(94 days)
Afinion™ HbA1c Dx is an in vitro diagnostic test for quantitative determination of glycated hemoglobin (% hemoglobin A1c, HbA1c) in human venous and capillary whole blood.
This test is to be used as an aid in the diagnosis of diabetes and as an aid in identifying patients who may be at risk for developing diabetes.
The measurement of % HbA1c is recommended as a marker of long-term metabolic control in persons with diabetes mellitus.
The Afinion™ HbA1c Dx is a fully automated boronate affinity assay for the determination of the percentage of hemoglobin A1c in human whole blood. The Afinion™ HbA1c Dx is a modification of the existing device, Alere Afinion™ HbA1c for use with the Alere Afinion™ AS100 Analyzer, with the addition of a diagnostic intended use.
The test begins with a blood sample collected with the integrated sampling device before the test cartridge is placed in the cartridge chamber of the Alere Afinion™ AS100 Analyzer. The sample is then automatically diluted and mixed with a solution that releases hemoglobin from the erythrocytes. After the hemoglobin is precipitated, the sample mixture is transferred to a blue boronic acid conjugate which binds to the cis-diols of glycated hemoglobin. This reaction mixture is soaked through a filter membrane and all precipitated hemoglobin, conjugate-bound and unbound (i.e. glycated and non-glycated hemoglobin) remains on the membrane. Excess conjugate is removed with a washing reagent. The analyzer measures the reflectance of the precipitate on the membrane as blue (glycated hemoglobin) and red (total hemoglobin) color intensities. The analyzer calculates a ratio proportional to the percentage of HbA1c in the sample and displays as the % HbA1c (NGSP).
The medical device is the Afinion™ HbA1c Dx, a fully automated boronate affinity assay for the determination of the percentage of hemoglobin A1c in human whole blood. It is an in vitro diagnostic test for the quantitative determination of glycated hemoglobin (% hemoglobin A1c, HbA1c) in human venous and capillary whole blood. Its intended use is as an aid in the diagnosis of diabetes and as an aid in identifying patients who may be at risk for developing diabetes. It is also used as a marker of long-term metabolic control in persons with diabetes mellitus.
1. Table of Acceptance Criteria and Reported Device Performance
| Performance Characteristic | Acceptance Criteria (from 21 CFR 862.1373 Special Controls) | Reported Device Performance |
|---|---|---|
| System Accuracy (Total Error) | Total Error (TE) ≤ 6.0% | Fingerstick Whole Blood Samples: |
| Range: 2.87% to 4.75% TE (using Weighted Deming or Passing-Bablok regressions from bias and precision from 3 studies combined). | ||
| Venous Whole Blood Samples (Internal Precision Study): | ||
| Range: 2.77% to 3.80% TE (using Weighted Deming or Passing-Bablok regressions from bias and internal precision studies). | ||
| Venous Whole Blood Samples (Point of Care, External Precision Study): | ||
| Range: 2.64% to 4.07% TE (using Weighted Deming or Passing-Bablok regressions from bias and external precision studies). |
All reported TE values are ≤ 6.0%. |
| Precision/Reproducibility | Not explicitly stated as a single numerical acceptance criteria in terms of SD or CV, but assessed through detailed studies. | Internal (Venous Whole Blood): Total %CV ranged from 1.32% to 1.74% across HbA1c levels and analyzers.
External (Venous Whole Blood): Total %CV ranged from 1.22% to 1.78% across HbA1c levels.
Fingerstick Samples (Combined Studies): Total %CV ranged from 1.30% to 2.03% across HbA1c levels. |
| Linearity/Reportable Range | No specific numerical acceptance criterion stated, but implicitly expected to cover the medical decision range. | Reportable range: 4.00-15.00 % HbA1c (DCCT/NGSP). Previously established in K050574. |
| Endogenous Interference | Significant interference defined as exceeding a 7% change in %HbA1c value from control. | No significant interference observed for tested substances (Bilirubin, Glucose, Intralipid, Rheumatoid factor, Total protein) at specified concentrations. |
| Hemolysis Interference | Significant interference defined as exceeding a 7% change in %HbA1c value from control. | No significant interference observed up to 24% hemolysis. Information codes related to hemolysis may occur above 14% hemolysis. |
| Drug Interference | Significant interference defined as exceeding a 7% change in %HbA1c value from control. | No significant interference observed for 20 tested drugs at specified concentrations. |
| Cross-reactivity with Hemoglobin Derivatives | Significant interference defined as exceeding a 7% change in %HbA1c value from control. | No significant interference observed for Acetylated hemoglobin, Carbamylated hemoglobin, Labile HbA1c, and Glycated albumin at specified concentrations. |
| Hemoglobin Variants Interference | Significant interference defined as exceeding a 7% change in %HbA1c value from reference method. | No significant interference for HbA2, HbS, HbC, HbE, HbD. Significant negative interference with HbF (highest concentration with no significant interference at 10.4% HbF). Device includes a prominent boxed warning for HbF. |
2. Sample Sizes Used for the Test Set and Data Provenance
- Accuracy/Method Comparison Study (Test Set):
- Sample Size: 120 subjects.
