(39 days)
The purpose of Dexcom Share Direct Secondary Displays is to notify another person, the Follower, of the patient's Dexcom G4 PLATINUM Continuous Glucose Monitoring System sensor glucose information. The Secondary Displays is intended for providing secondary notification of a continuous glucose monitoring system and does not replace real time continuous glucose monitoring (G4 PLATINUM System) or standard home blood glucose monitoring.
The Dexcom Share Direct Secondary Displays is not intended to modify or analyze data received from the continuous glucose monitor system. Nor is it intended to instruct, or to transmit information to the continuous glucose monitor system. The Dexcom Share Direct Secondary Displays is not intended to serve as a replacement for a primary display device for a continuous glucose monitoring system. The Dexcom Share Direct Secondary Displays is not intended to receive information directly from the sensor or transmitter of a continuous glucose monitoring system.
The Dexcom Share Direct Secondary Displays is a family of mobile applications (apps) that allows for the real-time monitoring of data from the Bluetooth enabled "Dexcom G4 PLATINUM Receiver with Share" by a third party by means of display on a compatible third party device with an internet connection. The system consists of two apps: one installed on the Continuous Glucose Monitor (CGM) patient's phone (the "Share2" app) and one installed on the phone of another person (the "Follow" app). Using the Share2 app, the patient can designate people ("Followers") with which to share their CGM data. The Share2 app receives real-time CGM data directly from the CGM receiver and transmits it to the Dexcom Cloud server (a web-based storage location). The Follow app. installed on a Follower's phone, can then download the CGM data and display it in real time.
The Dexcom Share Direct Secondary Displays was preceded by the Dexcom Share System. In the Dexcom Share Direct Secondary Displays, the CGM receiver is Bluetooth enabled and can communicate directly with the Share2 app on a patient's smart phone. In the predecessor, the Dexcom Share System, the CGM receiver must be placed in a Bluetooth enabled docking station which in turn transmits real-time CGM data to the Dexcom Share app on the CGM patient's mobile device. The Share app then transmits the internet to the Dexcom Share Cloud, from which the Follow app installed on a follower's phone can retrieve the data. The Dexcom Share Direct Secondary Displays eliminates the need for a docking station: the CGM receiver transmits information directly to the Share2 app on the CGM patient's phone. The Share2 app functions identically to the Share app, with the exception that it cannot communicate with the docking station. The same Follow app is used in both cases, as its function has not changed.
The Dexcom Share Direct Secondary Displays device performs the following functions:
- . Share2 App:
- o Transmit real-time glucose values and ancillary information (e.g. trends, alarms and notifications) to a third-party device/display/data storage system.
- Follow App:
- o Receive and display real-time glucose values.
- o Receive and display glucose trend information.
- o Receive and deliver notifications and alarms.
The Dexcom Share Direct Secondary Displays is used for secondary real-time monitoring of a patient's continuous glucose monitoring system. It is not intended to be the primary method of patient monitoring and it does not make treatment recommendations. It does not receive information directly from the sensor or primary transmitter of the CGM system; it receives information from the primary display device (in this case, the CGM receiver). It does not serve as a replacement for a primary display device. It does not make treatment recommendations, calculate insulin doses, or control any other medical devices.
The Dexcom Share Direct Secondary Displays device is a software-based device and its primary function is to transmit and display real-time glucose values and related information from a Dexcom G4 PLATINUM Continuous Glucose Monitoring System to a third-party device. As such, the acceptance criteria and performance evaluation focused on usability, data transmission accuracy, and compliance with regulatory requirements rather than traditional analytical or clinical performance metrics associated with glucose measurement.
Here's a breakdown of the requested information based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Safety and Effectiveness (Usability) | The Dexcom Share System (predecessor) was found to be adequately safe and effective for the intended users, its intended use, and use environment based on usability testing involving task-based scenarios. This was deemed applicable to the Dexcom Share Direct Secondary Displays because the interface and functionality of the Share and Follower apps remained unchanged. |
| Accuracy of Data Transmission | Bench testing demonstrated that data is transmitted accurately. Data from thirty (30) G4 PLATINUM Receivers with Share were uploaded from the Receivers to compatible Apple devices using the Share2 app and Follow app. Data on the Follow app were shown to be identical to the data on the receiver. |
| Software Development & Risk Mitigation | FDA reviewed and found acceptable the applicant's Hazard Analysis and software development processes, including level of concern, software description, device hazard analysis, software requirements specifications, architecture design chart, software design specification, traceability analysis, software development environment description, verification and validation testing, and revision level history for the Follow App. |
| Cybersecurity | Devices must include appropriate measures to protect against unauthorized access to data and unauthorized modification of data (Special control 1). This risk is mitigated by verification and validation, and design control activities. |
| Labeling - Treatment Decisions | Labeling must prominently and conspicuously display a warning: "Dosing decisions should not be made based on this device. The user should follow instructions on the continuous glucose monitoring system." (Special control 2). This mitigates the risk of incorrect treatment decisions. |
| Labeling - Self-Monitoring | Labeling must include a limitation: "This device is not intended to replace self-monitoring practices advised by a physician." (Special control 3). This mitigates the risk of over-reliance on followers. |
| Compliance with General Controls | The device must comply with general controls, including design controls. This ensures factors like safe design, manufacturing, and labeling practices. |
| No Modification/Analysis of Data | The device is not intended to modify or analyze data received from the continuous glucose monitor system, nor to instruct or transmit information to the continuous glucose monitor system. The device met this as it receives and displays data, not analyzes or modifies. |
| Not a Primary Display/Treatment Device | The device is not intended to serve as a primary display, receive information directly from sensor/transmitter, or control medical devices/calculate insulin doses. The device description confirms it receives data from the primary display device (CGM receiver) and its intended use explicitly states it does not make treatment recommendations or replace primary display. |
2. Sample Size Used for the Test Set and Data Provenance
- Usability Study (Clinical Study):
- Sample Size: 20 participant pairs (Sharer + Follower).
