On this page
- What is an expected event in healthcare operations?
- What counts as an actual event?
- How do triggers, expected events and evidence line up?
- How does reconciliation work?
- What counts as an exception?
- How should confidence be expressed?
- How is completion verified?
- How does reconciliation show up in daily work?
Key takeaways
- An expected event is care the record implies should happen; an actual event is evidence that it did.
- Appointments show intent, while encounters and procedures show completion, so they carry different weight as evidence.
- Reconciliation matches each expected event to actual events by patient, care type, location and time window.
- Unmatched or ambiguous items become exceptions with a confidence level, an owner and a recorded outcome.
An expected event is care that the record implies should happen: a consult after a referral, a procedure after an order, a follow-up after a recall date. An actual event is evidence that it did happen: an appointment, an encounter or a documented procedure. Comparing the two, patient by patient, is the most reliable way to find care that stalled without anyone noticing.
Comparing expected vs actual events in healthcare operations is simple to state but detailed to execute. This guide covers how both are defined, how they are reconciled, and how exceptions, confidence and verification should work. It is part of our EHR and data intelligence series, and it is the core mechanism behind healthcare revenue leakage intelligence.
What is an expected event in healthcare operations?
An expected event is a specific piece of care that should follow something already documented in the record, within a defined window: a consult after an accepted referral, a procedure after a signed order, a return visit after a recall date. If no matching appointment, encounter or procedure appears, the care may have stalled.
The trigger is something a person documented. The expected event is what that documentation implies. The window is set by the practice, and by clinicians wherever it touches clinical timing.
Three triggers account for most expected events in specialty practices.
Expected visit after a referral
When an inbound referral is accepted, the practice expects to contact the patient and then see them for a consult. Many practices set two checkpoints, a contact window and a visit window, with lengths they choose. Where the work breaks down is covered in why referrals don't get scheduled.
Expected procedure after an order
When a provider signs a procedure or diagnostic order, the practice expects that procedure to be scheduled and performed. The window depends on the type of order and the ordering provider's instructions. Orders that stay open past their window are what tracking unresolved medical orders is about.
Expected recall or follow-up
When a provider sets a recall or a checkout follow-up, the practice expects a return visit or surveillance procedure around the due date. Intervals depend on the patient's findings and the practice's approved guidance, so the window should come from what the provider documented, not from a universal rule.
What counts as an actual event?
An actual event is a record that shows the expected care moved forward or was completed. Not all actual events carry the same weight.
- Actual appointment: a booked appointment shows intent. It is progress, and it should remove the patient from a call list, but it is not completion. Appointments are canceled and missed.
- Actual encounter: a completed visit shows the patient was seen. For a referral or a follow-up visit, it is usually the completion evidence.
- Actual procedure: a documented procedure shows the ordered or recalled care was performed. For orders and surveillance recalls, it is usually the completion evidence.
Treating a booked appointment as completion is one of the most common measurement mistakes. It makes conversion look better than it is and hides canceled care.
How do triggers, expected events and evidence line up?
| Trigger | Expected event | Where the window comes from | Evidence of the actual event |
|---|---|---|---|
| Inbound referral accepted | Contact, then a consult visit | Referral checkpoints the practice sets | Contact logged, then a completed encounter |
| Procedure order signed | Procedure scheduled and performed | Order type and provider instructions | Appointment booked, then a documented procedure |
| Diagnostic order signed | Test performed and resulted | Order type and provider instructions | A procedure or result record, or an outside report |
| Recall due date set | Return visit or surveillance procedure | The interval the provider documented | Encounter or procedure on or after the due date |
| Checkout follow-up requested | Return visit booked and completed | Checkout instructions | Appointment booked, then a completed encounter |
| Procedure canceled | Procedure rebooked | Rebooking window the practice sets | New appointment, then a documented procedure |
How does reconciliation work?
Reconciliation is the process of matching each expected event to the actual events that could satisfy it. It runs on a schedule, so items open and close as the record changes.
- Trigger recordedA referral, order, recall or checkout follow-up is documented in the EHR.
- Expected eventThe rule defines what should follow and within which window.
- Search evidenceAppointments, encounters and procedures are checked for the same patient and care.
- ReconcileMatched items close; booked items wait; unmatched items open.
- Score confidenceEach open item gets a confidence level based on evidence found.
- Verify completionThe item closes when an encounter, procedure or recorded reason appears.
Reconciliation runs on a schedule, so items open and close as the record changes.
