How to Choose an ICD-10-CM Code: A Step-by-Step Coding Workflow

Almost every recurring coding error described across these guides is a step left out of one workflow: a code taken from the Index without opening the Tabular; a 7th character missed; an Excludes1 read as advice; a manifestation code sequenced first; a status code added to a complication code that already carries it. The Guidelines describe the workflow in two sentences and then spend a hundred pages on its details. Section I.B.1:

Official source

“To select a code in the classification that corresponds to a diagnosis or reason for visit documented in a medical record, first locate the term in the Alphabetic Index, and then verify the code in the Tabular List. Read and be guided by instructional notations that appear in both the Alphabetic Index and the Tabular List.”

“It is essential to use both the Alphabetic Index and Tabular List when locating and assigning a code. The Alphabetic Index does not always provide the full code. Selection of the full code, including laterality and any applicable 7th character can only be done in the Tabular List. A dash (-) at the end of an Alphabetic Index entry indicates that additional characters are required. Even if a dash is not included at the Alphabetic Index entry, it is necessary to refer to the Tabular List to verify that no 7th character is required.”

And Section I.B.2 sets the bar for the finished code: “Diagnosis codes are to be used and reported at their highest number of characters available and to the highest level of specificity documented in the medical record. … A three-character code is to be used only if it is not further subdivided. A code is invalid if it has not been coded to the full number of characters required for that code, including the 7th character, if applicable.”

• Alphabetic Index instructions — “see” (mandatory), “see also” (optional), “with” subterms (a presumed relationship), bracketed manifestation codes (sequenced second), nonessential modifiers in parentheses.
• Tabular List instructions — Includes notes and inclusion terms; Excludes1 (not coded here) and Excludes2 (not included here); Code First / Use Additional Code (a sequencing pair) and Code Also (no order); the 7th-character tables and placeholder X; laterality characters.
• Official Guidelines — the conventions above (I.A), the general rules (I.B: specificity, symptoms, multiple coding, combination codes, sequela, laterality), chapter-specific rules (I.C), and the principal / first-listed selection rules (II, IV).
• CMS editing logic — the Medicare Code Editor, which re-checks several of these rules on inpatient claims. It is downstream of the classification, never a substitute for it.
• Documentation-dependent reasoning — the provider’s diagnostic statement is the basis of code assignment (I.A.19); clinical criteria and laboratory values are not.

1. Identify the documented condition. Code the provider’s diagnostic statement; a symptom when no diagnosis was established (I.B.4); an uncertain diagnosis as established on an inpatient discharge but not in the office (II.H, IV.H).
2. Locate the main term in the Alphabetic Index — the condition, not the anatomy — and read the nonessential modifiers.
3. Follow all subterms and cross-references. A “see” must be followed; a “see also” only if the current entry lacks the code; a “with” subterm presumes the relationship (I.A.15).
4. Verify the code in the Tabular List. Open the category first and read its Includes and Excludes notes; then the code.
5. Read every instructional note at the code and its ancestors: Code First, Use Additional Code, Code Also, Excludes1, Excludes2, and any 7th-character table (the Excludes guide and the Code First guide walk each).
6. Confirm required specificity — the highest number of characters available and the highest specificity the record supports (I.B.2); an unspecified code only when the record gives no more (I.B.18).
7. Add laterality where the code carries it — bilateral when both sides, unspecified side rarely (I.B.13).
8. Add the 7th character where required, in the seventh position, with placeholder X if needed (I.A.4–5) — by phase of care, not visit count.
9. Determine sequencing — the Tabular’s Code First / Use Additional Code notes first, then Section II (inpatient) or IV (outpatient).
10. Add required associated codes — the manifestation, the organism, the stage, the control agent, the external cause — whatever the notes and chapter guidelines call for (I.B.7).
11. Check the code relationships — an Excludes1 pair on the same claim, a manifestation code leading, a status code beside the complication that includes it.
12. Validate the final code set — Claim Check applies the classification’s notes and the published CMS edits to the set; it does not decide what the record says.

1. Documentation: type 2 diabetes with CKD stage 3a, on metformin. Coding question: one code or two, and in what order? Analysis: Index “Diabetes › type 2 › with › chronic kidney disease” → E11.22 (the “with” presumption); Tabular E11.22 carries Use Additional Code for the N18 stage; N18 carries Code First for E11.22; E11 asks for Z79.84. Result: E11.22 → N18.31 → Z79.84. Why: the notes at both ends fix the order; the stage and control agent are the associated codes the notes require.
2. Documentation: closed fracture of the shaft of the right radius, first orthopedic visit two days after the emergency department applied a splint. Coding question: which 7th character? Analysis: active treatment is still under way; the character follows phase, not provider. Result: the S52.3- code for the documented fracture type with 7th character A. Why: I.C.19.c.1 assigns the initial-encounter character for each encounter of active treatment.
3. Documentation: office visit, cough and fever, “probable pneumonia,” imaging pending. Coding question: pneumonia or symptoms? Analysis: outpatient uncertain diagnosis (IV.H). Result: R05.9 and R50.9. Why: the outpatient rule codes to the highest certainty for the visit; the same chart on an inpatient discharge would be J18.9.

• Stopping at the Alphabetic Index.
• Skipping the category’s notes and reading only the code.
• Confusing Excludes1 with Excludes2, or Code Also with Code First.
• Assuming “initial encounter” means first visit.
• Missing laterality or the 7th character — the code is invalid without them.
• Reporting a code at fewer characters than the classification provides.
• Choosing a specific code the record does not support, or ordering tests to reach one.
• Sequencing by habit rather than by the notes and Section II / IV.

• The provider’s diagnostic statement, and its certainty.
• Type, site, side, phase of care, cause — each maps to a character or a note.
• The relationship between two conditions, unless the classification presumes it (“with”).
• The reason for the encounter, which drives sequencing.