The first decision is what should reconcile
A charge report starts after a charge exists. To find an omission, I need an independent source of completed activity. Depending on the service, that might be a procedure log, a medication administration record, a completed visit, or an interface transaction. Scheduling alone is usually insufficient because a scheduled service may be canceled, changed, or never performed.
I would select one service line, department, or charge family first. With the department and revenue integrity team, I would define the qualifying completed event, the expected charging method, and the legitimate reasons an event should produce no separate charge. Bundled services, canceled work, replacements, and hospital versus professional billing responsibilities need explicit treatment.
The unit of analysis matters. One encounter can contain several billable events, and one service can generate multiple charge lines. A one-to-one match is only appropriate when the actual charging design supports it.
What I would ask the teams to provide
- Completed activity with patient, encounter, department, service date, and a usable event or source identifier.
- Charge transactions, including posting dates, units, corrections, reversals, and the relevant hospital and professional components.
- Charge mappings, source-system or interface processing information, and exceptions that failed before posting.
- Charge review workqueue configuration, routing rules, and the department's reconciliation procedure.
I would confirm how identifiers connect across these sources before calculating a missing-charge rate. A missing match can reflect a bad join, a timing difference, or a legitimate charging exception. It is a lead for investigation, not automatically lost revenue.
How I would trace the charge in Epic
I would follow a small number of expected events through charge generation, the Charge Router where applicable, destination routing, charge review, and the posted transaction. The question at each stage is specific: did the source generate the charge, did the system accept it, did it reach the intended billing destination, and did it post with the expected attributes?
A Charge Router review exception and a charge already sitting in an HB charge review workqueue are different points in the process. Resolving one does not prove the other is clear. I would also check HB and PB independently when the service has both facility and professional components.
For a rule-driven issue, I would request the applicable CER records and the configuration that calls them. The extract needs enough context to interpret the rule and its property parameters. I would pair that build review with actual source and transaction examples. An inventory of rules cannot tell me whether an expected source event ever occurred.
That trace helps distinguish a workflow omission from an interface failure, mapping gap, routing defect, or review backlog. Each finding should identify the point of failure, the evidence, and who can correct it.
The control needs an owner and a resolution path
I would build the reconciliation around exceptions the team can investigate. The output should explain which completed events lack expected charges, which charges remain unresolved, and which exceptions are legitimate. A department owner needs a clear way to resolve each category and document the outcome.
Then I would test the control against difficult cases: a canceled service, duplicate source activity, a reversed charge, corrected units, a late interface message, and a service that produces both HB and PB charges. Those examples show whether the control catches actual gaps without repeatedly flagging correct activity.
The review cadence should follow the service's charging timeline and billing risk. I would agree on the expected interval, escalation point, and evidence of completion rather than impose a daily review on every department. When the underlying defect can be corrected, the ongoing exception volume should fall.
The measures should separate completeness, speed, and value
| Measure | How I would use it |
|---|---|
| Unreconciled event rate | Qualifying completed events without the expected charge outcome divided by qualifying events, after agreed exclusions and processing time. |
| Charge lag | Time from the relevant service event to posting. Review the median and older exceptions, using consistent timestamps. |
| Control accuracy | Confirmed defects as a share of reviewed exceptions, with false positives and repeat exceptions tracked separately. |
| Financial result | Incremental collectible reimbursement and subsequent collections after corrections, with reversals and payment methodology considered. |
Gross charges are useful for sizing an inventory, but they are not a savings calculation. A missing charge may have limited or no incremental payment under a particular reimbursement method. Some late corrections also create rebilling work or may fall outside applicable limits. I would validate financial value with the billing and reimbursement teams before attaching an ROI figure.
The strongest reconciliation process makes it possible to answer three questions: did the organization capture the expected activity, did it reach billing on time, and did the correction improve a measurable financial or operational outcome?