Search Results
Found 1 results
510(k) Data Aggregation
(261 days)
Go Dose System
The Go Dose System, comprised of the Go Dose Pro applications, is for use in home and clinical settings to aid in the review, analysis, and evaluation of historical blood glucose test values to support type 2 diabetes mellitus management. The Go Dose System is a mobile application for use with iPad or iPhone mobile devices. The Go Dose System provides recommendations for titrating prandial Humalog dosing one meal at a time using blood glucose values entered by the patient. Go Dose Pro applications are not intended to replace the care provided by a licensed health care provider or to provide any diagnosis on patient blood glucose values. Go Dose Pro applications are for prescription use only.
Go Dose is for prescription use only by persons who are:
- Diagnosed with type 2 diabetes mellitus (DM)
- Aged 18-85 years
- Failing to achieve glycemic targets despite optimization of insulin glargine, with or without metformin
- Requiring treatment intensification with Humalog 100 units per mL (U-100)
- For whom the set blood glucose target range of 85-114 mg/dL (4.7-6.3 mmol/L) is appropriate
Go Dose should be used multiple times per day, as directed by your healthcare provider, for entering blood glucose values and getting Humalog dosing recommendations.
The Go Dose daily adjusted dosing algorithm was tested using Humalog 100 units per mL (U-100) in patients with type 2 diabetes who were also taking insulin glargine, with or without metformin. The algorithm for Go Dose is not designed for the titration of NPH, regular insulin, or basal and analog insulins at concentrations other than 100 unit/mL.
Go Dose Pro is intended for use by health care providers who have experience prescribing mealtime insulin, including providing direction to patients within the scope of a preplanned treatments to prescribed insulin. Health care providers should use the Go Dose Pro to aid in the review, analysis and evaluation of historical blood glucose test values to support diabetes management in patients that meet the following criteria:
- Diagnosed with type 2 diabetes mellitus (DM)
- Aged 18-85 years
- Failing to achieve glycemic targets despite optimization of insulin glargine, with or without metformin
- Requiring treatment intensification with Humalog 100 units per mL (U-100)
- For whom the set blood glucose target range of 85-114 mg/dL (4.7-6.3 mmol/L) is appropriate
The Go Dose design is based on a dosing algorithm that was tested using Humalog 100 units per mL (U-100) in patients with type 2 diabetes who were also taking insulin glargine, with or without metformin. The algorithm for Go Dose is not designed for the titration of NPH, regular insulin, or basal and analog insulins at concentrations other than 100 unit/mL.
The Go Dose System, comprised of Go Dose Pro mobile medical applications (apps), aids patients under the supervision of their health care provider (HCP) in the management of type 2 diabetes mellitus (T2DM).
This medical device system is based on a "paper-based" algorithm approach that was studied for safety and efficacy in 2 multinational, multicenter, randomized, open-label, parallel-group studies within a single clinical trial, F3Z-MC-IOQC (Trial IOQC) as described in Edelman et al. 2014. This study enrolled adult patients with T2DM who were not under glycemic control after optimizing the dose of insulin glargine for fasting blood glucose, with or without metformin. The medical device provides HCPs and patients with an evidence-based approach to initiate and titrate Humalog® 100 units/mL (insulin lispro) therapy for patients continuing insulin glargine, with or without metformin.
The HCP determines the meal selection for Humalog titration and the Go Dose Pro app, and sends the dose recommendation to the patient's Go Dose app. The patient uses the Go Dose app to view the initial dose recommendation. The Go Dose app uses blood glucose values manually entered by the patient as an input to determine ongoing individualized dose recommendations. This titration process continues until the HCP determines that the dose can be stabilized (i.e., no further titration required) or stopped (i.e., discontinue that dose). At that point, the HCP will determine whether to select a new meal for dose titration, stop any additional use of the Go Dose System, or select additional stable pre-meal dose(s).
The Go Dose System is a mobile medical application designed to aid in the review, analysis, and evaluation of historical blood glucose test values to support type 2 diabetes mellitus management. It provides recommendations for titrating prandial Humalog dosing.
Here's an analysis of the acceptance criteria and the study proving device performance:
1. Table of Acceptance Criteria and Reported Device Performance
The FDA 510(k) summary doesn't explicitly list a table of "acceptance criteria" against "reported device performance" for the Go Dose System as one might find for a hardware device's technical specifications (e.g., accuracy of a measurement). Instead, the submission relies on the historical clinical validation of the underlying algorithm and extensive software verification and validation.
The "acceptance criteria" are implied through the statement that "the predetermined acceptance criteria were met" during summative and supplemental validation studies (Page 11). The primary performance validation for the algorithm itself is derived from the "Trial IOQC" study. The device's performance, therefore, is directly linked to the outcomes of this clinical trial in terms of its ability to initiate and titrate prandial insulin lispro (Humalog) therapy and improve diabetes management.
