6.2 ICD-10-CM Structure & Conventions

Diagnosis Coding and Medical Necessity

ICD-10-CM stands for International Classification of Diseases, 10th Revision, Clinical Modification, maintained jointly by the CDC's National Center for Health Statistics and CMS, with updates effective October 1 each year (with an additional April 1 implementation window available). It is the diagnosis code set that answers the payer's core question: why was the service performed? On the CMS-1500 / 837P claim, the diagnosis codes in Box 21 are tied to procedure lines in Box 24E by pointers, establishing medical necessity.

A biller who recognizes structural and convention errors stops preventable medical-necessity denials.

Code Structure

An ICD-10-CM code is 3 to 7 characters, alphanumeric, and always begins with a letter:

A decimal point follows the third character. The code set is grouped into 21 chapters by body system or condition type — for example, Chapter 9 (I codes) for the circulatory system. Chapter 19 (S00-T88) covers injury, poisoning, and consequences of external causes, and Chapter 20 (V, W, X, Y codes) captures external causes describing how an injury happened (place, activity, intent). External-cause codes are never sequenced first and never stand alone.

Default Codes vs. Combination Codes

A default code is the code listed beside a main term in the Alphabetic Index; it represents the condition most commonly associated with that term when documentation gives no further detail. A combination code is a single code that classifies two diagnoses, a diagnosis with an associated complication, or a diagnosis with an associated manifestation — for example, one code capturing both type 2 diabetes and diabetic nephropathy (E11.21). Combination codes must be used when one exists; reporting two separate codes instead is an error.

Sequencing Instructions

Two instructional notes drive code order:

• "Code first" — an underlying/etiology condition must be sequenced before this code (often a manifestation code printed in italics with brackets).
• "Use additional code" — an additional code is required to fully describe the condition, sequenced after this code.

Ignoring these notes produces sequencing denials and incorrect medical-necessity logic, especially in etiology/manifestation pairs.

Excludes1 vs. Excludes2 — A High-Yield Distinction

The two Excludes notes are routinely tested and routinely confused on the CPB exam.

Excludes1 is an absolute prohibition: the two codes describe conditions that cannot occur together (e.g., a congenital form vs. an acquired form of the same condition). Excludes2 simply means the excluded condition is not part of the code; if the patient genuinely has both, the biller reports both codes. A 2019 ICD-10-CM guideline exception allows two Excludes1 codes together only when the conditions are clearly unrelated — but for exam purposes, treat Excludes1 as "do not code together."

Placeholder X and 7th-Character Extensions

Some codes require a 7th character but have fewer than six preceding characters. The placeholder X fills the empty positions so the 7th character lands correctly in the seventh slot. For example, the poisoning code T36.0X1A uses X as a fifth-character placeholder; without it the 7th-character "A" would sit in the wrong position and the code would reject as invalid.

For injury and many Chapter 19 codes, the 7th-character extension identifies the encounter:

Fractures use an expanded set (e.g., A, D, G, K, P, S for healing status), but the A/D/S pattern is the foundation. A sequela (S) code requires two codes: the sequela code plus the code for the residual condition.

Common Biller Traps

• Submitting an injury code missing its required 7th character — the most frequent ICD-10-CM rejection.
• Forgetting the placeholder X, pushing the 7th character into the wrong slot.
• Reporting an external-cause (V/W/X/Y) code as the first-listed/primary diagnosis — never allowed.
• Coding two separate diagnoses when a combination code is mandated, or ignoring a "code first" note and sequencing the manifestation first.

Specificity and Laterality

ICD-10-CM rewards specificity. Many codes require a character for laterality — 1 for right, 2 for left, 3 for bilateral, and 0 or 9 for unspecified. Payers increasingly deny unspecified codes when a more specific code is documented, so a biller who sees a vague "unspecified site/side" code on a claim with a detailed note should query for the specific code. Reporting a laterality of "unspecified" when the chart says "left" is a documentation-to-code mismatch that fails medical-necessity edits.

Worked Example: Sequencing an Etiology/Manifestation Pair

A patient has type 2 diabetes with diabetic chronic kidney disease and stage 3 CKD. The diabetes code (E11.22) carries a "use additional code" note directing the biller to also report the stage of CKD (N18.3x). The biller sequences E11.22 first, then N18.3x, because the diabetes is the underlying condition and the CKD stage further describes it. Reversing the order, or omitting the CKD code, breaks the sequencing rule and can trigger a denial for incomplete coding. This is the everyday application of "code first" and "use additional code" that the CPB exam tests with scenario questions.