The complaint should become a traceable example
I would begin with a targeted sample of patient questions, corrected statements, and balances that staff repeatedly need to explain. A caller saying the bill ignores a payment may indicate unapplied cash, a timing difference, a template problem, or a misunderstanding of which account the payment covered. Those causes require different fixes.
I would identify the exact statement date and version rather than compare an old statement with only today's account balance. Payments, insurance responses, transfers, and adjustments can change the account after a statement is generated. The question is whether the amount and explanation were appropriate for the information available at that point and whether subsequent updates reached the patient clearly.
The evidence should connect liability and presentation
My request would include account and guarantor relationships, patient and insurance liability, payments and their application, adjustments, payment plans, statement holds, and statement run history. I would ask for the actual rendered statement, the relevant template configuration, and print-vendor output where applicable. A screenshot of the current billing screen is not enough to establish what the patient received.
I would include HB and PB when the organization uses combined statements or a Single Billing Office. The team should explain which balances appear together, how related accounts are handled, and which settings govern inclusion. Patient financial services, payment posting, and the statement build owner need to validate the examples together because a presentation concern can originate in a different operational workflow.
The Epic review needs eligibility and content
I would first examine statement qualifications and hold behavior in the applicable Enterprise Billing Profile and related settings. I would confirm which accounts and balance types enter the run, how payment plans are considered, and when a pending insurance condition should affect patient communication. Changing the template cannot correct a claim that moved to patient responsibility incorrectly.
I would then review the SmartText and SmartLinks used to display amounts. In particular, the treatment of undistributed patient payments must match the amount-due presentation. I would compare the configured display with actual account examples rather than assume every balance SmartLink handles those credits identically. If a print vendor renders the statement, I would verify the final vendor output as well as the information Epic exports.
The fix should address the originating workflow
A missing patient payment may need application or matching work. A balance transferred prematurely may need payer follow-up or correction of the liability workflow. A statement showing confusing historical activity may need a display change after the underlying balance is validated. I would assign the correction to the owner who can prevent recurrence, with patient services managing the communication.
The organization should agree on what the statement is intended to tell a patient: the amount currently due, which services it relates to, how recent payments are reflected, and how to raise a question. That agreement should inform both the content and the exception process. I would avoid suppressing difficult balances broadly just to reduce calls, because that can hide unresolved account problems and delay appropriate billing.
Testing should use realistic account combinations
I would test a straightforward self-pay balance, pending insurance, a recently posted patient payment, an undistributed payment, an active payment plan, and a credit after a prior statement. The set should include the organization's actual HB and PB combinations and any affected guarantor types. Each expected result needs a validated amount and a readable explanation.
A staff member who answers billing questions should review the rendered output. If the team changes statement qualification or cycle settings, I would confirm that accounts still receive appropriate communication and do not remain held unintentionally. Where MyChart and mailed statements present related information, I would compare those views. Consistent, accurate information matters more than a design that looks clean in only one channel.
I would retain the generated test statement with the transaction snapshot used to validate it. That pairing lets reviewers distinguish a display defect from account activity that posted after the statement was produced.
Measurement should distinguish accuracy from volume
I would measure confirmed statement defects per reviewed or generated statement population, with the denominator clearly identified. Complaints are useful leads, but not every complaint proves an error and not every error generates a call. A small targeted audit should therefore supplement service logs and complaint categories.
I would track corrected statements, repeat contacts, payment-application delays, inappropriate balance transfers, and accounts held beyond their next intended communication point. Payment timing can be monitored alongside those measures, but a reduction in calls alone does not prove that patients received accurate information. The strongest result connects fewer validated defects with clearer service resolution and a reconciled patient balance. That gives the organization a better explanation for both its operational improvement and its collections trend.