Search Results

EndoTool SubQ 2.1 is a software application for use by trained healthcare professionals to calculate and recommend an individual patient's next dose of insulin to be administered subcutaneously to manage blood glucose levels in patients with hyperglycemia including adult and pediatric patients (age 2 years and above and 12 kg or more). The software is designed to recommend the insulin dose(s) and when indicated a carbohydrate dosed on the prescribing healthcare Provider's nutritional regimen, insulin regimen, target glucose range, and patient specific characteristics.

EndoTool SubQ 2.1 logic is not a substitute for clinical reasoning but an aid for trained healthcare professionals based on obtained glucose readings and entered clinical data. Final dose recommendations for a patient must be made only after consideration of the full clinical status of the patient. No medical decision should be made based solely upon the results provided by this software program.

EndoTool SubQ 2.1 is a software application for use by trained healthcare professionals to calculate and recommend an individual patient's next dose of insulin to be administered subcutaneously to manage blood glucose levels in patients with hyperglycemia including adult and pediatric patients (age 2 years and above and 12 kg or more). The software is designed to recommend the insulin dose(s) and when indicated a carbohydrate dose based on the prescribing healthcare Provider's nutritional regimen, insulin regimen, target glucose range, and patient specific characteristics.

The EndoTool® SubQ 2.1 Glucose Management system includes security, software, and technical support features. Each user has an individual User Identification (ID) and Password to access portions of the application. EndoTool SubQ 2.1 is designed to safeguard the confidentiality, integrity, and availability of electronic protected health information of patients according to the Health Insurance Portability and Accountability Act (HIPAA) privacy rules.

EndoTool SubQ 2.1 is packaged in a user friendly, stand-alone program. The application is installed on Windows Server 2008 R2 or newer. The end-user should access the application using a compatible web browser, such as Internet Explorer 9 and higher, or Google Chrome 58 and higher. The application was developed for use on Personal Computers (PCs), network servers, and terminal server environments. As EndoTool SubQ 2.1 data is time sensitive, it is also imperative that all PCs and servers be set with the correct date and time using UTC.

EndoTool SubQ can utilize barcode scanning in Code 39 format (also known as Alpha39, Code 3 of 9, Code 3/9, Type 39, USS Code 39, or USD-3) for patient identification/verification.

The provided text does not contain detailed information about specific acceptance criteria for the output of the EndoTool SubQ 2.1 software or a study demonstrating that the device meets such criteria. Instead, it focuses on demonstrating substantial equivalence to a predicate device through comparison of technological characteristics and general software verification and validation activities.

Here's an analysis of the information that is available and what is missing:

1. Table of Acceptance Criteria and Reported Device Performance

Not available in the provided text. The document states that "Software verification and validation testing was conducted per IEC 62304 and FDA's software validation guidance to demonstrate the performance of the device is substantially equivalent to the predicate." It then lists types of testing performed (requirements-based, module and integration, data integration, algorithm test cases, static analysis, regression testing), but it does not provide specific quantitative acceptance criteria (e.g., target accuracy, precision, or other performance metrics for insulin dose recommendations) or the reported performance values against these criteria.

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

Not explicitly available. The document mentions "Algorithm test cases (automated and manual)" as part of performance testing, which implies a test set was used. However, the size of this test set, the nature of the data (e.g., patient records, simulated data), and its provenance (country of origin, retrospective or prospective collection) are not specified.

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

Not available. Since the details of the "Algorithm test cases" are not provided, there is no information on how ground truth was established for these cases, nor the number or qualifications of any experts involved.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set

Not available. This information is dependent on the details of how the ground truth was established, which is not provided.

5. If a multi-reader, multi-case (MRMC) comparative effectiveness study was done, and its effect size

No, a MRMC comparative effectiveness study was not explicitly mentioned and does not appear to have been performed. The document focuses on the software's performance and its substantial equivalence to a predicate device, not on comparing human reader performance with and without AI assistance. The device is described as "an aid for trained healthcare professionals," not a tool to enhance human reading of medical images or data in a comparative effectiveness study context.

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

Yes, implicitly. The "Algorithm test cases (automated and manual)" would involve evaluating the algorithm's output independently. The device's description as a "software application for use by trained healthcare professionals to calculate and recommend an individual patient's next dose of insulin" implies standalone algorithmic calculations. However, the performance metrics from this standalone evaluation are not provided.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

Not explicitly available. For "Algorithm test cases," the ground truth would likely be a clinically validated optimal insulin dose or a set of expected outcomes based on established medical protocols for various patient scenarios. However, the specific method of establishing this ground truth is not detailed.

8. The sample size for the training set

Not available. The document describes a software application that calculates insulin doses. While such applications often involve algorithms that might be "trained" or developed based on data, there is no mention of a distinct "training set" or its size in the context of machine learning or AI. The software is described as having an "algorithm" and being "designed to recommend the insulin dose(s)", suggesting it operates based on predefined logic and parameters rather than a continuously learning model with a training phase.

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

Not applicable/Not available. As above, the concept of a "training set" and establishing ground truth for it is not explicitly discussed for this type of software, which appears to be rule-based or algorithmic rather than a machine learning model requiring labeled training data.

In summary: The provided document is a 510(k) summary focused on demonstrating substantial equivalence of EndoTool SubQ 2.1 to a predicate device, EndoTool SubQ (K180366). It outlines general software validation practices and a comparison of features but lacks the specific performance data, acceptance criteria, and details of clinical studies that would typically be described for a diagnostic AI device requiring quantitative performance metrics against a defined ground truth. The device is a "Predictive Pulmonary-Function Value Calculator" (Regulation Name) but its "Indications for Use" clearly state its function is for "calculat[ing] and recommend[ing] an individual patient's next dose of insulin... to manage blood glucose levels." This discrepancy might be a clerical error in the regulatory classification.

Ask a specific question about this device