICD-10.
This matters for two reasons. One, US clinicians collaborating internationally will encounter ICD-11 codes in research, in cross-border consultations, and in international guidelines. Two, the eventual US transition will benefit from accumulated international experience — both in what worked smoothly and in where adoption was harder than expected.
The BMJ and other international journals have published ongoing analyses of ICD-11 adoption in countries that have completed or are well into the transition.
What to expect
Three reasonable expectations for the next several years.
Continued parallel use. US public health and international research will increasingly reference ICD-11. Clinical care and billing will remain ICD-10-CM-dominant. Knowing both at a basic level will be valuable for physicians involved in research, public health, or international collaboration.
Gradual EHR vendor preparation. Major EHRs will quietly build ICD-11 support over the next several years in anticipation of eventual US adoption. Physicians will see the codes appearing as alternatives in some contexts.
Eventual CMS announcement of an ICD-11-CM transition timeline. When that happens, the implementation timeline will likely be 18–36 months of preparation, mirroring the ICD-9 to ICD-10-CM transition that happened in 2015. The announcement is not imminent, but the groundwork is being laid.
The bottom line
For most US physicians, ICD-11 is not a 2026 problem. It is, however, a coming reality that rewards thoughtful preparation. The simplest preparation is the same thing that already produces good ICD-10-CM coding: specific, structured clinical documentation that captures the actual reasoning of each encounter.
When the eventual transition arrives, the physicians and groups that document well will find the change manageable. The ones that have leaned on vague documentation propped up by aggressive coders will face a harder adjustment. The case for documentation discipline only gets stronger as coding systems become more granular.
