Detection Lifecycle

Clinical Alerts evaluates qualifying network document exchange activity for enrolled patients and generates notifications when clinically meaningful activity is inferred.

Alert processing follows a defined lifecycle:

Enrollment eligibility
Network query detection
Document retrieval
Encounter inference
Alert generation
Delivery

Each stage must complete successfully before a notification is delivered.

Detection Model

Alert generation requires all of the following conditions:

The patient is actively enrolled for Clinical Alerts at the time processing occurs.
A qualifying network query is detected for that patient.
One or more clinical documents are successfully retrieved.
Retrieved documents contain sufficient clinical context to support inference of clinically meaningful activity.

Detection is independent of delivery method and payload format. Delivery configuration does not affect alert eligibility or detection timing.

Observation Window

Alert evaluation occurs within a defined observation window:

The window begins when a qualifying network query is detected.
The window remains active for up to 10 days.
If qualifying document retrieval and analysis occur within that window, an alert may be generated.
If no qualifying documents are retrieved during the window, no alert is generated.

Observation windows are evaluated per patient and per qualifying query event. Overlapping windows may occur if multiple qualifying queries are detected for the same patient.

Encounter Inference

Clinical Alerts derives encounter-level context from retrieved clinical documents. It does not receive explicit encounter events from source clinical systems.

Based on document content, Clinical Alerts may infer:

Encounter class (for example, inpatient, emergency, outpatient, observation)
Facility or organization context
Encounter timing information, when available
Patient association

Inferred encounter context reflects document-derived activity and should not be treated as an authoritative or complete source-system encounter record.

Alert Generation

An alert is generated only when:

Enrollment requirements are satisfied.
Qualifying document retrieval occurs.
Document analysis completes successfully within the active observation window.
Retrieved content supports inference of clinically meaningful activity.

Alert generation is prospective. Alerts are not generated retroactively for activity that occurred before enrollment, before Clinical Alerts was activated for your organization, or outside an active observation window.

Alerts indicate that qualifying activity occurred. They do not include full clinical documents.

Delivery

Once generated, alerts are delivered using the configured delivery method:

FHIR Subscription (REST-hook)
Secure File Transfer Protocol (SFTP)

Delivery configuration determines transport and payload format only. It does not affect detection logic, eligibility, or timing of alert generation.

Timing Characteristics

Clinical Alerts is event driven and asynchronous.

Alert timing reflects:

When document exchange occurs on the network
When documents become available for retrieval
Completion of document analysis and inference processing

Alerts may be delivered hours or, in some cases, days after the underlying care event. Ordering relative to other clinical events is not guaranteed, and notification timing does not necessarily correspond to the exact time care occurred.

Limitations and Constraints

Clinical Alerts depends on network participation and document availability.

If participating organizations do not exchange documents for an encounter, no alert can be generated.
Administrative, financial, scheduling, or eligibility documents do not generate alerts.
Some unstructured or incomplete documents may not contain sufficient information to support encounter inference.
Alerts are not exhaustive and should not be treated as a complete encounter history.

Clinical Alerts complements, but does not replace, HL7 v2 ADT messages.