- Data Provenance: The document does not specify the country of origin. It was a usability study with simulated use, making it prospective in its design for evaluation, though the data itself is on user interactions rather than patient outcomes.
- Bench Testing (Data Transmission Accuracy):
- Sample Size: Data from thirty (30) G4 PLATINUM Receivers with Share.
- Data Provenance: Not specified, but likely laboratory-based bench testing. Retrospective in the sense of comparing data already generated by the receiver with data displayed by the app.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
- Usability Study (Clinical Study): The document does not explicitly state the number or qualifications of "experts" used to establish ground truth for this usability study. The "ground truth" here would likely be defined by the successful completion of predetermined task-based scenarios by the participants and assessment of their responses by study investigators or researchers, rather than clinical experts establishing a medical diagnosis or outcome.
- Bench Testing (Data Transmission Accuracy): The ground truth was the data directly from the G4 PLATINUM Receivers. This implies a comparison against the source data, without the need for independent expert review to establish truth.
4. Adjudication Method for the Test Set
The document does not describe a formal adjudication method (like 2+1 or 3+1 consensus) for either the usability study or the bench testing.
- Usability Study: Participants' responses to task-based scenarios were "scored," suggesting an objective measure against predefined correct actions or satisfactory outcomes.
- Bench Testing: Data on the Follow app was simply "shown to be identical" to the data on the receiver. This is a direct comparison, not requiring adjudication.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, If So, What Was the Effect Size of How Much Human Readers Improve with AI vs Without AI Assistance
No, an MRMC comparative effectiveness study was not done. This device is a data management software, not an AI-assisted diagnostic or interpretive tool that would involve human "readers" interpreting cases. Its function is to display glucose data, not to interpret or provide diagnostic information.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
Yes, the device's core function is largely standalone in terms of its data transmission and display capabilities. The "Bench testing" that showed data on the Follow app was identical to the data on the receiver demonstrates the standalone performance of the software in accurately replicating the glucose data without human intervention in the data processing itself. The usability study evaluated human interaction with the system, but the data transmission component itself is algorithmic.
7. The Type of Ground Truth Used
- Usability Study: The ground truth for the usability study was the successful completion of pre-defined task-based scenarios by participants, representing effective and safe interaction with the system's interface and functionality.
- Bench Testing: The ground truth for data transmission accuracy was the glucose data directly displayed on the primary Dexcom G4 PLATINUM Receiver itself.
8. The Sample Size for the Training Set
The document explicitly states: "The interface and functionality of the Share and Follower apps remains unchanged between these two systems, so it was determined that a new clinical study was not required for the Dexcom Share Direct Secondary Displays." This implies that the usability study performed for the predecessor device (Dexcom Share System) served as the primary clinical evaluation.
- Given it's a software device managing data display, the "training set" doesn't typically apply in the same way it would for an AI/ML algorithm that learns from data.
- If we consider the development and testing cycle of the software to be analogous to "training," then the software was developed and tested internally by Dexcom (as indicated by the acceptable software development processes) using an unspecified amount of internal data and testing environments. The provided document does not specify a distinct "training set" size for the software development.
9. How the Ground Truth for the Training Set Was Established
As noted above, a traditional "training set" with ground truth in the context of machine learning isn't applicable here. The software's "ground truth" during its development and internal testing phases would be derived from:
- Functional Requirements: The software was developed to meet specific functional requirements (e.g., transmit real-time glucose values, display trends, deliver notifications) as defined by engineering specifications.
- Verification and Validation Data: Internal testing (bench testing, unit testing, integration testing) would compare the software's output against expected outcomes based on the input data, with the "ground truth" being the correct data or system behavior as defined by the product's design and intent. The bench testing performed for accuracy (comparing app data to receiver data) would be a part of this validation.
§ 862.1350 Continuous glucose monitor secondary alarm system.
(a)
Identification. A continuous glucose monitor (CGM) secondary alarm system is identified as a device intended to be used as a secondary alarm for a CGM to enable immediate awareness for potential clinical intervention to help assure patient safety.(b)
Classification. Class II (special controls). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 862.9. The special controls for this device are:(1) Devices being marketed must include appropriate measures to protect against unauthorized access to data and unauthorized modification of data.
(2) The labeling must prominently and conspicuously display a warning that states “Dosing decisions should not be made based on this device. The user should follow instructions on the continuous glucose monitoring system.”
(3) The labeling for the device must include a statement that reads “This device is not intended to replace self-monitoring practices as advised by a physician.”