K Number
K160673
Manufacturer
Date Cleared
2017-06-07

(454 days)

Product Code
Regulation Number
868.1890
Panel
HO
Reference & Predicate Devices
Predicate For
N/A
AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
Intended Use

The Insulin Algorithms System (IA System) is a glycemic management software support program designed to provide insulin dosing recommendations to the clinician in order to adjust and maintain blood glucose levels in insulin-dependent diabetic patients within a clinician determined target range. The IA System evaluates a set of current and cumulative patient blood glucose values together with biometric information (including age, weight, and gender) and suggests changes (if necessary). The system is designed for use by healthcare professionals. The IA System is indicated for use in adults and pediatric patients ages 7 and above who weigh more than 36 lbs.

The IA System is not intended to be used to provide insulin dosing recommendations to the clinician in order to adjust insulin treatment for patients hospitalized in an acute care setting. Also, the IA System is not intended to be used to provide insulin dosing recommendations for patients on insulin pump therapy or IV insulin therapy. Nor is IA System intended to provide recommendations for mealtime injections of insulin that are based on carb counting.

The IA System is not a substitute for, but rather an adjunct to clinical reasoning. No medical decision should be based solely on the recommended guidance provided by this software program.

Device Description

The Insulin Algorithms System (IA System) takes basic biometric information (age, height, weight, gender), insulin dosing information (prescribed insulins, injection times, and amounts), meal information (meal times, before meal glucose targets, after meal glucose targets), along with a collection of glucometer readings and processes this information though a custom algorithm to create a recommendation as to what changes to the patient's insulin regimen should be made. The algorithm was developed using the insulin dosing conventions taught and used in

The IA System is a software solution. It is comprised of two separate software systems: a server application containing all algorithms and mitigations and a client application that serves as an interface to the server. All data are stored and analyzed on the server application. Each user has an individual identification and password to access the application. The IA System is designed to safeguard the confidentiality, integrity, and availability of electronic health information in conformance with the Health Insurance Portability and Accountability Act (HIPAA).

AI/ML Overview

The provided text from the FDA 510(k) submission for the Insulin Algorithms System (IA System) does not contain detailed information about specific acceptance criteria and the results of a study proving the device meets those criteria, as typically presented in a clinical validation report.

The document primarily focuses on:

  • The FDA's decision regarding substantial equivalence to predicate devices.
  • The device's indications for use.
  • A high-level description of the software and its development process (software verification & validation).
  • A brief mention of usability testing.

Therefore, many of the requested fields cannot be fully populated as the information is not present in the provided text.

Here's an attempt to answer the questions based on the available information, with specific notes where information is missing:

1. Table of Acceptance Criteria and Reported Device Performance

This information is not explicitly provided in the document. The document mentions "Software verification & validation testing was conducted" and "All features provided for the IA System have been verified to operate as specified," but it does not detail specific quantitative acceptance criteria (e.g., accuracy percentages, error margins for insulin dosing recommendations) or their corresponding performance metrics.

Acceptance Criteria (e.g., target accuracy, precision)Reported Device Performance (e.g., achieved accuracy, precision)
(Not explicitly stated in the document)(Not explicitly stated in the document)
Functions operate as specified (qualitative)All features verified to operate as specified (qualitative)
Usability for target users (qualitative)Medical professionals understood scope/limit and used with ease/confidence (qualitative)

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

This information is not provided in the document. The text mentions "Usability testing" and "Software verification & validation testing," but does not specify the sample size of patient data or users involved in these tests, nor the origin (country, retrospective/prospective) of any data used.

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

This information is not provided in the document. The document mentions "usability testing indicates that the target group of medical professionals understood the scope and limit of the application," but it doesn't describe their role in establishing a "ground truth" for the device's recommendations. For an algorithm providing medical recommendations, this would typically involve expert review of cases to determine the "correct" or "optimal" recommendation.

4. Adjudication Method for the Test Set

This information is not provided in the document. Since the method for establishing ground truth isn't described, an adjudication method for disagreements among experts also cannot be determined from the text.

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

A MRMC comparative effectiveness study was not specifically mentioned or detailed in the provided text. The device is described as a "software support program designed to provide insulin dosing recommendations to the clinician," implying it's an adjunct tool. However, there's no study comparing human readers (clinicians) with and without AI assistance to quantify an "effect size" of improvement.

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

The document describes the IA System as "a glycemic management software support program designed to provide insulin dosing recommendations to the clinician." It is explicitly stated: "The IA System is not a substitute for, but rather an adjunct to clinical reasoning. No medical decision should be based solely on the recommended guidance provided by this software program." This strongly implies that a standalone performance study, where the algorithm's recommendations are directly applied without human review, was not the focus or, if done, is not detailed here as the device is designed for human-in-the-loop use. The "Software verification & validation" may have included isolated algorithm testing, but a study of its standalone clinical performance is not described.

7. Type of Ground Truth Used

This information is not explicitly stated for the validation of the algorithm's output. The algorithm itself was "developed using the insulin dosing conventions taught and used in..." (the sentence cuts off at the page break, but implies established medical practice). For actual performance testing, the specific method for determining the "correct" insulin recommendation for comparison purposes is not detailed.

8. Sample Size for the Training Set

This information is not provided in the document.

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

This information is not provided in the document. It is mentioned that "The algorithm was developed using the insulin dosing conventions taught and used in" (cuts off), implying that established medical guidelines and conventions formed the basis, but the specific process of creating a "ground truth" for a training set (if one was used for machine learning, which isn't explicitly stated) is not detailed.

§ 868.1890 Predictive pulmonary-function value calculator.

(a)
Identification. A predictive pulmonary-function value calculator is a device used to calculate normal pulmonary-function values based on empirical equations.(b)
Classification. Class II (performance standards).