Conditions Qualify for CCM

Which conditions qualify for CCM?

Conditions qualify for CCM when a Medicare patient has two or more chronic conditions expected to last at least 12 months, or until death, that place the patient at significant risk of death, acute exacerbation or decompensation, or functional decline. CMS gives examples such as diabetes, hypertension and COPD, but no closed list of qualifying diagnoses.

The definition has three parts, and all three must hold: a count (two or more), a duration (12 months or longer) and a risk (death, exacerbation or decline). A screening tool can test the first two from structured data. The third needs clinical judgement, which is why software should propose candidates rather than enroll them.

What does CMS say about eligible patients?

CMS says the criteria in the CPT code descriptors are the only diagnostic test. It states the rule in its Chronic Care Management Services booklet (MLN909188) and repeats it in its CCM billing FAQ.

• No fixed list: The CMS FAQ says CCM is not limited to a specific list of illnesses; any condition that meets the descriptor criteria can count.
• Examples, not limits: The MLN booklet names conditions such as Alzheimer's disease and related dementia, arthritis, asthma, atrial fibrillation, cancer, COPD, depression, diabetes, hypertension and substance use disorders.
• Suggested signals: CMS notes that practitioners may use criteria suggested in CPT guidance, such as the number of illnesses, the number of medications, or repeat admissions and emergency department visits.
• Complex CCM: Patients billed under the complex codes 99487 and 99489 also need moderate or high complexity medical decision making.

Payer matters too. These are Medicare's rules; Medicaid programs and commercial plans that pay for care management set their own criteria, so a screen built for Medicare should take the payer as an input rather than assume it.

How do you turn a problem list into an eligibility screen?

Start from the coded problem list, group diagnoses into chronic condition categories, count distinct categories per patient, and keep anyone with two or more as a candidate. Then add risk signals so the care team reviews the highest-need patients first.

• 1. Pull active problems: Read problem list entries with an active status, coded in ICD-10-CM or SNOMED CT, and drop resolved or entered-in-error rows.
• 2. Map to condition groups: Map each code to a chronic condition group, so that two diabetes codes count as one condition, not two.
• 3. Apply duration: Exclude groups that are usually self-limiting, and flag codes whose chronicity depends on context, such as a single episode of depression.
• 4. Count groups: Keep patients with two or more distinct chronic groups as candidates.
• 5. Add risk signals: Attach admissions, emergency visits, medication count and recent abnormal results from the past year.
• 6. Check exclusions: Remove patients already enrolled with another billing practitioner, in hospice, or who declined CCM recently.
• 7. Route for review: Send the ranked list to a clinician, who confirms the risk element and decides whether to offer the service.

Keep the mapping table under version control with a named owner. ICD-10-CM changes every October, and a map that misses new codes quietly drops patients from the list without anyone noticing.

Which data points make the screen more accurate?

Encounter history, medication lists and lab results make the screen more accurate than diagnosis codes alone. Problem lists are often stale or incomplete, so supporting data confirms that a condition is active and shows which patients carry the most risk.

• Encounter diagnoses: A diagnosis billed at two or more visits in the past year confirms an active condition that the problem list may have missed.
• Medication list: Active prescriptions such as insulin, anticoagulants or inhaled steroids corroborate a condition and feed the medication count that CPT guidance suggests.
• Utilization: Admissions and emergency visits in the last 12 months are a direct signal of exacerbation risk.
• Lab trends: Rising creatinine, a poorly controlled A1c or a falling ejection fraction mark conditions that are getting worse.
• Functional status: Fall history, mobility aids and activities of daily living assessments capture the functional decline part of the rule.
• Care gaps: An overdue annual wellness visit identifies patients who will also need an initiating visit before CCM can start.

Store every signal with its source and date. When an auditor or a clinician asks why a patient was flagged, the answer should come from the record, not from a score nobody can explain.

What should the reviewing clinician see on the worklist?

