(80 days)
The Bio-Rad VARIANT II Hemoglobin A1c Program is intended for the percent determination of hemoglobin A1c in human whole blood using ion-exchange highperformance liquid chromatography (HPLC).
The Bio-Rad VARIANT II Hemoglobin A1c Program is intended for Professional Use Only.
Measurement of percent hemoglobin A1c is effective in monitoring long-term glucose control in individuals with diabetes mellitus.
The VARIANT II Hemoglobin Testing System uses the principles of high performance liquid chromatography (HPLC). The VARIANT II Hemoglobin A1c Program is based on chromatographic separation of Hemoglobin A1c on a cation exchange cartridge.
Here's a breakdown of the acceptance criteria and the study details for the Bio-Rad VARIANT™ II Hemoglobin A1c Program, based on the provided text:
Acceptance Criteria and Reported Device Performance
The provided document describes a substantial equivalence study comparing a new device (VARIANT™ II Hemoglobin A1c Program (270-2101NU)) to a predicate device (VARIANT™ II Hemoglobin A1c Program (270-2101)). Therefore, the acceptance criteria are implicitly defined by the demonstration that the new device performs equivalently to the predicate device across various metrics. The study aims to show that the new device is "substantially equivalent" rather than meeting specific numerical performance targets against a predefined standard.
Table of Acceptance Criteria (Implied by Predicate Equivalence) and Reported Device Performance:
| Performance Metric | Acceptance Criteria (Implied - Equivalence to Predicate) | Reported Device Performance (New Device: 270-2101NU vs. Predicate: 270-2101) |
|---|---|---|
| Accuracy (Method Correlation) | The new device should show comparable correlation to the predicate device across the clinical range of HbA1c values. | Evaluated using 42 EDTA whole blood patient samples (4.2% to 10.9% HbA1c). The document states, "The results are presented in the following regression table." However, the regression table itself is missing, making it impossible to report specific intercept and slope values that would demonstrate correlation. |
| Precision | The new device should exhibit comparable precision (within-run, within-device, total) to the predicate device. | New Device (270-2101NU): |
- Low Patient (n=240, Mean 5.5%): Within run %CV = 0.9, Within Device %CV = 1.60
- High Patient (n=160, Mean 8.8%): Within run %CV = 0.6, Within Device %CV = 1.38
Predicate Device (270-2101): - Low Patient (n=80, Mean 5.4%): Within run %CV = 1.46, Total Precision %CV = 2.14
- High Patient (n=80, Mean 13.7%): Within run %CV = 0.65, Total Precision %CV = 1.68
Conclusion: "the precision results...are equivalent." |
| Linearity (Linear Range) | The linear measuring range of the new device should be comparable to the predicate device. | New Device (270-2101NU): 3.1 - 18.5 % HbA1c
Predicate Device (270-2101): 1.3 - 18.9 % HbA1c
(Interpretation: Ranges are similar, supporting substantial equivalence.) |
| Interfering Substances | The new device should demonstrate similar resistance to common interfering substances at comparable concentrations. | New Device (270-2101NU) & Predicate (270-2101): - Bilirubin: No interference up to 20 mg/dL
- Lipids (Triglycerides): No interference up to 6000 mg/dL
- EDTA: No interference up to 11X EDTA
- Hemoglobin F: 10% (New), 15% (Predicate)
(Interpretation: Generally similar, with a slight difference in HbF tolerance for the new device which isn't explicitly flagged as non-equivalent.) |
Study Details:
-
Sample size used for the test set and the data provenance:
- Accuracy (Method Correlation):
- Sample Size: 42 EDTA whole blood patient samples.
- Data Provenance: Not explicitly stated, but the samples are described as "patient samples," implying human origin. Country of origin and whether retrospective or prospective are not mentioned.
- Precision:
- New Device (270-2101NU): 240 samples for "Low Patient" and 160 samples for "High Patient."
- Predicate Device (270-2101): 80 samples for "Low Patient" and 80 samples for "High Patient."
- Data Provenance: "EDTA whole blood patient samples." Country of origin and whether retrospective or prospective are not mentioned, but the samples were analyzed at "three Bio-Rad sites" for the new device and on "one VARIANT II Hemoglobin Testing System" for the predicate.
- Accuracy (Method Correlation):
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The study is a method correlation and precision study for an in vitro diagnostic device, not an AI or imaging device requiring expert interpretation of results.
- No human experts are mentioned as establishing ground truth in the context of diagnostic performance for this type of device. The ground truth for HbA1c values would be considered the measured value obtained from the method itself (or a reference method if specified).
- The study compares the new device's measurements to the predicate device's measurements, implying the predicate device's results serve as a form of reference for comparison, and the predicate itself is standardized "Traceable to the Diabetes Control and Complications Trial (DCCT) reference method and IFCC. Certified via the National Glycohemoglobin Standardization Program (NGSP)."
-
Adjudication method (e.g. 2+1, 3+1, none) for the test set:
- This concept of "adjudication" is typically relevant for studies where human interpretation or labeling is involved, such as in image analysis or clinical trials with subjective endpoints.
- For this in vitro diagnostic device study, the outputs are quantitative measurements of HbA1c. Therefore, no adjudication method in the traditional sense was used or needed for the test set.
-
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.
- This study is for an in vitro diagnostic (IVD) device, specifically an automated assay, which does not involve human readers interpreting AI outputs. The device directly measures a biomarker (HbA1c).
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Yes, a standalone performance evaluation was done. The entire study assessed the performance of the Bio-Rad VARIANT™ II Hemoglobin A1c Program (an automated analytical instrument) in generating HbA1c measurements, without human intervention in the result generation itself. The "algorithm" here refers to the instrument's internal processes for chromatographic separation and calculation, not an AI algorithm in the contemporary sense.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc):
- The "ground truth" for the performance evaluation is established through comparison to a legally marketed predicate device (Bio-Rad VARIANT™ II Hemoglobin A1c Program (270-2101)).
- Both the new and predicate devices are stated to be "Traceable to the Diabetes Control and Complications Trial (DCCT) reference method and IFCC. Certified via the National Glycohemoglobin Standardization Program (NGSP)." This indicates that the measurements themselves are standardized against established reference methods, which serve as the ultimate "ground truth" for HbA1c values.
-
The sample size for the training set:
- This is a study for an IVD device, not an AI/machine learning model. Therefore, the concept of a "training set" for an algorithm is not applicable. The device's operational parameters are based on its physical and chemical design, calibrated using established standards, not trained on a dataset.
-
How the ground truth for the training set was established:
- As explained above, there is no training set in the AI sense for this device. The device operates based on principles of high-performance liquid chromatography calibrated against reference methods (DCCT, IFCC, NGSP).
§ 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).