Medical Coding Examples: Outpatient ICD-10-CM Cases

Medical coding examples show how documented care becomes a code—and why a small change in the record can change the answer. For example, documented dysuria can support R30.0; an unconfirmed outpatient urinary infection does not become a confirmed diagnosis simply because it appears in the differential. The useful skill is explaining the decision, not memorizing a code beside a disease name.

This guide uses seven original fictional cases: six diagnosis-coding decisions and one example of when an office-visit service code cannot yet be selected. They contain no real patient records. Code entries were checked against the official April 2026 ICD-10-CM files on September 11, 2026; separate physician/coder review has not been completed. Final code selection requires qualified human review of the complete record.

Vero publishes this guide and offers clinical documentation software. These are teaching cases, not measured product results. See the editorial and correction policy.

Choose the code set and effective release

Medical coding is broader than diagnosis coding. In the United States, ICD-10-CM describes diagnoses, symptoms and reasons for encounters. CPT/HCPCS describe services and other reportable items; ICD-10-PCS is used for hospital inpatient procedure reporting. A diagnosis code does not tell you which office-visit level, procedure or supply code to bill. CMS overview of coding and classification systems.

All diagnosis examples below assume a U.S. outpatient encounter dated September 11, 2026. The April 1, 2026 FY2026 update applies to encounters from April 1 through September 30, 2026. FY2027 files apply beginning October 1, 2026; advance publication does not make them effective early. For hospital inpatient coding, use the applicable discharge-date rules and inpatient guidance. CMS release dates and files, CDC ICD-10-CM files.

For Canada, do not transplant these ICD-10-CM answers into an ICD-10-CA workflow. CIHI identifies version 2022 ICD-10-CA/CCI and associated standards for use with 2026–2027 data. Canadian facility reporting and provincial physician billing also involve different requirements. Identify the destination data set and local billing system before adapting a case. CIHI classification release information.

Seven medical coding examples

Read the documented facts first, decide what you would code or hold, then open the reasoning. Each answer addresses the stated coding question; the whole claim still needs review for other conditions, services, sequencing and payer requirements. For detailed search mechanics, use the separate ICD-10-CM lookup guide.

Reproduce the source trails: Open the official April 2026 XML archive. Inside it, use icd10cm-index-April-1-2026-XML.xml for the Index routes and icd10c-tabular-April-1-2026.xml for the code entries and parent instructions below. Search the named main term, follow its nested subterms, then locate each code in the Tabular file. Parentheses denote nonessential modifiers; the arrows below show successive Index levels. The listed instructions explain each decision, not every rule that could apply to a complete record.

1. Dysuria with a possible urinary infection

Record: At an outpatient visit, the clinician documents dysuria and writes “possible urinary tract infection; culture pending.” The final assessment available for this exercise does not establish an infection. No other reportable condition is supplied.

Decision: Should the coder select a urinary infection code, report the symptom or add both automatically?

Error to catch: Turning “possible” into “confirmed” while transferring the assessment into billing. Preserve the original wording in the review trail. Official April 2026 Guidelines, Section IV.H.

2. An elevated reading versus established hypertension

Record A: The clinician assesses an elevated blood-pressure reading and explicitly states that hypertension has not been diagnosed. Record B: At a different fictional visit, the final assessment documents essential hypertension under active management, with no documented pregnancy, secondary cause, hypertensive heart disease or kidney disease.

Decision: The screen shows a blood-pressure value in both records. Does that make the diagnosis code the same?

Error to catch: Inferring a new diagnosis from a vital sign or dropping an established condition because a single result looks normal. Compare the provider’s assessment with the code definition and inherited notes. April 2026 XML archive — R03.0 and I10 entries.

3. Type 2 diabetes with stage 3b chronic kidney disease

Record: The provider explicitly documents type 2 diabetes with diabetic chronic kidney disease, stage 3b. Both are addressed at the encounter. The example supplies no medication-status information and asks only for the diabetes/CKD code pair.

Decision: Is E11.9 enough because “type 2 diabetes” appears in its description?

Error to catch: Choosing an uncomplicated-diabetes code and losing the documented complication, or reporting the combination code without its required stage detail. Keep both the E11.22 instruction and the N18 parent instruction in the evidence trail. April 2026 XML archive — E11.22 and N18 entries.

4. An ankle sprain during active treatment or routine healing

Record: The diagnosis is a sprain of the right ankle, with the ligament unspecified. No associated open wound is documented. In version A, the patient receives active treatment. In version B, active treatment is complete and the visit provides routine care during recovery.

Decision: Does “first time this clinician has seen the patient” decide the seventh character?

