Encounter data export completeness depends on what a product stores, what an export operation is scoped to include, and how the output can be reconciled against the source record. The ONC criterion for electronic health information export does not prescribe a fixed encounter dataset or transport standard. Instead, it requires that a certified module enable timely creation of an export file containing all of a single patient's electronic health information that can be stored at the time of certification [1]. Vendor documentation then defines the practical meaning of completeness through dataset lists, file formats, and reconciliation instructions. A complete export is therefore a documented boundary, not an absolute copy of every possible clinical artifact. Organizations evaluating completeness should compare the certification scope, the product-specific data dictionary, the export level, and the supporting identifiers that allow the output to be matched back to source records.

The certification boundary for export completeness

The ONC criterion establishes a temporal and storage boundary for what must be exported. The export must include all of a single patient's electronic health information that can be stored at the time of certification by the product of which the health IT module is a part [1]. This phrasing limits the scope to data held in the module in its original form as presented for certification. It does not include data subsequently generated from unique post-certification deployments [1]. For an encounter export, that means completeness is measured against the certified product's stored encounter model, not against every custom field or downstream calculation that a particular customer may have added after deployment.

The criterion also leaves transport methods and data standards unspecified [1]. It requires that the export be electronic and in a computable format, and that a publicly accessible hyperlink to the format documentation be included with the exported files [1]. This creates a situation where two certified products can both satisfy the criterion while producing materially different encounter outputs. One may export encounter records as FHIR resources, another as proprietary JSON files tied to internal database tables. Neither approach is inherently more complete under the regulation. The completeness question shifts to whether the vendor's published format documentation allows a recipient to understand which encounter fields are present and how they relate to the source system.

How vendor data dictionaries define encounter scope

athenahealth's ambulatory clinical EHI export documentation lists encounters as one dataset within a broader export that can include allergies, lab results, procedures, and other clinical data [2]. The documentation states that each JSON file contains the relevant data if present for a patient or multiple patients depending on the export use case [2]. Patient identifiers such as first name, last name, date of birth, Social Security number, Athena patient ID, enterprise ID, phone numbers, and address are added at the beginning of each line to support patient matching and reconciliation across datasets [2]. This design treats encounter completeness as dependent on the presence of data in the source record, not on a guaranteed population of every field.

Veradigm's EHI Export Data Guide takes a different approach by documenting output files against internal database tables. The encounter file is described as corresponding to the ENCOUNTER table with EncounterId as the primary key [6]. Field definitions include EncounterId, ProviderId, ProviderName, PatientID, PatientName, VisitTypeID, and other encounter-level attributes [6]. Related files such as EncounterPlan.json provide details about plans or orders created from individual encounter assessments [6]. This table-oriented documentation makes the completeness boundary explicit: a recipient can see which source table each file represents and which fields are included. The encounter export is complete relative to the documented schema, but the documentation does not claim that every possible encounter-related concept is present.

A suggested validation scenario for encounter export completeness

A hypothetical validation could select a known encounter with a documented set of associated records, then compare the export output against the source system. The team would first identify the encounter ID in the source EHR and note the expected related artifacts: the encounter record itself, any encounter documents, lab results ordered during the encounter, and procedure results. After running the export, the team would confirm that the encounter file contains the expected EncounterId and that patient identifiers match the source record [6]. For document-based artifacts, the team would check that the corresponding document folder exists and contains the expected files [2].

The validation would then test the population boundary by selecting a group of patients and confirming that encounters referenced by exported resources appear in the output [4]. For a database-oriented export, the team would run the vendor-supplied mapping queries to connect multimedia files to their corresponding data entries [3]. Any encounter present in the source but absent from the export would be investigated against the documented inclusion rules. This scenario is a suggestion for internal quality assurance, not a claim that any vendor has validated this specific workflow. It provides a structured way to turn the documentation's completeness boundaries into testable checks.

Export levels and the population boundary

Elation's FHIR bulk data export documentation describes three export levels: system level, all patients, and group of patients [4]. The system level exports data from the FHIR server whether or not it is associated with a patient, while the all patients operation obtains resources pertaining to all patients, and the group operation targets members of a specified group [4]. The documentation notes that for patient level and group level exports, patients are always exported regardless of search parameters if they are referenced by resources present in the response [4]. This means encounter completeness at the population level is influenced by the chosen endpoint and by which resources reference the encounter.

The Elation documentation also states that the server provides Encounter, Location, Organization, and Practitioner resources as they are referenced as must support elements in required resources [4]. This is a narrower statement than a claim that every encounter in the system will appear. The export contains encounters that are referenced by the resources included in the response. A recipient evaluating completeness must understand this reference-based inclusion rule. If an encounter exists but is not referenced by an exported resource, the documentation does not state that it will be included. The population boundary is therefore defined by the export level and the reference relationships within the FHIR data.

Reconciliation artifacts and source-to-export matching

Oracle's patient population EHI export documentation describes a process that extracts all data for all patients in the target Millennium or Oracle Health EHR system and transforms it where necessary for consumption [3]. The export includes core Millennium EHR data stored in relational database tables, multimedia storage data, document imaging data, longitudinal plan data, and Oracle Health EHR native data [3]. For single tenant systems, a copy of the full Millennium EHR database is delivered via a customer-supplied secure encrypted device [3]. Multimedia files are downloaded in their native format, and mapping of individual multimedia files to corresponding data entries can be accomplished using queries supplied as part of the extract [3].

The MHS GENESIS encounter specification illustrates how a downstream repository reconciles exported encounter data. The MDR GENESIS Encounter table is based on data received from Oracle Bulk Data Extract 3.0 feeds, with the source system identified as the MHS GENESIS Oracle Millennium database [7]. Raw source views include encounter, encounter alias, encounter personnel relationship, schedule appointment, and schedule event actions [7]. The specification describes transformation rules for fields such as the encounter key, derived as a character version of encntr_id [7]. This demonstrates that completeness in a real deployment is verified through a documented transformation pipeline, not by assuming that the raw export alone is sufficient.

What remains unresolved in encounter export completeness

The ONC criterion explicitly states that it does not specify a predefined data set and that the amount of EHI will differ by developer and product [1]. This means there is no single authoritative list of encounter fields that must appear in every certified export. A buyer cannot compare two products by checking for the same field names. Instead, the comparison must be made against each product's own data dictionary and the certification scope. The criterion also does not specify transport methods or data standards [1], so the same encounter data may be delivered as FHIR NDJSON, proprietary JSON, or database extracts depending on the vendor.

Another unresolved area is the treatment of encounter-related documents and multimedia. athenahealth's documentation indicates that for each JSON dataset containing supporting files or documents in zip folders, a corresponding document folder in the same name will be present [2]. Oracle's documentation separates multimedia files from the relational database extract and requires separate download via a secure URL [3]. These different handling approaches mean that encounter completeness may require combining multiple artifacts. The documentation does not establish a common method for verifying that all referenced documents were successfully transferred. Organizations must design their own reconciliation checks against the vendor's published format specifications.