ICD-10-CM guide
ICD-10-CM Specificity: When an Unspecified Code Is Correct
How to code to the highest supported specificity, when an unspecified code is the right answer, and how to check that a code is complete.
Reviewed 2026-09-16 | 2 min read
New coders often treat unspecified codes as mistakes and specific codes as wins. The guidelines say something more practical: report the code that most accurately reflects what is known about the patient at that encounter, coded to the highest level of specificity the record supports.
Codes grow character by character
An ICD-10-CM code can have three to seven characters. Three-character codes are category headings that may subdivide into fourth, fifth, sixth, and seventh characters. A three-character code is used only when the classification does not subdivide it further.
A code is invalid when it is not coded to the full number of characters required for that code, including a seventh character when one applies.
Code to the highest supported specificity
Within the characters the record supports, go as specific as the documentation allows. Laterality, encounter type, and healing status are common examples. If the record documents the right knee, a code for an unspecified knee is not the best answer.
The limit is the documentation, not your knowledge of codes. A more specific code the record does not support is not a better code.
Unspecified codes are not errors
Sign, symptom, and unspecified codes have acceptable and even necessary uses. When a diagnosis is known but its type is not, such as pneumonia without a documented organism, the unspecified code most accurately reflects what is known.
The guidelines are explicit that selecting a specific code the medical record does not support is inappropriate, and that care should not be ordered just to justify a more specific code.
How to check that a code is complete
- Find the term in the Alphabetic Index first, then verify the code in the Tabular List.
- Read the Tabular entry for required additional characters and for notes that direct you to a more specific code.
- Confirm that every element the code names, such as site, laterality, acuity, or status, appears in the documentation.
- When the record lacks an element the code requires, move to the code whose elements are all supported.
Where coders go wrong
- Stopping at the category code when the Tabular List subdivides it.
- Choosing an unspecified code out of habit when the record actually supports specificity.
- Choosing a specific code the record does not support, which adds risk without improving accuracy.
- Ignoring laterality or a seventh character because the code looked complete at five characters.
Questions coders ask
- Are unspecified codes wrong?
- No. An unspecified code is correct when it describes the encounter more accurately than any specific code the documentation supports.
- Should I ask the provider for more detail?
- When the record is incomplete, a documentation query is the right path. The guidelines describe complete and consistent documentation as a shared responsibility of the provider and the coder.
- How do I know a code needs more characters?
- The Tabular List shows subdivisions and required characters. A code that is missing required characters is invalid, even if the description looks close to the note.
- Can specificity ever hurt?
- Only when it goes beyond the record. Specificity that the documentation supports is always the goal.
Practice these rules
Sources
ICD-10-CM Official Guidelines for Coding and Reporting FY 2027
Section IV.F.1, ICD-10-CM codes with 3, 4, 5, 6 or 7 characters
ICD-10-CM Official Guidelines for Coding and Reporting FY 2027
Section IV.F.2, Use of full number of characters required for a code
ICD-10-CM Official Guidelines for Coding and Reporting FY 2027
Section IV.F.3, Highest level of specificity
ICD-10-CM Official Guidelines for Coding and Reporting FY 2027
Section I.B.18, Use of Sign/Symptom/Unspecified Codes