Error to catch: Submitting the incomplete stem S93.401 or choosing D solely because a prior appointment occurred. The code expresses the documented treatment phase. Official Guidelines, Section I.C.19.a.

5. Colon screening versus evaluation of symptoms

Record A: An asymptomatic patient attends specifically for colon cancer screening. The supplied record has no relevant family history or discovered condition. Record B: A patient is referred for evaluation of a documented symptom; the request is to investigate that symptom, not routine screening.

Decision: Can the same screening diagnosis be used because both pathways might involve the same test?

Error to catch: Letting the name of a planned test override its documented purpose. Record the reason for the service before working on its procedure coding. Official Guidelines, Section I.C.21.c.5.

6. Follow-up after a resolved urinary infection

Record: Treatment for a urinary tract infection is complete. At a follow-up examination, the clinician documents that the infection has resolved, no symptoms remain and no further treatment is needed. The purpose is surveillance after completed treatment.

Decision: Should the old active-infection code be copied forward?

Error to catch: Treating yesterday’s problem list as today’s assessment. Reconcile active conditions with resolved conditions before selecting codes. April 2026 XML archive — Z09 and Z87.440 entries; Guidelines I.C.21.c.8, pages 100–101.

7. A diagnosis does not determine the office-visit level

Record: A note says “diabetes follow-up,” lists diagnoses and contains several pages of templated history. The exercise does not supply sufficient medical-decision-making information or qualifying practitioner time for an office/outpatient E/M level.

Decision: Can a longer note or the diabetes code justify a particular visit level?

CMS’s May 2026 E/M guide emphasizes documentation support and medical necessity, and says documentation volume should not determine the service level. CMS E/M documentation, page 22; service-level selection, page 24.

From a worked example to a reviewable record

A code with no reasoning is difficult to audit. A useful review record connects the actual documentation to the effective source, the decision and the person responsible for final selection. Vero’s editorial workflow below assigns a concrete output to each handoff.

1. 1. Clinician → final assessment: preserve the diagnosis, certainty, site, relationships and reason for the encounter.
2. 2. Coder → candidate and source: record the date-effective Index path, complete Tabular code and relevant parent instructions.
3. 3. Clinician and coder → resolved question: clarify conflicting or missing information through the approved process.
4. 4. Qualified reviewer → selection: confirm the supported codes, required companions, sequence and remaining limitations.
5. 5. Billing or reporting team → destination: verify that the approved selection, service date and version reach the intended claim or data submission.

For the diabetes/CKD example, the record would identify the explicit clinical relationship, the stage 3b statement, the E11.22 additional-code instruction and the N18 code-first instruction. “Diabetes code found” would lose the reason for the second code. Conversely, the E/M example should finish with a documented hold and the missing evidence, not a guessed answer.

Keep the underlying clinical record in the authorized system. A separate training log can use case IDs such as EX-03 and the relevant coding rule. If your team later adds real cases, use its approved de-identification process and access controls; removing a name alone does not establish de-identification. HHS de-identification guidance.

A quality checklist for coding examples

Use this editorial checklist to explain a decision consistently. It is deliberately a review aid: checking every box records the work completed, while the authorized reviewer remains responsible for the final selection. The source checks correspond to the official Index/Tabular conventions, setting-specific guidance and CMS documentation requirements used above.

MEDICAL CODING REVIEW RECORD
Case ID:
Encounter or discharge date:
Jurisdiction / care setting / code set:
Effective release:
Documented facts and certainty:
Index route:
Tabular candidate and inherited instructions:
Required companion codes and sequence:
Missing evidence / clarification:
Final disposition: select / hold / clarify
Reviewer, role and date:

REVIEW CHECKS
[ ] Identify the jurisdiction, code set and care setting.
[ ] Select the release effective for the encounter or discharge date.
[ ] Use the completed clinical documentation, including uncertainty.
[ ] Record the Alphabetic Index term and applicable subterms.
[ ] Verify the complete code in the Tabular List.
[ ] Check parent-category instructions and exclusions.
[ ] Confirm documented site, side, stage and treatment phase.
[ ] Apply required companion codes and sequencing instructions.
[ ] Separate active disease, history, screening and follow-up.
[ ] Keep diagnosis coding separate from service-level selection.
[ ] Resolve missing or conflicting evidence through the approved query process.
[ ] Record qualified human review and the final disposition.

