The report needs to identify which amounts differ
I would first ask whether the report compares billed charges, expected allowed amounts, payer-reported allowed amounts, or actual payments. A lower payment can reflect patient responsibility, another payer's liability, a denial, a recoupment, or a contractual variance. Those scenarios should not enter one recovery calculation without review.
The unit of analysis also matters. A contract may price a whole claim, a day, a service line, or a combination of methods. I would agree on the correct level with managed care and reimbursement specialists, then examine the complete claim and payment history. Comparing a partial payment to a whole-claim expectation can create a false underpayment. Summing several snapshots of the same unresolved difference can exaggerate the opportunity.
The evidence should connect terms, build, and transactions
My initial request would include the applicable contract terms and effective dates, amendments, payer and plan mapping, the claim as billed, expected reimbursement details, and the remittance. I would include subsequent payments, reversals, adjustments, replacement claims, and correspondence for the sample. The data needs to show the position after the relevant payment cycle rather than an arbitrary interim balance.
I would request examples of correct payments as well as suspected underpayments. Correct cases help establish whether a contract line, fee schedule, modifier treatment, or claim attribute explains the difference. The operational team should also identify the evidence normally accepted in a payer escalation. A technically calculated expectation still needs a contractual explanation a reviewer can follow.
I would retain the effective contract version with the sample calculation so a later amendment does not silently change the basis of the original finding.
The Epic review should explain the calculation
In Epic Contract Maintenance, I would review the contract selected for the account, its effective content, component selection, pricing methods, fee schedules, and relevant extensions. Ordering can matter when contract lines are evaluated. I would ask the contract builders to demonstrate how the sample reaches the expected amount and which inputs determine that result.
I would then inspect variance thresholds and actions. The organization needs to understand what difference triggers follow-up, whether a contractual adjustment is posted or suppressed, and which payer retains responsibility during investigation. These choices affect both the working balance and the reporting view. I would not assume the reported underpayment represents a comparison of actual cash amounts, because an allowed-amount variance can be a different calculation.
The investigation should separate payer and internal defects
For each sample, I would assign a supported disposition: a likely payer calculation issue, incorrect contract build, inaccurate claim data, remittance mapping or posting error, a clinical or coding dispute, or an unresolved question. I would validate that classification with the appropriate owners. It should not become a contest over which team keeps the balance.
A contract build defect should return to a governed correction process, with a defined approach to reevaluation. A payer issue needs the relevant terms and calculation in its escalation. A DRG downgrade based on clinical validation requires a clinical investigation rather than only a request to pay the billed DRG. Once those categories are separated, the organization can estimate a recovery opportunity without treating every system discrepancy as payer debt.
Testing should include the boundaries of the payment method
I would use a controlled test set covering the contract's main reimbursement methods, effective-date changes, relevant modifiers or units, and any claim-level limits. I would include a corrected claim, a supplemental payment, and a case that should produce no variance. A specialist should confirm both the contractual interpretation and the expected transaction behavior.
If the build change affects previously priced accounts, the team needs an approved reevaluation population and reconciliation of resulting adjustments. I would compare the before and after calculation on the same examples, verify the follow-up destination, and inspect whether account or bucket statuses change. A successful calculation test does not automatically prove that the operational follow-up and financial reporting remain correct.
The benefit should be measured after validation
I would report validated collectible differences separately from the gross variance inventory. The validation yield is the share of reviewed cases with a supported recoverable discrepancy. I would also show false positives by cause, because repeated internal pricing defects can consume more staff effort than a handful of large payer errors.
For recovery, I would follow a defined cohort through actual additional payments, offsets, final dispositions, and the time required to obtain them. The denominator and observation window should be visible. Funds already received through a normal payment cycle should not be attributed to the project without evidence. Preventing an incorrect contractual adjustment and collecting an old underpayment can both be valuable, but they are different outcomes. A defensible result explains what was corrected, how collectibility was established, and which dollars actually arrived.