A match usually requires four things to line up: the same patient, the right kind of care, a plausible location, and a date inside the window after the trigger. The "right kind of care" is the step that takes the most setup. Local appointment types and order codes rarely match each other directly, so they need to be mapped once to shared categories and kept current as new codes appear.
The outcome of reconciliation for each expected event falls into one of four states:
- Completed: completion evidence was found. The item closes.
- In progress: an appointment is booked inside the window. The item waits and is checked again.
- Open: no matching actual event and the window has passed. The item becomes an opportunity.
- Exception: the evidence is unclear, conflicting or outside the data. The item goes to review.
What counts as an exception?
An exception is any expected event that cannot be confidently placed in one of the other states. Common examples include:
- Two possible matches inside the window, such as two procedures of a similar type.
- A procedure documented under a local code that has not been mapped yet.
- Care completed at an outside facility with no report in the chart.
- A patient who declined, transferred care or moved away.
- A canceled order that was replaced by a different order for related care.
Exceptions should not be dumped into the same list as clear opportunities. They need a reviewer, and each resolution should be recorded with a reason, such as "completed elsewhere," so the same item does not reappear on the next run. Each recorded reason also improves the rules over time.
How should confidence be expressed?
Confidence should tell staff how much to trust an item before they act on it. A practical approach uses a few plain levels rather than a precise-looking score:
- High: a clear trigger, a passed window, and no matching appointment, encounter or procedure anywhere in the available data.
- Medium: a clear trigger, but some evidence is missing or partial, such as an outside referral with no report.
- Low: the trigger itself is ambiguous, or the mapping for the relevant codes is incomplete.
High-confidence items can go straight to a prioritized work queue. Medium and low items often go to a reviewer first. Confidence also belongs in prioritization, alongside estimated value and urgency, so staff spend time where the evidence is strongest.
How is completion verified?
Completion is verified only when evidence of the actual event appears, or when staff close the item with a recorded reason. Phone calls, messages and booked appointments are progress. The loop closes when the encounter or procedure is documented.
Where claims or health information exchange data is connected, completion matching can also recognize care performed elsewhere more precisely. Without those sources, a staff disposition with a reason is the honest fallback. Either way, clinicians make clinical decisions; the reconciliation only reports what the data shows.
How does reconciliation show up in daily work?
In daily work, the result is a short list of patients with open items, each showing the trigger, what is missing, the evidence, a suggested action and an estimated value. Schedulers, referral coordinators and MAs work that list; supervisors review exceptions; leaders see how many items were found and how many closed.
A practice can start small: one trigger, such as signed procedure orders, one service line and a spreadsheet refreshed weekly. The limits show up quickly, mainly in code mapping and in keeping the schedule check current, and that is usually when teams look at software. Nuvory runs this reconciliation nightly with rules a clinician at the practice signs off, and ranks open items through patient opportunity intelligence, illustrated in these worked patient examples; how Nuvory works shows the sequence. Why built-in reports struggle with the same job is covered in EHR analytics vs. revenue leakage intelligence.
Expected vs. actual events checklist
- List the triggers your practice wants to track: referrals, orders, recalls and checkout follow-ups.
- For each trigger, define the expected event and where its window comes from.
- Decide which actual events count as progress and which count as completion.
- Map local appointment types and order codes to the care they represent.
- Agree how care completed elsewhere is recorded, with a reason.
- Set confidence levels and decide which items go to a reviewer instead of a call list.
- Have a clinician sign off any rule that depends on a clinical interval.
Frequently asked questions
Why not treat a booked appointment as completion?
Because appointments are canceled, rescheduled and missed. A booked appointment should remove a patient from a call list, but the expected event stays open until an encounter or procedure is documented. Otherwise canceled care looks complete and leakage is undercounted.
Who decides the window for each expected event?
The practice does. Referral and rebooking windows are operational choices the practice approves. Recall and follow-up windows come from what the provider documented, because intervals depend on the patient's findings and the practice's approved guidance. Rules touching clinical intervals should be signed off by a clinician.
What happens when care was completed at another facility?
If claims or HIE data is connected, matching can often recognize it. If not, staff close the item with a recorded reason, such as completed elsewhere, and attach any outside report. The reason keeps the item from reopening and shows up in reporting.
Is this the same as a recall list?
No. A recall list shows patients with a due date. Expected-versus-actual reconciliation checks each due date against appointments, encounters and procedures, removes patients already booked, flags exceptions and verifies completion. The recall list is one input, not the result.