Search Results

The Health-e-Connect System with IDA is intended for use in the home and clinical settings by people with diabetes and healthcare providers as an aid in the review, analysis, and evaluation of historical glucose test results and associated usage data in support of an effective diabetes management program.

The Insulin Dose Adjustment (IDA) feature is intended only for insulin-requiring Type 2 diabetes patients to provide the physician with two reference doses.

The IDA feature is not indicated for patients who utilize insulin pumps and it is limited to adults with Type 2 diabetes on fixed dose regimen of insulin.

The Health-e-Connect System with IDA is for patients under the supervision of a physician / healthcare provider trained in the management of diabetes. Final drug dose recommendations for a patient must be made only after careful consideration of the full clinical status of the patient. No medical decision should be based solely upon the results provided by this software program.

The ALRT Health-e-Connect System (HeC) allows healthcare providers, ALRT staff, and other authorized caregivers to remotely monitor the blood glucose values of patients with diabetes and therefore can assist healthcare providers in making adjustments to the patient's care plan based upon trends in the patient's blood glucose data. There are no physical, electrical, biocompatible, or sterility specifications for this device as it is software only.

The original Health-e-Connect system (K102063) performed two functions:

1. A data management tool and
2. A communication platform (Health-e-Connect Remote Care System).

Modification:
The proposed modification is to add a module - Insulin Dose Adjustment (IDA) - to the Healthe- Connect System.

This is an additional module of software that monitors patient blood glucose levels uploaded from the patient's blood glucose meter, to ascertain whether the patient's current insulin dose may or may not be optimal. If trends in the patient's blood glucose are outside of the guidelines set by the AACE and ADA, the patient is flagged for a potential insulin dose adjustment. The IDA system will then employ the AACE and ADA algorithms to calculate reference doses that can be compared with the patient's current insulin dose. If there is an inconsistency between the patient's current insulin dose as compared to the reference doses calculated by AACE and ADA algorithms, this discrepancy will be flagged and an alert sent to the managing HCP requesting an insulin dose review.

The provided text describes a 510(k) submission for the "Health-e-Connect System with IDA (Insulin Dose Adjustment)". While it details the device, its intended use, and a comparison to predicate devices, it does not contain the specific acceptance criteria or an explicit study that proves the device meets those criteria with quantitative performance metrics.

The document discusses "Verification and validation testing" and a "Human Factors / Usability study" but does not provide details on the specific performance outcomes of these tests in relation to predefined acceptance criteria. It focuses on demonstrating substantial equivalence to predicate devices rather than proving performance against specific numerical targets.

Therefore, I cannot populate the requested table or answer most of the questions directly from the provided text.

Here's what can be inferred or stated based on the given information, with limitations:

1. Table of acceptance criteria and reported device performance:

2. Sample size used for the test set and the data provenance:

• Sample size for test set: Not specified.
• Data provenance: Not specified (e.g., country of origin, retrospective/prospective). The document mentions the device monitors "blood glucose levels uploaded from the patient's blood glucose meter," implying real-world data, but no specifics on the test set's origin.

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

• Not specified. The IDA feature uses "AACE and ADA algorithms to calculate reference doses." This suggests the "ground truth" for the calculated doses is based on these established clinical guidelines, rather than expert human interpretation of individual cases for the purpose of a test set.

4. Adjudication method for the test set:

• Not specified. Given that the IDA feature calculates reference doses based on established algorithms (AACE and ADA), it's unlikely a traditional human adjudication process for a test set was applied in the same way it would be for an AI model that interprets medical images. The "ground truth" is algorithmic.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done:

• No, an MRMC study is not mentioned. The device's function is to provide "reference doses" to a physician, not to interpret complex medical data like images that would typically necessitate an MRMC study for human reader performance evaluation.

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

• The "Nonclinical Performance Testing" and "Verification and validation testing" suggest an evaluation of the software's functionality, which would be a form of standalone testing for the algorithm's calculation accuracy of reference doses. However, no specific performance results are provided. The function of the IDA feature is to "employ the AACE and ADA algorithms to calculate reference doses," so its standalone performance would be about its fidelity to these algorithms.

7. The type of ground truth used:

• The ground truth for the IDA feature's calculations appears to be based on the guidelines and algorithms established by the American Association of Clinical Endocrinologists (AACE) and the American Diabetes Association (ADA). The device calculates "reference doses" using these established guidelines.

8. The sample size for the training set:

• Not specified. The document primarily describes an algorithmic approach based on existing clinical guidelines rather than a deep learning model requiring a large training set in the conventional sense.

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

• Not explicitly a "training set" in the machine learning sense. The "ground truth" for the IDA feature's logic is derived from and established by the AACE and ADA guidelines/algorithms. The system "employs" these existing expert-developed algorithms.

Ask a specific question about this device

The Health-e-Connect System is intended for use in the home and clinical settings by people with diabetes and healthcare providers as an aid in the review, analysis and evaluation of historical glucose test results and associated usage data in support of an effective diabetes management program.

