(319 days)
The Avie™ Alc test is a point of care system that quantitatively measures %A1C (glycated hemoglobin) in capillary or venous whole blood samples. The test is for prescription use and physician-directed home use to monitor glycemic control in people with diabetes mellitus. The device cannot be used in patients with hemoglobinopathies of HbF, HbC, and HbD.
The Avie A1C test system is for the quantitative measurement of %A1C in fingerstick whole blood at the point-of-care (Professional or Physician Directed Home Use). The system consists of a small instrument ("Reader), a single-use diluent solution vial, single-use reagent test cartridges, and control levels I and II solutions. To perform one test, the Reader is turned on by the push of a button on its top surface, and a self-check (optics and software) is performed automatically. If a malfunction is detected, the Reader displays an error message on the LCD display. When the LCD display denotes "Ready" a test cartridge is inserted into an unambiguous slot in the reader. About four uL of whole blood is delivered into a vial containing wash solution or diluent, via a transfer pipette, and the vial is inverted five times for mixing. At the instruction of the reader (LCD read-out), three drops of blood mixture are applied to a well in the cartridge: the reactions then proceed automatically. There are no further procedural steps, and results in %A1C are displayed on the reader's LCD screen in approximately three minutes. The concentration of A1C is expressed as a percentage of the total hemoglobin as follows: %HBA1C = (HBA1C ÷ TOTAL HB) ×100
Here's an analysis of the acceptance criteria and study information for the Avie™ A1C Test System, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Precision | Percent coefficient of variation within 5% |
| Self: Avie = 0.99 (NGSP) + 0.001 r = 0.94 | |
| Accuracy (estimated bias) | Versus NGSP: Y=0.992 ± 0.001, R=0.94 N=155 (This is from the comparison table against the predicate, likely representing the target for the new device. The clinical study results provide the actual performance.) |
| Professional: Avie = 0.95 (NGSP) + 0.004 r = 0.94 | |
| Self: Avie = 0.99 (NGSP) + 0.001 r = 0.94 | |
| Combined professional vs. self-testing (109 subjects): Y=0.98S+0.003 r=0.94 | |
| Trained vs. Untrained operators (42 subjects): Y=1.05x -0.1899 r²=0.95 | |
| Linearity | Linear between 5 and 14 %A1C |
| Interference | Insensitive to high physiological levels of triglyceride and bilirubin, high therapeutic levels of various over-the-counter pharmaceuticals, and hemoglobin levels ranging from approximately 9 g/dL to 20 g/dL. (However, sensitive to high levels of Metformin and glybenclamide, and high levels of hemoglobinopathies). |
| Operational Condition Limits | Operable between temperatures of 18°C to 28°C |
| NGSP Certification | Certified |
2. Sample Size Used for the Test Set and Data Provenance
- Clinical Study Test Set:
- Approximately 150 subjects for the main clinical study (professional vs. self-testing analysis).
- 109 subjects for the combined professional vs. self-testing sensitivity study.
- 42 subjects for the untrained operator study.
- 60 subjects for the comparison against the A1C Now+ predicate.
- 60 subjects for the comparison of venous to capillary samples.
- Data Provenance: The document does not explicitly state the country of origin for the clinical data. It mentions "four sites" and "three sites" for the clinical studies and "NGSP laboratory in Columbia. MO" for ground truth testing. This suggests the studies were likely conducted within the United States. The studies are described as clinical studies, implying they are prospective as participants actively performed tests.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
- The ground truth for the test set was established by sending collected venous blood samples to an NGSP laboratory in Columbia, MO.
- The document does not specify the number of experts (e.g., lab technicians, certified professionals) within the NGSP laboratory who processed and analyzed these samples.
- The qualifications of these experts are implied by the NGSP (National Glycohemoglobin Standardization Program) certification, which signifies adherence to standardization and quality control for HbA1c testing. Specific individual qualifications (e.g., years of experience as a clinical chemist) are not provided.
4. Adjudication Method for the Test Set
- The document does not describe an explicit "adjudication method" in the context of expert consensus for image or diagnostic interpretation typically found in AI/ML device studies.
- Instead, the ground truth was derived from a single, standardized laboratory method performed by an NGSP-certified laboratory. The device's performance was then compared to this reference standard. Therefore, there was no multi-reader adjudication process as such.
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 done in the context of human readers improving with or without AI assistance.
- This device (Avie™ A1C Test System) is a diagnostic instrument for quantitative measurement of A1C, not an AI or imaging diagnostic aid that assists human readers.
- The studies involved comparing the device's readings to laboratory reference methods and evaluating performance by professional and self-testing users, and trained vs. untrained operators, but not human interpretation aided by AI.
6. If a Standalone (i.e. algorithm only without human-in-the loop performance) Was Done
- Yes, the primary evaluation of the Avie™ A1C Test System's performance (its "algorithm only") was done in a standalone manner against the NGSP reference method. The "Professional" and "Self" testing studies reflect the device's performance when used by different types of operators (humans-in-the-loop), but still measure the device's output against the ground truth.
- The regression statistics (e.g., "Avie = 0.95 (NGSP) + 0.004 r = 0.94") directly reflect the standalone performance of the device's measurement compared to the reference.
7. The Type of Ground Truth Used
- The ground truth used was laboratory reference measurements of %A1C from venous blood samples, performed by an NGSP-certified laboratory. This is a highly standardized and accepted method for determining HbA1c levels.
8. The Sample Size for the Training Set
- The document does not provide information regarding a separate training set. This is typical for a diagnostic device like the Avie™ A1C Test System, which is based on a fixed immunoassay methodology rather than a machine learning or AI model that requires explicit training data. The device is likely calibrated at the factory.
9. How the Ground Truth for the Training Set Was Established
- As there's no mention of a traditional "training set" for an AI/ML model, the concept of establishing ground truth for it is not applicable here. The device's calibration would be established by the manufacturer using a separate, internal process traceable to NGSP standards, but this is distinct from an AI model's training data.
§ 864.7470 Glycosylated hemoglobin assay.
(a)
Identification. A glycosylated hemoglobin assay is a device used to measure the glycosylated hemoglobins (A1a , A1b , and A1c ) in a patient's blood by a column chromatographic procedure. Measurement of glycosylated hemoglobin is used to assess the level of control of a patient's diabetes and to determine the proper insulin dosage for a patient. Elevated levels of glycosylated hemoglobin indicate uncontrolled diabetes in a patient.(b)
Classification. Class II (performance standards).