Sift Healthcare
Abstract data visual for an article on finding Medicare Advantage underpayments in hospital remittance data

Where Medicare Advantage Underpayments Hide in Your Remits

A Medicare Advantage claim can be approved, adjudicated and paid, and still come in thousands of dollars short with no “denial.” That’s the pattern behind most Medicare Advantage underpayments we’re seeing right now. Aetna’s level-of-severity policy is the clearest example. Since January 1, Aetna has reviewed urgent and emergent inpatient stays of one to four midnights for its Medicare Advantage and dual-eligible members against MCG criteria. Stays that miss criteria keep their inpatient status but get paid at a rate Aetna calls comparable to observation. The admission is approved, but the DRG doesn’t pay.

Hospitals pushed back early. The AHA asked Aetna to rescind the policy before it took effect, and a federal judge in Pennsylvania kept Jefferson Health’s challenge in open court rather than sending it to arbitration. But the case could take a long time. The policy is still in effect, and these claims are in your remit now. Aetna isn’t the only problem. One large revenue cycle firm told Thomson Reuters that MA plans pay roughly 90% of what contracts at 100% of Medicare imply. Plans that ignore the two-midnight rule get to a similar result through status changes. Aetna’s version happens to be the easiest to find, so it’s a good place to start.

Identifying Medicare Advantage Underpayments

These cases carry one code pairing, RARC N610 with CARC 45. N610 tells you the payer-based payment based on what it considered the appropriate level of care. CARC 45 says the charge exceeded the fee schedule or contracted amount, and it’s the same code your team sees on thousands of routine contractual adjustments every month.

The second code is where these cases disappear. Most revenue cycle systems treat CARC 45 as a contractual write-off, so the case never creates a denial record, never lands in a denial workqueue and never touches the denial rate leadership reviews. Expected reimbursement often still reflects the inpatient DRG, leaving an unexplained variance on the account.

A variance report won’t catch it on its own

Most advice on how to detect payer underpayments assumes the variance surfaces. Model the contract, compare expected to paid, work the gap. That holds when a payer pays the wrong amount on the right basis. Here the payer pays on a different basis and labels the difference contractual.

Plenty of teams suppress CARC 45 variances to keep their queues workable. If yours does, the contract model never sees these accounts. Finding them starts with the remark code, not the dollar amount.

The gap is pretty big

Across the low-severity Aetna Medicare cases Sift has analyzed, the gap between expected and actual reimbursement has run about $6,000 per case. On the systems we reviewed, those cases sat with routine payer adjustments, and most never generated a denial record. On one account, $6,000 is an annoyance. Repeated across every short Aetna Medicare stay that misses MCG, it adds up to a figure your denial report has never shown your CFO, because the report only counts what posts as a denial.

How to Find Medicare Advantage Underpayments in Your Data

  1. Pull the pairing. Take 12 months of Aetna Medicare inpatient remits and filter for N610 with CARC 45. Sum expected minus paid, then count how many of those cases ever created a denial record. The ones with a gap and no denial record are your working list for appeals and payer escalation.
  2. Make it visible going forward. Configure N610 plus CARC 45 to create a denial record or its own variance flag, so these stop clearing with contractual write-offs.
  3. Reconcile status. When a claim pays at an observation-level rate, and the account stays inpatient, expected reimbursement is overstated, and so is net collection rate. Decide how these accounts should sit before quarter close.
  4. Match remits to UR correspondence. UR or case management often holds Aetna’s notification before finance ever sees the remit. Peer-to-peer requests and appeals can get some of these paid at the expected rate, and the letter’s date matters for both.
  5. If you’re an academic medical center, settle the IME question. How IME dollars are handled when a claim pays at an outpatient-comparable rate but the patient stays inpatient belongs with billing compliance and legal, not a workqueue default.
  6. Run the same cut on your other MA plans. Segment by plan, admission source and length of stay. If another plan starts paying short stays below DRG with a level-of-care remark, this data will show it before a policy bulletin does.

The pairing is narrow, which makes it a useful test. If your team can’t say how many N610 cases Aetna paid this year, the same blind spot is probably holding other losses that post as paid. We cover four more in the 2026 Denial Deep Dives guide, a helpful resource for thinking through where health systems are losing revenue outside of denials.

Picture of Bethany Grabher

Bethany Grabher

Bethany Grabher leads HR and Communications at Sift Healthcare, where she turns complexity into clarity and ideas into action. A lifelong ideator, she explores how culture, strategy, and technology shape the way healthcare organizations grow and lead.

Post-Bill Recovery Details

Pre-Bill Denial Prevention

Concurrent Denial Prevention

Harnessing the power of AI

Sample Insights Report

ML-Driven Denials Intelligence

Driving Denial Reduction And Revenue Recovery Improvement

Recovery Interest

RevProtect Interest

Request an Insights Analysis

Schedule A Demo