Search Results

The d-Nav® System calculates the next dose of insulin to aid in optimizing insulin management.

The d-Nav System contains two user-interactive software elements:

• The patient user interface software is intended for use by adults with Type 2 diabetes as an aid in optimizing insulin management. It resides on a hand-held device, e.g. cellular phone or enabled glucose meter, and is used to enter glucose event data and receive a recommended insulin dose.

· The HCP user interface software tool is intended for use by Health Care Providers (HCPs) to set up the patient software for its intended use. Setup consists of entering the physician-prescribed, patient-specific starting insulin dose instructions (insulin prescription) and sending the patient user software. Insulin instructions include the treatment algorithm (treatment plan), insulin drug, and dose(s).

The d-Nav System also contains the d-Nav Get-Dose Library that provides the next insulin dose.

The System can receive glucose measurement data entered manually into the patient user software or automatically via the cloud from a linked blood glucose meter. The d-Nav Get-Dose Library Recommend Dose function resides locally on the phone while the d-Nav Get-Dose Library Update Insulin Instruction may reside locally on the phone or be hosted in the cloud. Configurations are as follows:

Model 1: Patient user software resides on a hand-held device and uses manual glucose measurement entry. The Get-Dose Library Update Insulin Instruction function resides locally within the device.

Model 2: Patient user software resides on a hand-held device and uses manual glucose measurement entry. The Get-Dose Library Update Insulin Instruction function resides in the cloud.

Model 3: Patient user software resides on a hand-held device and uses automated glucose measurement entry. The Get-Dose Library Update Insulin Instruction function resides locally within the device.

Model 4: Patient user software resides on a hand-held device and uses automated glucose measurement entry. The Get-Dose Library Update Insulin Instruction function resides in the cloud.

Use of the d-Nav System is limited to Health Care Providers who have been trained by Hygieia trained person on the use of the d-Nav System, including setup of the patient's Phone App.

The d-Nav System is a software-based, prescription-only product designed to provide the next insulin dose recommendation as an aid for personal insulin management. The product integrates the Health Care Provider (HCP) prescribed starting insulin dose instructions with automated dosing quidance to the patient based on comparing reqularly measured blood glucose data trends to a device specified target range. The d-Nav System contains two userinteractive software elements; the d-Nav Phone App and the d-Nav Website.

• . The Phone App is for use by persons with Type 2 diabetes as an aid in optimizing insulin management. The Phone App resides on a cellular phone and is used by the patient to enter qlucose event data and receive a recommended insulin dose. The blood qlucose data is obtained from an over-the-counter, cleared Blood Glucose (BG) device and entered into the software system either through manual BG data entry using the Phone App kevpad or via a cloud-pushed mechanism from a linked blood qlucose meter. The Phone-App allows change in insulin dose recommendations to be sent to the patient without concurrence from the prescriber.
• The d-Nav Website is for use by Health Care Providers that have been trained by Hygieja or . a Hygieia trained trainer on the use of Phone App and website to set up the patient's Phone App software for its intended use. Setup consists of entering the physician prescribed, patient-specific starting insulin dose instructions and sending the information to the intended patient's Phone App. Insulin instructions include the treatment algorithm (treatment plan), insulin druq, and dose(s).

The d-Nav System also contains the d-Nav Get-Dose library that provides the next insulin dose.

The System can receive glucose measurement data entered manually into the patient user software or automatically via the cloud from a linked blood qlucose meter. The d-Nav Get-Dose Library Recommend Dose function resides locally on the phone while the d-Nav Get-Dose Library Update Insulin Instruction may reside locally on the phone or be hosted in the cloud.

Here's a breakdown of the acceptance criteria and the study details for the d-Nav System, extracted from the provided text.

The document primarily focuses on demonstrating substantial equivalence to a predicate device ("My Insulin Doser (MID) / Intelligent Dosing System (IDS)") rather than presenting a detailed clinical study with specific performance metrics and acceptance criteria in the format typically used for de novo or PMA applications. Therefore, some information, particularly quantitative acceptance criteria and detailed study results with effect sizes, is not explicitly provided in the typical sense of a clinical trial report. Instead, the performance data presented refers to verification and validation of the software system and human factors testing, aiming to confirm that the new device does not raise new safety or efficacy concerns compared to the predicate.

