Can you code a rule-out diagnosis at an outpatient visit?

No. At an office visit, you code what you know today: the symptom, sign, or abnormal result that brought the patient in. A diagnosis the note calls "rule out," "probable," or "suspected" stays in the note and off the claim.

The doctor's thinking belongs in the assessment, where the next clinician can follow it. The claim carries only what's confirmed. This is the official ICD-10-CM rule for outpatient visits, and HIPAA requires following these guidelines, so it applies to Medicare, Medicaid, and commercial claims alike.

What does the outpatient rule say?

Section IV.H of the FY2027 guidelines spells it out: "Do not code diagnoses documented as 'probable', 'suspected,' 'questionable,' 'rule out,' 'compatible with,' 'consistent with,' or 'working diagnosis' or other similar terms indicating uncertainty." Instead, code "to the highest degree of certainty for that encounter/visit, such as symptoms, signs, abnormal test results, or other reason for the visit."

Section IV.D backs it up. Symptom codes "are acceptable for reporting purposes when a diagnosis has not been established (confirmed) by the provider." Many of them live in the R00–R99 chapter.

Hospital inpatient coders follow a different rule, and the guidelines call that out. If someone on your billing team came from a hospital, this is the habit to retrain.

What do you code when the diagnosis isn't confirmed?

Here are four primary care notes, what goes on that day's claim, and what changes when the answer comes back. Every code is valid in the FY2027 code set.

What the note says	What goes on today's claim	When the result comes back
"Cough and fever for 4 days. Rule out pneumonia. Chest x-ray ordered."	R05.1 acute cough and R50.9 fever	If the doctor reads the x-ray during the visit and documents pneumonia, it's confirmed today, so J18.9 goes on this claim. If the read comes later, the diagnosis goes on the claim for the visit where the doctor documents it.
"Burning with urination and frequency. Probable UTI. Culture sent, nitrofurantoin started."	R30.0 dysuria and R35.0 frequency	If the follow-up note documents acute cystitis, that visit gets N30.00, plus B96.20 when the culture grew E. coli. Leave N39.0 off; its Excludes1 note covers cystitis.
"Fatigue and constant thirst. Suspect diabetes. A1c and fasting glucose ordered."	R53.83 other fatigue and R63.1 polydipsia	If the doctor diagnoses type 2 diabetes, that visit gets E11.9 or the more specific code the note supports. If the results show prediabetes, it gets R73.03.
"Heartburn after meals for 2 months. Working diagnosis GERD. Trial of omeprazole."	R12 heartburn	If the follow-up note documents GERD, that visit gets K21.9.

Each row follows one habit. The claim for a visit reflects what was known at that visit. A result that arrives tomorrow gets documented in a result note or the next visit's note. It doesn't go back and rewrite the visit where the diagnosis was only suspected.

The note doesn't have to lose anything to make this work. Here's an assessment that keeps the doctor's reasoning and still codes cleanly:

Cough and fever x4 days, crackles at right base. Suspect community-acquired pneumonia. Chest x-ray today. If positive, start antibiotics and recheck in 48 hours.

The claim gets R05.1 and R50.9. The suspicion, the test, and the plan all stay in the chart, where they belong.

Why not just code the suspected diagnosis?

Because a diagnosis code is a statement about the patient, and it sticks. Once E11.9 goes on a claim, the insurance company has it on file. Once it's on the problem list, it rides onto the next claim too. If the A1c comes back at 5.4%, the chart now says diabetes for a patient who doesn't have it, and fixing that takes a corrected claim plus a cleaned-up problem list.

It also opens a gap an auditor can see. A claim code that the note itself calls "rule out" is a code the note doesn't support.

Suspected diagnoses cause a second problem when the answer does come back. Say the problem list already carries R73.03 from last year, and a suspected diabetes gets coded as E11.9 before anyone resolves the old entry. Now the claim has prediabetes and diabetes side by side. R73's Excludes1 note rules out every diabetes code, so the pair can get the claim denied or kicked back. See how Excludes1 pairs work.

Does an unconfirmed diagnosis still count toward the visit level?

Yes. Coding the symptom doesn't shrink the visit. In the AMA's medical decision making (MDM) table, "1 undiagnosed new problem with uncertain prognosis" counts as a moderate problem, the same problem level a 99214 calls for. The AMA defines it as a problem in the differential diagnosis "that represents a condition likely to result in a high risk of morbidity without treatment." A possible pneumonia can fit. A cough the doctor expects to clear on its own doesn't.

So the level comes from the work the note shows, and the diagnosis codes come from what's confirmed. See what else a 99214 note needs to show.

What should you check before the chart is signed?

• Does every code on the claim describe something confirmed today, or a symptom or finding?
• When the assessment says "suspect" or "rule out," does the claim carry the symptom instead of the diagnosis?
• Do the ordered tests and the follow-up plan match the uncertainty in the note?
• When a new diagnosis replaces a placeholder on the problem list, such as prediabetes or a symptom code, is the old entry resolved?

The last question matters for every chronic condition, not only new ones. See which chronic conditions belong on today's claim.

Cortex Lens catches the pair that a suspected diagnosis often leaves behind. When E11.9 lands on a Practice Fusion chart while R73.03 prediabetes is still on the problem list, Lens flags the two as codes that can't be billed together, and your team fixes the list before the claim goes out.

Sources: Uncertain diagnoses, symptom codes, and HIPAA adoption: CMS and NCHS, ICD-10-CM Official Guidelines for Coding and Reporting, FY2027, Sections IV.D and IV.H, pages 113–114, and page 1. The FY2027 guidelines apply October 1, 2026 through September 30, 2027. Code titles and Excludes1 notes: FY2027 ICD-10-CM tabular list. Problem levels in medical decision making: AMA, 2023 CPT E/M Descriptors and Guidelines, Table 1 on pages 8–10 and definitions on page 15.