The reviewing clinician should see the evidence behind each flag on one screen: the qualifying conditions, the risk signals and the reason the patient ranks where they do. A bare list of names forces a chart dive for every row, and reviewers stop using it within weeks.

• Qualifying groups: The two or more chronic condition groups found, with the source code, the date first recorded and the last visit that addressed each one.
• Risk summary: Admissions, emergency visits, medication count and the most recent abnormal result, each with a date.
• Billing context: Payer, last visit date, whether an initiating visit is needed, and any CCM or PCM billed elsewhere this year.
• Prior outreach: Earlier offers, declines and the reason given, so the team does not call someone who said no last month.
• Decision buttons: Offer CCM, offer PCM, not appropriate, or needs more review, each with a short required reason.

Feed the decisions back into the screen. If reviewers keep rejecting one condition group as low risk, that is a signal to change the ranking, not to keep sending the same patients. The reasons also give the program a defensible record of why some patients were offered the service and others were not.

What has to happen before an eligible patient is enrolled?

Before the first CCM claim, every eligible patient needs documented consent, and some also need an initiating visit. Screening only finds candidates; enrollment follows the CMS service requirements in MLN909188.

• Initiating visit: Required for new patients and for patients not seen in the previous year, held during a comprehensive E/M visit, an annual wellness visit or an initial preventive physical exam.
• Consent: Written or verbal consent before billing, with the patient told about cost sharing, that only one practitioner can bill CCM in a calendar month, and that they can stop at any time.
• Certified EHR: Demographics, problems, medications and medication allergies recorded in certified EHR technology.
• Care plan: An electronic comprehensive care plan that is established, revised or monitored as the codes describe.

Build the screen so a confirmed candidate flows straight into these steps: an outreach task, a consent form, and a scheduling request for the initiating visit when the last qualifying visit is more than a year old.

When does PCM fit better than CCM?

Principal care management (PCM) fits when one high-risk condition drives the work. CMS describes PCM as care for a single chronic condition expected to last at least 3 months that places the patient at significant risk of hospitalization, exacerbation, functional decline or death.

A screen should therefore produce two lists, not one. A cardiology patient with heart failure whose other problems are managed elsewhere belongs on a PCM list. A patient with two or more chronic groups and a primary care relationship belongs on the CCM list. A clinician confirms which service matches the care actually delivered.

• CCM: Two or more chronic conditions lasting 12 months or longer, a comprehensive care plan, usually led by primary care.
• PCM: One complex condition lasting 3 months or longer, a disease-specific care plan, often led by a specialist.
• Neither: Patients whose conditions are stable and low risk, where outreach time is better spent on someone else.

Where does automated CCM screening go wrong?

Automated CCM screening goes wrong when it counts codes instead of conditions or skips the risk test. These are the failure points that show up most in chart review.

• Counting codes, not conditions: Several codes for one disease push a patient over the threshold of two.
• Stale problem lists: Resolved problems never closed in the chart keep patients on the list for years.
• Acute codes treated as chronic: Injuries, infections and pregnancy codes slip into the count.
• No risk element: The screen counts conditions but never checks risk, so stable patients crowd the worklist.
• Duplicate billing: The screen ignores enrollment with another practice, and two practitioners bill CCM for the same month.
• Silent auto-enrollment: Software marks patients as enrolled before consent is documented.

Each of these surfaces later as a denied claim, an audit finding or a care team that stops trusting the list. Test the screen against a sample of charts reviewed by a nurse before go-live, and repeat that review after every mapping change.

How Agnotic can help

Agnotic Technologies builds CCM eligibility screening for the client as part of its care management software work: condition mapping tables, worklists ranked by risk, links into consent and initiating visit workflows, and audit trails that show why each patient was flagged.

Agnotic connects the screen to the EHR through FHIR APIs or HL7 v2 feeds, so problem lists, encounters and medications stay current, and designs the review step so a clinician makes the final call on every enrollment.

Sources

• CMS MLN909188: Chronic Care Management Services
• CMS: Frequently Asked Questions about Physician Billing for Chronic Care Management Services