The backlog needs a precise definition
I would first agree on the population. DNB, or do-not-bill, holds, Stop Bills, charge review exceptions, and claim edits can all delay billing, but they operate at different points. A total balance labeled unbilled does not explain which barrier is active or who can resolve it.
One account can have several holds and appear in more than one workqueue. Summing workqueue balances can therefore overstate the financial inventory. I would retain the individual hold detail while reconciling the executive view to unique accounts and the organization's agreed unbilled definition.
Age also needs a defined starting point. Days since discharge, days since an edit first appeared, and days in the current workqueue measure different things. All can be useful, provided the report names the clock it is using.
I would request both configuration and account history
For an Epic review, I would ask for the active DNB and Stop Bill inventory, workqueue configuration, owning areas, supervisors, routing criteria, and representative account histories. I would also request the relevant rule records, or CER extract, where rules control the check or routing. Some checks use other configuration or extension logic, so a CER extract is only part of the review.
For each linked rule, I would want its context, criteria, referenced properties and parameters, and where the rule is used. I would compare that intended behavior with accounts that entered, remained in, or unexpectedly bypassed the queue. The relationship between the check and the queue matters as much as either item by itself.
The account history should show when the barrier appeared, what users changed, whether another team received the account, and what happened when the account was reevaluated. Without that sequence, it is easy to mistake a symptom for the cause.
The review should explain why work stalls
- The check catches a legitimate issue, but the responsible team lacks the information, access, or training to resolve it.
- The routing criteria send an account to a team that cannot take the next action.
- A broad rule or outdated parameter catches accounts outside the intended population.
- Several teams review the same issue, with no agreed sequence or return handoff.
- The underlying data is corrected, but processing timing or reevaluation behavior delays the account's release.
These findings lead to different actions. A staffing problem needs a capacity response. A routing defect needs a configuration change. A missing documentation handoff needs an operational agreement. Adding another workqueue can make all three harder to see if it simply duplicates the existing inventory.
A smaller queue is only useful if the control still works
I would validate a proposed change with the operational and application owners before implementing it. That includes confirming which accounts should remain held, which should pass, which team should receive each exception, and what event should clear the barrier.
My acceptance scenarios would include an account that legitimately fails the check, an account that should pass, an account with multiple concurrent holds, and a corrected account waiting for reevaluation. I would also test the handoff between teams and verify that legitimate billing controls remain intact.
The goal is to reduce avoidable holds and repeat work while preserving the checks that protect claim accuracy. Deactivating a control or transferring accounts out of a queue can improve a dashboard without improving billing readiness. I would follow the accounts through to the next meaningful outcome.
How I would measure a durable improvement
| Measure | How I would use it |
|---|---|
| New hold inflow | Accounts newly hitting the targeted barrier, normalized to the relevant discharge or billing-ready population. |
| Hold duration | Time from first occurrence to resolution, with median and older cases shown separately. |
| Repeat work | Accounts returning with the same issue, duplicate reviews, transfers, and touches required to resolve it. |
| Billing progression | Time to claim submission after the targeted hold clears, including other remaining barriers. |
| Financial impact | Expected collectible balances and subsequent cash timing, with gross charges and overlapping queue balances identified. |
I would report new work and resolved work alongside ending inventory. A one-time cleanup can reduce the balance while the same volume of avoidable holds continues to arrive. A lasting improvement changes that pattern and gives the owner a way to detect recurrence.
Faster billing can improve working capital without creating an equal amount of recurring margin. Staff time released is another useful outcome, but it becomes an expense reduction only when spending actually changes. The measure should reflect the benefit the organization achieved, with enough follow-up to show that the backlog is not rebuilding.