Consumer Mobile Health Application Functional Framework, Release 2
2.0.1 - CI build International flag

Consumer Mobile Health Application Functional Framework, Release 2, published by HL7 International / Mobile Health. This guide is not an authorized publication; it is the continuous build for version 2.0.1 built by the FHIR (HL7® FHIR® Standard) CI Build. This version is based on the current content of https://github.com/HL7/cmhaff-ig/ and changes regularly. See the Directory of published versions

Requirements: APU.5 Data Authenticity, Provenance, and Associated Metadata (Function)

Official URL: http://hl7.org/ehrs/uv/cmhaffr2/Requirements/CMHAFFR2-APU.5 Version: 2.0.1
Standards status: Informative Computable Name: APU_5_Data_Authenticity__Provenance__and_Associated_Metadata

This category is about providing assurance that consumer data is secure when it is moved between the consumer’s device(s) and other locations. This category is about the attribution of sources of data (provenance) and assurance of data authenticity.

Description I:

Related Regulations and Standards: The following are examples of standards for data provenance. Even though they are specific to the FHIR and CDA standards respectively, the principles and data elements -- for recording entities and processes involved in producing or delivering a resource (data) – may be applicable. • FHIR Provenance Resource http://www.hl7.org/FHIR/provenance.html • HL7 CDA® R2 Implementation Guide: Data Provenance, Release 1 - US Realm http://www.hl7.org/implement/standards/product_brief.cfm?product_id=420 • ISO 21089 Health Informatics – Trusted End-to-End Information Flows. https://www.iso.org/standard/35645.html Offers a guide to trusted end-to-end information flow for health(care) records and to the key trace points and audit events in the electronic entity/act record lifecycle (from point of record origination to each ultimate point of record access/use).

Criteria N:
APU.5#01 SHALL

The system SHALL conform to best practices for data authenticity, provenance, and associated metadata.

APU.5#02 SHALL conditional

IF the system creates an AI-derived output that is stored, displayed, exported, or exchanged THEN the system SHALL label the output as AI-derived and retain provenance sufficient to identify the source data references, model or service identifier and version, time of generation, and whether a human reviewed or modified the output.

APU.5#03 SHALL conditional

IF the system itself originates data, as defined in ISO 21089, THEN the system SHALL provide the customer with the option to review, irreversibly destroy, reject, or discard that data.

APU.5#04 SHALL conditional

IF the system only receives data as a pass-through and cannot store data THEN the system SHALL provide the customer with a review option that displays the data prior to executing the pass-through, including the option to irreversibly stop and block the pass-through.

APU.5#05 SHOULD conditional

IF the system receives data and stores it THEN the system SHOULD provide the customer with a review option that permits only appending data or free-text comments to the received data as author, while preserving the original received data intact with its original provenance. The user may comment that data are erroneous but does not have the option to delete the original data.

APU.5#06 SHALL conditional

IF the system displays AI output derived from consumer or device data THEN the system SHALL preserve or reference the original source data without overwriting it and SHALL distinguish AI interpretation, summarization, prediction, or generated content from the original data.

APU.5#07 SHALL conditional

IF the AI function produces a meaningful confidence, uncertainty, or quality signal THEN the system SHOULD preserve and present the signal in a manner appropriate to the intended user and SHALL NOT present a confidence value as a guarantee of correctness.

Notes:

The following are examples of standards for data provenance. Even though they are specific to the FHIR and CDA standards respectively, the principles and data elements – for recording entities and processes involved in producing or delivering a resource (data) – may be applicable.