DRG validation and the single-MCC claim: when one code moves the payment

Billed DRG and regrouped DRG

A hospital claim arrives with diagnosis and procedure codes and, usually, the MS-DRG the hospital expects. Validating it is two different jobs.

• Coding validation. Run the codes, present-on-admission flags, age, sex and discharge status through the grouper version for the discharge date. If the result differs from the billed DRG, common causes are a version mismatch, a code that did not make it onto the claim, or a data-entry error.
• Clinical validation. Check that the medical record supports the conditions behind the codes. A code can be valid, correctly sequenced and correctly grouped, and still not be supported by the documentation.

How one code moves the DRG

Many base DRGs split into levels: with a major complication or comorbidity (MCC), with a complication or comorbidity (CC), or without either. The level comes from the secondary diagnoses, and heart failure shows how much can ride on one of them.

Take a stay with acute on chronic systolic heart failure (I50.23) as the principal diagnosis and two secondary diagnoses that carry no severity, type 2 diabetes without complications (E11.9) and hyperlipidemia (E78.5). Here is how MS-DRG version 44 groups the same stay with E43, with E44.0, and with no malnutrition code:

One secondary diagnosis is the difference between $9,681.89 and $19,069.87 for the same four-day stay.

Check the coding before the malnutrition code. If the patient also has hypertension, ICD-10-CM presumes it is linked to the heart failure: the stay is coded I11.0 (hypertensive heart disease with heart failure) first, with I50.23 as a secondary diagnosis. I50.23 then counts as an MCC on its own, and the engine groups the stay to MS-DRG 291 with or without a malnutrition code.

The single-MCC claim

A single-MCC claim is one whose DRG level rests on exactly one MCC. If that code is not supported, the claim drops a level or two, with nothing else on the claim to hold it up. That makes it the natural first place for a payment integrity review to look, and federal auditors have tested exactly this profile.

In July 2020 the HHS Office of Inspector General reported on claims coded with severe malnutrition (E43 or E41, both MCCs) where removing that one code would have changed the DRG. Those claims, for fiscal year 2016 and 2017 discharges, totaled $3.4 billion in Medicare payments across 224,175 claims. In a sample of 200, hospitals billed 173 incorrectly, and the inspector general estimated about $1 billion in overpayments. Other audits have tested whether patients received the more than 96 consecutive hours of mechanical ventilation that MS-DRGs 207 and 870 require.

None of this means a single MCC is wrong. It means a single MCC carries the whole payment difference, so it deserves the review.

When an MCC does not count

A secondary diagnosis on the MCC list does not always raise the level. Two rules can take it away:

• CC exclusions. A secondary diagnosis closely related to the principal diagnosis does not count. Sepsis is the textbook case: with unspecified sepsis (A41.9) as the principal diagnosis, severe sepsis without septic shock (R65.20) is excluded, and the stay groups to MS-DRG 872, not 871.
• Present on admission. For the hospital-acquired conditions Medicare tracks, a CC or MCC that was not present on admission does not count toward the DRG.

A validation tool has to apply both, for the right grouper version, before it calls a claim a mismatch. This site’s regrouping applies CC exclusions. When the code that sets the severity is marked not present on admission, it flags the claim for review against the hospital-acquired condition list rather than applying that list itself.

A validation routine

1. Regroup every claim under the version for its discharge date and compare the result with the billed DRG.
2. Remove each CC and MCC in turn and regroup, to find the claims whose level rests on one code.
3. Price both the billed DRG and the lower one, so the review queue is ordered by dollars at stake rather than claim count.
4. Send those records for clinical review with the specific code and the criteria used; findings must come from the documentation.
5. Track findings by code and by hospital, and let them decide which claims you review next.

The same routine applies to Medicaid claims paid by APR-DRG. A payer without the licensed APR-DRG grouper relies on the severity level the hospital reports, and one severity level can double an APR-DRG payment.

Clinical validation asks whether the documentation supports a diagnosis. It should never start from the assumption that a code is wrong because it raises the payment.