ICD-10-CM diagnostic coding: the first-listed diagnosis

All five cases test one rule, and once you have it they take minutes. The *first-listed* diagnosis is an outpatient term — the inpatient equivalent is the principal diagnosis, and they follow different rules — and Section IV of the ICD-10-CM Official Guidelines governs it. The single most consequential clause is this: in the outpatient setting, do **not** code a diagnosis documented as probable, suspected, questionable, rule out, or working. Code instead the condition to the highest degree of certainty known — signs, symptoms, abnormal test results, or the reason for the encounter. That clause decides three of the five cases here, and it is the opposite of the inpatient rule, which is why students who have memorised one confidently get the other wrong. Write the rule at the top of your worksheet before you code anything.

The second rule the set is built around is that the first-listed diagnosis is *the condition chiefly responsible for the services provided during the encounter*, not the patient's most serious problem and not the one listed first in the note. Case 3 is the cleanest test of this: the patient attends for hypertension follow-up and the prescription is renewed, while the nurse separately observes bruising that is not evaluated or treated. The encounter was for hypertension. The bruising may be clinically important — and it may warrant a screening code or a safeguarding conversation — but it did not drive the services provided, so it is not first listed. Ask of every case: what did the provider actually do, and for what? The bruising is a good example of something a coder must notice without coding: the clinical concern is real and belongs in the record, but the coding question is narrower than the clinical one.

Case 1 rewards care with chronology and laterality. The presenting problem is left knee pain with *probable* arthritis, so the probable diagnosis is not coded and the pain is. Everything else in the case is history: an injury twenty years ago and surgery on the *right* knee ten years ago, neither of which is the reason for this encounter. Laterality matters in ICD-10-CM in a way it did not in ICD-9, and the case deliberately puts the old surgery on the opposite knee to catch a reader skimming for a diagnosis. Case 2 works the same way at higher stakes: the suspected myocardial infarction was ruled out by normal EKG and enzymes, so it is not coded, and the confirmed discharge diagnosis of GERD is what the encounter established. Both cases reward the same habit — read the whole vignette, then decide what the encounter was for, rather than coding the first named condition.

Case 4 is the one worth arguing carefully, because it looks like case 2 and behaves differently. The physician documented "likely strep throat" — likely is an uncertain qualifier, so under Section IV that diagnosis is not coded — and the swab result is not back. What is documented with certainty is the presenting symptom, sore throat, which is the reason for the encounter. Case 5 is the counterpart: the lump was removed and pathology *returned* a definitive answer, lipoma, before coding. If the result is available and confirmed at the time of coding, the confirmed diagnosis is used rather than the presenting sign. The pair together is the whole lesson: what matters is the level of certainty documented, not how convincing the story sounds. Treat any case where a specimen has been sent as a question about timing, because the same clinical picture yields a different code depending on whether the result is back.

Practically, work each case in three steps and show them. Identify the reason for the encounter in one phrase. Strip out anything qualified as probable, suspected, likely or rule out. Then code what remains to its highest documented certainty, using the Alphabetic Index first and verifying in the Tabular List — never coding straight from the index, which is where invalid and non-specific codes come from. If the assignment asks only for the diagnosis in words rather than the code, still name the guideline section you applied. An answer sheet that says "sore throat" earns the mark; one that says "sore throat, because Section IV.H bars coding uncertain outpatient diagnoses" demonstrates that the next case will also be right. The rationale line also protects you in a marked worksheet: a right answer with the wrong reasoning is worth less than a wrong answer that shows the rule being applied.