EX-03 — FICTIONAL COMPLETED TEACHING RECORD
Case ID: EX-03 (no patient record)
Encounter date: September 11, 2026
Jurisdiction / care setting / code set: United States / outpatient / ICD-10-CM
Effective release: April 1, 2026 FY2026 update

Documented facts and certainty:
The provider explicitly documents type 2 diabetes with diabetic chronic kidney disease, stage 3b. Both conditions are addressed. Medication status and other associated conditions are not supplied.

Index routes:
Diabetes, diabetic > type 2 > with > chronic kidney disease > E11.22
Disease, diseased > kidney > chronic > stage 3b > N18.32

Tabular candidates and inherited instructions:
E11.22: type 2 diabetes with diabetic CKD; add the CKD-stage code.
N18.32: CKD stage 3b; parent N18 directs coding associated diabetic CKD first.
Parent E11: check applicable long-term medication-use codes.
Parent N18: check applicable cachexia and kidney-transplant status.

Required companion codes and sequence:
E11.22 followed by N18.32. This establishes this pair's order, not the first-listed diagnosis or complete code set for a real claim.

Missing evidence / clarification:
Medication status, other reportable conditions and the complete encounter require review. Do not infer stage from a laboratory value or invent medication use.

Final disposition:
The teaching answer supports E11.22 + N18.32 in that order. Hold any real submission until the complete record and other applicable requirements have been reviewed.

Reviewer, role and date:
No qualified coding sign-off recorded. Editorial source check: September 11, 2026.

Reproduction sources:
CDC archive: https://ftp.cdc.gov/pub/Health_Statistics/NCHS/Publications/ICD10CM/2026-update/icd10cm-April-1-2026-XML.zip
Index file: icd10cm-index-April-1-2026-XML.xml
Tabular file: icd10c-tabular-April-1-2026.xml
Locate the Index paths and E11, E11.22, N18 and N18.32 entries above.

When two reviewers disagree, classify the disagreement before counting it as an error. They may be using different release files, different versions of the clinical note or different assumptions about the encounter. Resolve those inputs first. Then document whether the disagreement concerned the code itself, a missing companion code, sequence, unsupported specificity or a service-level assumption.

Implement the cases in a clinic workflow

Begin with a small calibration session

Give the clinician and coder the same seven case descriptions with the answers closed. Ask each to record a candidate or a hold, the source and the fact that could change the answer. Review disagreements together. This exercise is more revealing than asking whether a code label looks plausible because it exposes the assumptions behind the selection.

The cases cover a narrow group of outpatient decisions. Add locally relevant cases before using the exercise for staff competency assessment: specialty conditions, documented complications, incomplete records and setting changes. Include cases where the correct operational action is clarification rather than a more specific code. Do not reward choosing a code when the information is missing.

Test software suggestions against the same evidence

Run synthetic cases through an approved test environment. Record the software version, release configuration, input and proposed output. Check whether a suggestion drops uncertainty, supplies an undocumented stage, confuses screening with diagnostic evaluation, omits a companion code or treats note length as service complexity.

Use the final clinician-approved note as the coding input. If an AI-generated draft introduced an unsupported diagnosis, correcting only the billing code leaves an inaccurate medical record behind. Send the problem back to documentation review as well. The clinical documentation guide covers source attribution and correction boundaries; the digital scribe guide explains draft-to-record review.

Measure quality before speed

For a local evaluation, define the denominator before reporting a result. Separate cases with complete required codes, cases with an incorrect extra code, cases needing a query and cases that failed to reach a decision. Also record review minutes and repeated manual work. One overall “accuracy” number can conceal the difference between missing a stage code and inventing a diagnosis.

Agree in advance who adjudicates disagreements and who can approve use in the production workflow. A faster suggestion is valuable only when the reviewed record and final selection remain supported. No product accuracy or time-saving result is reported by these teaching cases.

Recheck at a release or workflow change

The next U.S. annual release begins October 1, 2026. Before using these cases for a later date, check every code and relevant instruction against the new files. Keep the earlier expected answer with its original date so the team can reproduce old decisions rather than overwriting history.

Repeat the affected tests after changing templates, diagnosis extraction, mapping rules or claim interfaces. The EHR integration guide provides a broader framework for checking what reaches the destination. A successful search in one application does not demonstrate that the complete approved code set, sequence and encounter date survived transfer.

Sources for repeating the checks

Use the dated release files, not a search snippet, when reproducing a worked answer. The source checks for this edition were completed September 11, 2026.

• CDC April 2026 release directory: official Guidelines, XML, Index and Tabular files
• CDC ICD-10-CM release and file overview
• CMS code-set overview
• CMS ICD-10 effective dates and annual releases
• CMS E/M documentation and service selection guidance, May 2026
• CIHI Canadian classification version information
• HHS guidance on de-identifying health information