What Is ICD-10-CM?

What ICD-10-CM is

ICD-10-CM stands for International Classification of Diseases, 10th Revision, Clinical Modification. The base classification, ICD-10, is published by the World Health Organization and used internationally for mortality and morbidity statistics. The “CM” is the United States clinical modification of it: an expanded version built for the level of clinical detail American healthcare reporting requires.

ICD-10-CM answers one question: what was the patient's condition. It is the diagnosis vocabulary behind claims, quality measures, registries and public-health statistics. It does not describe what was done to the patient — that is the job of a procedure code set such as ICD-10-PCS or HCPCS Level II.

ICD-10-CM replaced ICD-9-CM for United States diagnosis reporting on October 1, 2015. It is the HIPAA-adopted diagnosis code set, which is why the same codes appear across payers, providers and clearinghouses rather than each organization keeping its own list.

Who maintains it

Two federal organizations share responsibility. The National Center for Health Statistics (NCHS), part of the CDC, maintains the diagnosis classification itself and publishes the official ICD-10-CM files. CMS maintains ICD-10-PCS and publishes the annual ICD-10 update materials that healthcare organizations work from. The two bodies run the ICD-10 Coordination and Maintenance Committee jointly: proposals for new, revised or deleted codes are presented at public meetings before they can reach a release.

Alongside the code set, the same organizations publish the ICD-10-CM Official Guidelines for Coding and Reporting, which set out how the codes are to be applied and sequenced. The guidelines are the authority on usage; a reference site like this one shows you what a code is, not when to report it.

What a code looks like

An ICD-10-CM code is 3 to 7 characters. The first character is always a letter; a decimal point is printed after the third character when the code is longer than three. The first three characters are the category, and everything after the decimal adds specificity — typically cause, anatomic site, severity, or the episode of care.

E 1 1 . 9

Category — type 2 diabetes mellitus Detail — without complications

A three-character category is a real entry in the code set, but it is usually not something you can report on its own — it exists to hold the more specific codes beneath it and to carry the notes that apply to all of them. The distinction is published in the official order file as a validity flag, which is where the billable/specific label on each code page comes from. For a character-by-character walkthrough, including the placeholder X, see How to Read an ICD-10-CM Code.

How the code set is organized

The Tabular List — the authoritative part of ICD-10-CM — is arranged as a strict hierarchy: chapter → block → category → subcategory → code. Chapters group conditions mostly by body system or by the nature of the problem, and each chapter owns a letter range. The FY 2027 Tabular List has 22 chapters.

A second file, the Alphabetic Index, lists clinical terms and points to candidate codes. The Index is a lookup aid; the Tabular List is what actually defines a code. MedCodeSearch imports the Index to improve search, but every code page shows the Tabular List entry.

The FY 2027 release loaded here contains 98,403 entries. Of those, 74,879 are flagged in the official order file as valid for submission and 23,524 are non-billable header codes that organize the rest. You can walk the whole structure from the ICD-10-CM index.

The notes attached to a code

Much of the meaning in ICD-10-CM lives in instructional notes printed beside the codes, not in the descriptions. Notes declared on a category apply to every code beneath it, which is why a code page here shows both its own notes and the ones it inherits, labelled with their source.

The exact effect of each note type is defined in the Official Guidelines, and the difference between Excludes1 and Excludes2 in particular carries real consequences. MedCodeSearch reproduces note text verbatim from the Tabular List and never paraphrases it. E11 is a good page to see this on: it carries Includes terms, an Excludes1 list, and a Use additional code instruction, all inherited by the codes beneath it.

How a release works

ICD-10-CM is republished once a year and takes effect on October 1, the start of the federal fiscal year, which is why releases are named by fiscal year rather than calendar year. The current release is FY 2027, effective October 1, 2026.

Each release restates the entire code set and comes with addenda describing what changed. Codes are added, descriptions are revised, and codes are deleted. A code that is deleted keeps mattering, because records coded under an earlier release still reference it — see What Happens When a Medical Code Is Deleted?. The full, per-release change lists are on Code Updates for FY 2027.

What ICD-10-CM is not

• Not a procedure code set. Inpatient hospital procedures use ICD-10-PCS; supplies, drugs and equipment use HCPCS Level II. See ICD-10-CM vs ICD-10-PCS.
• Not a statement of coverage or payment. A code existing, and even being valid for submission, says nothing about whether a payer will reimburse a claim carrying it.
• Not a diagnosis. A code classifies a documented condition. The clinical documentation is what a code is assigned from.
• Not identical to ICD-10 elsewhere. The U.S. clinical modification is more granular than the WHO base classification, and other countries maintain their own modifications.