CRC Risk Adjustment Coding study materials
CRC Risk Adjustment Coding study resources

Six Conditions, One Word: Does "Continue" Meet MEAT?

CRC Risk Adjustment Coding · Put your knowledge to work in context.

All case studies

Six Conditions, One Word: Does "Continue" Meet MEAT?

An internal review pulls the September 3, 2025 annual wellness visit note of a 72-year-old female member from a 14-provider group practice. The note is template-generated. The entire Assessment/Plan reads: "1. CAD - continue. 2. CHF - continue. 3. COPD - continue. 4. CKD stage 4 - continue. 5. Dementia - continue. 6. Depression - continue." Vitals were entered by the medical assistant: BP 138/84, weight 61 kg, BMI 21. The physician spent 12 minutes with the patient and signed the note; no labs, medication changes, or interval history are documented. The medication list shows furosemide, donepezil, tiotropium, and atorvastatin. A separate cardiology visit on March 12, 2025 states "CHF, EF 35%, titrate carvedilol," and an oncology note from 2016 says "breast cancer, completed treatment." The problem list has 14 entries, including "H/O breast cancer" and "tobacco use."

1. Apply the MEAT framework to each of the six conditions and decide what, if anything, can be submitted from the September 3 visit itself.

Read the worked answer

MEAT requires evidence that the provider Monitored, Evaluated, Assessed, or Treated the condition at that encounter. A bare "continue" with no data, no status statement, and no management decision meets none of the four for most conditions.

  • CAD, COPD, dementia, depression: No monitoring data, no severity or functional status, no medication decision. Not supportable from this note.
  • CHF: The only support is the medication list (furosemide), which is a standing order, not a decision made at this visit. Weak to absent. However, the March 12 cardiology note is a face-to-face encounter by an acceptable provider in the same data year documenting CHF with an EF of 35% and a titration plan - that record supports the diagnosis independently, and the September note's medication list corroborates ongoing management.
  • CKD stage 4: A stage is named, but there are no renal function values, no assessment of progression, and no plan. "Continue" is not an assessment. Not supportable from this note; look for a nephrology or lab-linked note in the year.
  • Breast cancer: Treatment completed in 2016. That is personal history (Z85.3), not active malignancy, and history codes do not map to the cancer HCC. Never re-code as active cancer.
  • Tobacco use: Only meaningful if the provider documented current use or dependence (e.g., F17.21x) with an assessment; a problem-list entry alone is not sufficient.

Net result: the September visit yields little on its own, and the compliant capture comes from source records elsewhere in the year.

2. The practice manager asks the coder to "just add an addendum saying everything was reviewed." Explain the compliance boundaries around addenda and queries and describe the compliant path.

Read the worked answer

An addendum is the provider's statement, not the coder's. A coder may never write, dictate, or draft clinical content for a provider's signature, and may never add a diagnosis the provider did not document. An addendum that appears only after an audit selection or a chart-review flag, that backdates a review which did not happen, or that uses boilerplate identical across dozens of charts is a documentation integrity failure - it can convert a coding problem into a false claims problem.

What is permitted:

  1. Retrieve source records. The cardiology note, nephrology note, or lab-linked progress note from the data year is the cleanest fix. A diagnosis supported by another face-to-face encounter in the same year is capturable without touching this note.
  2. Send a compliant query during the normal course of business. The query must be yes/no or multiple choice as appropriate, cite the clinical indicator (e.g., "furosemide on the medication list, EF 35% in March"), present options including "unable to determine," avoid introducing a diagnosis not already implied, and be answered and signed by the provider.
  3. Leave it out if unanswered. No response means no code.

What is not permitted: coaching the provider toward a specific HCC, adding a diagnosis "to be safe," or altering the original note. The query and any response must be retained with the record.

3. What systemic risk does this pattern create under RADV, and what documentation improvement would you recommend that does not turn the record into a coding artifact?

Read the worked answer

If the plan submitted all six conditions from this note, five of six would likely fail validation on medical-record review, because "continue" does not demonstrate assessment at the encounter. Multiply that across a panel and the exposure is not one bad note - it is a chart-level error rate applied to a population. RADV validates the submitted diagnosis against the record; an unsupported diagnosis is an improper payment, and the plan's own internal review can become discoverable evidence that it knew.

The deeper risk is a template that produces identical language for every patient. That is a documented pattern auditors specifically look for, and it also destroys the clinical usefulness of the record.

Recommendations that build real documentation rather than coding artifacts:

  • Restructure the AWV template so each chronic condition requires a status statement plus one of: interval data, a medication decision, or a referral rationale. "Continue" should not be a selectable option.
  • Give providers a one-page MEAT job aid tied to the conditions most common in their panel, not to HCC numbers.
  • Move diagnosis capture upstream: prospective review before claim submission lets the coder retrieve the cardiology note instead of querying after the fact.
  • Track query response rates and template-driven note rates as quality metrics, and audit for copy-forward separately from coding accuracy.
  • Educate on the boundary: documentation drives the code, never the reverse.