This page is part of the Application Data Exchange Assessment Framework and Functional Requirements for Mobile Health (v0.1.0: STU 1 Ballot 1) based on FHIR R4. . For a full list of available versions, see the Directory of published versions

Attribution

Features in the Attribution category ensure that metadata about reported measurements are included to identify the device creating a measurement, the author creating manually entered data, and that devices are not permanently linked to a given user and can be readily replaced at need.

This metadata is typically considered to be Provenance information (related to the Provenance FHIR Resource of the same name). Provenance information helps establish authenticity of the recorded information, can be used to identify issues related to a specific device or type of device.

When this metadata is not available, operational procedures can be used to obtain some measure of assurance about the source data, but do not provide nearly as strong assurance as when it is.

Many mobile devices are small, readily lost or broken. Anyone whose run their fitness tracker through the laundry can attest to the fact they may need to be replaced. The ability to connect the measurement to the device allows receivers of measurement data to rule in or out a change in device as a reason for change in resulting measurements. Apps, devices and infrastructure should enable such capabilities.

Some devices easily accommodate use by multiple members of the same household to use them (e.g., weight scales and blood pressure cuffs). Additional requirements in this category ensure that devices and apps which work with them can support this sort of use. Even devices considered to be personal use (a smart watch or fitness tracker) might be shared as one person “tries it out” for a bit to see if it’s something they want to acquire for themselves. Devices can also be handed down or off to others as users upgrade and acquire new technology.

Feature: Measurement Device Can be Identified

The device that performs the measurement **SHALL** be uniquely identified and that identifier
**SHALL** persist throughout that devices lifetime.

Each type of device has attributes with regard to precision, accuracy and quality which can impact the interpretation of measurements taken by the device. Knowledge of the device helps greatly in interpreting the results.

Scenario: Device is Identified

GIVEN

<Device> performs a measurement

AND

<Device> is uniquely identified with <Identifier>

WHEN

<Measurement> has been obtained

THEN

<Identifier> can be found from <Measurement>

Scenario: Device Identifier Persists

GIVEN

<Device> that can perform a measurement

AND

<User1>

AND

<User2>

AND

an <App1> used by <User1>

AND

an <App2> used by <User2> (can be the same app as for User1)

WHEN

a <Measurement1> is taken for <User1> by <App1>

AND

a <Measurement2> is taken for <User2> by <App2>

THEN

there is an <Identifier1> in the <Device> record associated with <Measurement1>

AND

there is an <Identifier2> in the <Device> record associated with <Measurement2>

AND

<Identifier2> = <Identifier1>

Feature: Identify Manually Entered Data And the Author of it

If an app enables recording of outside measurements that app **SHALL** identify measures as that have been manually entered by the user of the app, and the user **SHALL** be identified also.

Scenario: Manually Entered Data Can be Detected

Manualy entered data shall be distinguishable from device recorded data.

GIVEN

An app that enables user entry of data manually,

AND

a <User>

AND

a manually entered <Measurement>

WHEN

<Measurement> is accessed by another system or transmitted to infrastructure

THEN

The <Measurement> can be identified as having been manually entered.

Scenario: The User Who Entered the Data Can be Identified.

The user manually entering data shall be recorded.

GIVEN

An <App> that enables user manual entry of data

AND

a <User>

AND

a manually entered <Measurement>

WHEN

<Measurement> is accessed by another system or transmitted to infrastructure

THEN

The <Measurement> can be identified as having been manually entered by <User>.