Suspected COPD, Prediabetes, and a Consultant Who Says "Capture Everything"
A 58-year-old man is seen October 21, 2025 for his first visit at a new practice. The assessment reads: "1. History of MI 2019, two stents, no chest pain; continue aspirin 81, metoprolol, atorvastatin. 2. Suspected COPD - 30 pack-years, quit 2019; will order PFTs; started tiotropium today and albuterol PRN. 3. Prediabetes, A1c 6.1% - recheck in 6 months. 4. BMI 35.2 - counseled on diet. 5. Father had colon cancer at 62 - screening due." The only respiratory finding documented is "mild end-expiratory wheeze." The plan also reads: "if PFTs confirm COPD, document at next visit." An external consultant has told the practice that every diagnosis mentioned anywhere in the chart should be submitted because "risk adjustment is about documenting everything the patient has." The practice's coder is asked to build the submission file from this note.
1. For each of the five assessment entries, state whether a diagnosis can be submitted and identify the governing rule.
Read the worked answer
- History of MI: Codeable as I25.2, old myocardial infarction, which is a current-status code rather than a personal-history Z code. The record documents the event, the stents, and ongoing therapy with no current symptoms, and the index directs "history of myocardial infarction" to old MI. Whether I25.2 maps to a payment HCC must be confirmed against the current model's mapping file; the mapping governs, not the code title.
- Suspected COPD: Not codeable. ICD-10-CM outpatient guidelines prohibit coding diagnoses documented as suspected, probable, questionable, rule out, or working diagnosis; code the signs and symptoms to the highest degree of certainty instead. The documented wheeze is a codable sign. The provider has explicitly deferred confirmation to PFTs, which settles the question.
- Prediabetes: Codeable as R73.03. It must never be coded as diabetes - the A1c of 6.1% does not meet diabetes criteria, and R73.03 maps to no diabetes HCC. Substituting E11.9 would be a fabricated diagnosis.
- Obesity: Codeable - the provider documented a BMI of 35.2 with counseling. Use the obesity code supported by the documentation plus Z68.35 as a secondary BMI code; BMI codes are never first-listed. Verify mapping rather than assuming an HCC.
- Family history of colon cancer: Codeable as Z80.0 for completeness, but it is family history, not disease in this patient, and generates no HCC.
The consultant's instruction is wrong on its face and, if followed, converts a documentation problem into a false claims exposure.
2. The provider argues: "I started tiotropium, so I treated the COPD - it should count." Address the argument and explain what the coder may and may not do here.
Read the worked answer
Treatment alone does not convert a suspected diagnosis into a confirmed one. Two separate requirements have to be met: the diagnosis must be established with an acceptable degree of certainty, and it must be documented and assessed at a face-to-face encounter. Starting a trial of tiotropium while awaiting PFTs is, by the provider's own words, a diagnostic trial. The provider wrote both "suspected" and "if PFTs confirm" - the record itself states the condition is unconfirmed. Coding it would contradict the note it came from.
What the coder may do:
- Capture the documented sign (wheeze) and any confirmed conditions.
- Send a non-leading, clinically grounded query if the record is ambiguous - but there is no ambiguity here; the provider has already answered, and re-querying is pressure, not clarification.
- Flag the chart for follow-up once PFTs are completed, so the diagnosis is captured at the next encounter if confirmed.
What the coder may not do:
- Code J44.9 on the theory that treatment implies diagnosis.
- Add "probable COPD" and let downstream review sort it out.
- Coach the provider by naming the HCC or the payment effect in a query.
If the PFTs do confirm COPD, the correct capture point is the visit at which the provider documents the confirmed diagnosis and assesses it - there is a natural, documented path here, and the compliant answer is simply to wait for it.
3. The provider asks how "history of MI" should be handled going forward, since the patient has had no chest pain for six years. Explain the coding logic and its limits.
Read the worked answer
Old MI is a continuing clinical status, not a resolved event to be filed away. I25.2 exists precisely for a documented history of myocardial infarction without current symptoms, and it is appropriate to continue reporting year over year as long as the provider documents and assesses it - which here means confirming the history, the stent status, and the ongoing antiplatelet and beta-blocker therapy. That is an assessment; a problem-list entry copied forward is not.
The limits matter more than the label:
- Do not substitute a personal-history Z code (for example a circulatory history code) for I25.2 unless the documentation actually frames it as history without ongoing relevance; the two carry different meanings, and the substitution may drop the condition out of the model.
- Do not assume it pays. Some ischemic heart disease codes map to a payment HCC and some do not; the mapping file decides. Coders should verify rather than reason from the code's name.
- Do not upgrade. Six years without symptoms is not an acute MI, and no code implying current acute ischemia should appear.
- Recapture annually. The diagnosis must be re-documented and assessed each data year; it does not carry over automatically.
The recurring failure mode with this kind of entry is a chart where the condition is real, but the year's note only carries a problem-list line. The fix is a status statement and a management decision in the encounter, not a coding workaround.