Acceptance Criteria and Reported Device Performance

The provided document does not explicitly present a table of quantitative acceptance criteria and corresponding numerical device performance results. The "performance data demonstrating substantial equivalence" section primarily discusses risk management, cybersecurity, software verification/validation, and human factors testing, all of which are qualitative or process-oriented rather than metric-based performance.

The acceptance criteria can be inferred as demonstrating that the d-Nav System is:

• Substantially equivalent to the predicate device in terms of intended use and technological characteristics.
• Does not raise new issues of safety and/or efficacy compared to the predicate.
• The software is verified and validated according to FDA guidance for a "Major" level of concern.
• Human factors testing supports the determination of substantial equivalence.

Reported Device Performance (as per the document):

• Risk analysis conducted in accordance with ISO 14971:2007.
• Cybersecurity evaluation conducted according to FDA guidance.
• Software verification and validation performed to FDA Guidance (May 11, 2005) for software with a "Major" level of concern, with test coverage including load testing.
• Human factors testing performed according to FDA Guidance (February 3, 2016), supporting a determination of substantial equivalence.
• Requirements traceability matrix demonstrates full coverage of requirements.

Essentially, the "performance" is stated as compliance with relevant standards and guidelines, and the absence of new safety/efficacy concerns compared to the predicate.

Study Details:

Given the nature of the submission (510(k) for substantial equivalence), the "study" described is primarily focused on software verification/validation and human factors, rather than a clinical effectiveness study.

1. A table of acceptance criteria and the reported device performance
As noted above, explicit quantitative acceptance criteria and performance metrics are not provided in a table. The performance demonstrated is primarily documented through adherence to regulatory processes and standards for software and human factors.

2. Sample sized used for the test set and the data provenance

• Test Set Sample Size: The document does not specify general "test set" sample sizes for clinical performance. For software verification and validation, it mentions "test coverage" but no specific number of tested configurations or data points. For human factors testing, sample size is not disclosed.
• Data Provenance: Not explicitly stated as real-world patient data. The context implies testing of the software system itself, possibly using simulated data or internal testing procedures. There is no mention of country of origin of data or whether it was retrospective or prospective patient data for performance evaluation.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

• This information is not provided as the submission does not detail a clinical study where expert consensus was used for ground truth. Ground truth for the functionality of the device is inherent in its design and adherence to medical guidelines for insulin dosing management.

4. Adjudication method for the test set

• This information is not applicable/provided as there is no multi-rater clinical study described requiring adjudication of expert opinions.

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 presented. The device is an insulin dose calculator, not an imaging AI diagnostic aid for human readers. Its purpose is to provide a recommended insulin dose, not to assist human interpretation of complex medical cases. Therefore, the concept of "human readers improve with AI" is not directly relevant to this device's function.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• The device inherently involves a human-in-the-loop (the patient and the HCP). The "d-Nav Get-Dose Library" is the algorithmic component. The software verification and validation would have tested the standalone performance of this library in calculating doses based on inputs. However, the document does not present specific performance metrics for the algorithm in isolation as a clinical outcome. Its "standalone" function is calculating doses, which is then integrated into the user interface for patient and HCP interaction.

7. The type of ground truth used

• The "ground truth" for this device's function is the clinically accepted principles of insulin management and dosing. The algorithm's calculations are based on established medical understanding of how insulin doses should be adjusted based on blood glucose readings and treatment plans. This is implicit in the device's purpose as a "predictive pulmonary-function value calculator" (which seems to be a regulatory classification misnomer in the initial letter, and should be "insulin dose calculator" as per the comparability table). The document does not refer to pathology, expert consensus (in a diagnostic sense), or outcomes data as external "ground truth" for its performance validation; rather, its validity stems from its adherence to therapeutic dosing principles.

8. The sample size for the training set

• This information is not applicable/provided. The d-Nav System is a rule-based or algorithm-based dose calculator, not a machine learning model that requires a "training set" in the conventional sense of AI/ML development. Its logic is programmed based on medical formulas and treatment plans, not learned from a dataset.

9. How the ground truth for the training set was established

• This information is not applicable/provided for the same reason as point 8.

Ask a specific question about this device