Implied Acceptance Criteria (from the clinical study and device's purpose):
| Acceptance Criteria (Implied) | Reported Device Performance (from Trial IOQC) |
|---|---|
| Effective initiation and titration of prandial Humalog dosing | The algorithm, implemented in the Go Dose System, demonstrated its ability to initiate and titrate prandial insulin lispro in adult patients with T2DM who were not under adequate glycemic control, comparing favorably to other patient-driven approaches. (Based on Edelman et al. 2014) |
| Safety of dosing recommendations (avoiding hypoglycemia/hyperglycemia risks) | The algorithm included safety limits, such as no dosing recommendation for hypoglycemia, a 1-unit dosage increase limit per day, and a maximum Humalog dose of 30 units per dose. The clinical trial results (Edelman et al. 2014) would implicitly support the safety profile. |
| Software functionality and reliability | Software verification and validation activities (including a Safety Assurance Case and Human Factors testing) demonstrated that software specifications were correctly implemented, and the system is substantially equivalent to the predicate device. |
| User-friendliness and effective patient/HCP interaction | Human Factors testing demonstrated substantial equivalence to the predicate device, implying usability and effective interaction for patients and HCPs. |
2. Sample Size Used for the Test Set and Data Provenance
- Test Set Description: No specific "test set" for the device software is described with a distinct sample size in the context of the clinical trial. However, the clinical algorithm was tested in a large clinical trial.
- Sample Size for Algorithm Validation (Trial IOQC): The details state that Trial IOQC enrolled "adult patients with T2DM who were not under glycemic control after optimizing the dose of insulin glargine for fasting blood glucose, with or without metformin." While the exact number isn't provided in the excerpt, the paper referenced (Edelman et al. 2014) would contain this information. The description mentions "2 multinational, multicenter, randomized, open-label, parallel-group studies within a single clinical trial." This implies a substantial sample size.
- Data Provenance: "multinational, multicenter" indicates data from multiple countries and clinical sites. The studies were prospective as they were described as "randomized, open-label, parallel-group studies within a single clinical trial."
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications
This information is not provided in the document. The "ground truth" for the algorithm was established through the clinical trial's outcomes, where patient blood glucose levels and insulin doses were collected and analyzed directly. There is no mention of a separate panel of experts establishing ground truth for the clinical algorithm retrospective evaluation against a "test set" in the traditional sense of an AI model's validation. The experts involved would be the clinical investigators and trial statisticians of the "Trial IOQC."
4. Adjudication Method for the Test Set
This information is not provided. Given the nature of the clinical trial (randomized, open-label, parallel-group), outcomes were likely adjudicated based on pre-defined clinical endpoints and statistical analysis rather than a consensus of experts retrospectively labeling data.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size
No, an MRMC comparative effectiveness study was not done. This type of study is typically used for image-based diagnostic aids where multiple human readers interpret cases with and without AI assistance. The Go Dose System is an insulin dosing calculator, not an image interpretation tool. Its effectiveness is measured by clinical outcomes related to glycemic control.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
Yes, the underlying algorithm's performance was evaluated in a standalone manner within the clinical trial. The "paper-based" algorithm, which is the foundation of the Go Dose System, was studied for safety and efficacy in "Trial IOQC" before being implemented in the mobile application. The mobile application then underwent software verification and validation, but the core clinical performance relied on the prior clinical validation of the algorithm itself.
7. The Type of Ground Truth Used
The ground truth used for the clinical validation of the algorithm (Trial IOQC) was outcomes data, specifically:
- Blood glucose values (objective physiological measurements).
- Patient health status, including glycemic control.
- Insulin dosing regimens.
8. The Sample Size for the Training Set
The document does not explicitly mention a "training set" for the algorithm in the context of machine learning. The algorithm's design is stated to be "based on a 'paper-based' algorithm approach that was studied for safety and efficacy in 2 multinational, multicenter, randomized, open-label, parallel-group studies within a single clinical trial, F3Z-MC-IOQC (Trial IOQC) as described in Edelman et al. 2014." This suggests a rule-based algorithm derived from established clinical evidence and refined through traditional clinical trials, rather than a data-driven machine learning approach requiring a distinct training set. The clinical trial itself serves as the validation of the algorithm's effectiveness.
9. How the Ground Truth for the Training Set Was Established
Since it's a rule-based algorithm and not a machine learning model with a distinct "training set," the concept of "ground truth for the training set" doesn't directly apply in the same way. The algorithm's rules and parameters would have been established based on medical knowledge, clinical guidelines, and insights gained from extensive research and prior clinical experience with insulin titration, potentially informed by observational data or previous pilot studies, eventually validated in Trial IOQC. The efficacy and safety were then prospectively evaluated in the aforementioned clinical trial (Trial IOQC) against observed patient blood glucose values and clinical outcomes.
Ask a specific question about this device
Page 1 of 1