The ALRT Health-e-Connect System (HeC) is an internet based blood glucose monitoring system that allows healthcare providers and patients the opportunity to review, analyze and evaluate the efficacy of a diabetes management program. It is expected that this functionality will significantly improve HbA1C levels in patients with diabetes. Note there are no physical, electrical, biocompatibility or sterility specifications for this device as it is software only. It performs two functions: it is a data management tool and a communication platform. The Health-c-Connect System is comprised of a home based application, legally marketed peripherals (blood glucose meters) and a server. The home based application software collects data from blood glucose meters and transmits the data over the home's existing internet connection where it is uploaded to the Health-e-Connect System's web-based servers. The server is a web-based application that collects, range checks, stores and displays historical patient blood sugar levels. It also allows patients, healthcare providers, patient relatives and other healthcare providers involved in the case to send messages to each other and share patient information. This communication is retrospective and not a real- time alert or alarm. The Healthe-Connect System is a tool to monitor patients remotely and motivate them through notifications.

The ALRT Health-e-Connect System is a software-only device designed as an internet-based blood glucose monitoring system. It allows healthcare providers and patients to review, analyze, and evaluate the efficacy of a diabetes management program. The device acts as a data management tool and communication platform, collecting data from legally marketed blood glucose meters and transmitting it to web-based servers for storage and display of historical patient blood sugar levels. It also facilitates communication among patients, healthcare providers, and patient relatives.

Here is an analysis of its acceptance criteria and the supporting studies:

1. Table of Acceptance Criteria and Reported Device Performance

2. Sample Size Used for the Test Set and Data Provenance

• Bench Testing: The test set for bench testing involved 10 different glucose meter models from various manufacturers (Bayer, Abbott, Roche, Lifescan). The data provenance is not explicitly stated as country of origin, but it can be inferred that these are commercially available meters. The testing appears to be prospective as it involved performing specific procedures (taking blood tests, uploading data, checking values, etc.) to verify the system's functionality.
• Usability / Human Factor Studies (HFS): The test set included 22 lay users and 2 healthcare providers. The data provenance is not specified regarding country of origin, but the study was conducted to evaluate the usability of the Health-e-Connect system. This was likely prospective through surveys and observations of user interactions.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications

• Bench Testing: For the "Correct Glucometer summary (checked against paper copy & calculations)" and "Glucose analysis is flawed and the results are not correct in the HeC-RCS" tests, the ground truth was established by checking against paper copies & calculations and by comparing the analysis code against other algorithms. While it implies expert knowledge in calculations and algorithm design, the number and specific qualifications of human experts involved in these direct "ground truth" determinations are not specified.
• Usability / Human Factor Studies: The "ground truth" for usability was established through the feedback and successful task completion of the 22 lay users and 2 healthcare providers. These individuals served as the "experts" in evaluating user experience. Their specific qualifications beyond being "lay users" and "healthcare providers" are not detailed (e.g., specific roles or years of experience).

4. Adjudication Method for the Test Set

• Bench Testing: The adjudication method for bench testing involved direct verification against established references (paper copies, calculations, other algorithms) and observations of successful data transfer and accurate association. There is no mention of a formal multi-expert adjudication method (like 2+1 or 3+1). The "Pass" results suggest a direct factual comparison rather than a consensus approach.
• Usability / Human Factor Studies: The usability study relied on surveys and observations of the 22 lay users and 2 healthcare providers. The conclusion states that the system was "user friendly, fast and easy to use," implying a qualitative assessment based on the collected feedback rather than a formal multi-reader adjudication process.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No MRMC comparative effectiveness study was done. The submission does not mention any study comparing human reader performance with and without AI assistance for tasks related to interpreting glucose data. The device is a data management and communication tool, not an AI for medical image or diagnostic interpretation.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study

A standalone performance study of the data processing algorithms was conducted as part of the bench testing.

• The test: "Glucose analysis is flawed and the results are not correct in the HeC-RCS".
• The procedure: "The code which analyzes the glucometers values is checked thoroughly against other algorithms calculating the same statistics to verify accuracy."
  This demonstrates an evaluation of the algorithm's accuracy in processing glucose values independently.

7. Type of Ground Truth Used

• Bench Testing: The ground truth used for bench testing included:
  • Manual calculations and paper records for verifying summary accuracy.
  • Direct comparison with values displayed on the physical glucometers for data transfer accuracy.
  • Comparison against other established algorithms for the accuracy of glucose analysis.
  • Manual logging and checking for data association.
• Usability / Human Factor Studies: The ground truth was user feedback and observed task completion, assessing aspects like user-friendliness, speed, and ease of use.

8. Sample Size for the Training Set

The document does not provide any information regarding a training set sample size. This is a software system for data management and display, not a machine learning or AI model trained on a dataset in the conventional sense for diagnostic or predictive purposes. Its functionality is based on established logical processes for data handling and communication.

9. How the Ground Truth for the Training Set Was Established

As there is no mention of a training set in the context of machine learning, the question of how its ground truth was established is not applicable based on the provided document. The software's functionality is based on direct implementation of data handling rules and communication protocols rather than learning from a labeled dataset.

Ask a specific question about this device