- Data Provenance: Samples from each study subject were tested with both fingerstick samples and fresh venous EDTA samples. Divided across three study sites. Retrospective or prospective is not explicitly stated, but the collection of fresh venous and fingerstick samples suggests it was prospective. The country of origin of the data is not specified.
- Precision (Internal, Venous Whole Blood): 4 levels of HbA1c patient samples. Each level tested with 2 replicates, twice a day for 20 days with 3 lots on 3 analyzers. (240 measurements per analyzer, 720 combined).
- Precision (External, Venous Whole Blood): 4 levels of HbA1c patient samples. 4 replicates analyzed twice a day for 10 days with 3 lots at each of 3 sites, using 3 analyzers.
- Precision (Fingerstick Samples):
- Study A (Accuracy study - within-run): 172 subjects (fingerstick samples in duplicate).
- Study C (Between-instrument and between-operator): 15-16 subjects per HbA1c level (total of 4 levels) across 3 sites. Each subject gave 6 fingerstick samples. Total of 90-96 fingerstick measurements per level.
- Between-Instrument Precision: 4 venous whole blood samples measured in 6 replicates on each of 14 analyzers with 1 test cartridge lot (total of 84 replicates per sample).
- Lot-to-Lot Variation: 18 EDTA venous whole blood samples spanning the reportable range. Each sample analyzed in 1 replicate with each of 3 test cartridge lots on the same analyzer.
- Endogenous/Hemolysis/Drug Interference, Hemoglobin Derivatives Cross-reactivity: Whole blood sample pools. Specific number of samples not given, but tested with 10 replicates for each condition/substance.
- Hemoglobin Variants Interference: 234 fresh EDTA whole blood samples containing 6 common hemoglobin variants. Also, 100 samples for HbA0, HbA1a, HbA1b components.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of those Experts
- For the Method Comparison study, the ground truth was established by an NGSP secondary reference laboratory method (Tosoh Glycohemoglobin test on the G8 HPLC analyzer). This method itself serves as the "expert" or gold standard. The document does not specify the number or qualifications of human experts involved in operating or verifying this reference method.
4. Adjudication Method for the Test Set
- The document implies that the reference method (Tosoh G8 HPLC) was considered the definitive ground truth for the method comparison study. There is no mention of a human adjudication method (like 2+1 or 3+1 consensus) for the test set results against the reference method. The comparison was statistical, using Deming and Passing-Bablok regressions.
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
- This device is an automated in vitro diagnostic test (HbA1c assay). It does not appear to involve "human readers" in the sense of image interpretation or other judgmental tasks where AI assistance might improve their performance. Therefore, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study comparing human readers with and without AI assistance is not applicable to this device.
6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was Done
- Yes, the performance studies described are for the "algorithm only" (device only) performance. The Afinion™ HbA1c Dx is a fully automated system for determining HbA1c. The performance metrics (precision, accuracy, interference, etc.) reflect the standalone operation of the device.
7. The Type of Ground Truth Used
- The ground truth used for the method comparison and hemoglobin variant interference studies was a reference method, specifically an NGSP secondary reference laboratory method (Tosoh Glycohemoglobin test on the G8 HPLC analyzer). For interference studies, the ground truth was derived from non-spiked control samples or reference samples without the interfering substance.
8. The Sample Size for the Training Set
- The document describes performance characteristics for the Afinion™ HbA1c Dx. It does not provide information on a separate "training set" for an algorithm, as this typically applies to machine learning models. Instead, the device is a chemical assay with established analytical principles. Therefore, a distinct "training set" size in the context of an AI/ML algorithm is not applicable as presented in this document. The "training" or development would refer more to the optimization and validation of the analytical method itself.
9. How the Ground Truth for the Training Set Was Established
- As a chemical assay rather than an AI/ML algorithm, the concept of a "training set" with established ground truth is not applicable in the same way. The device's analytical method (boronate affinity assay) is based on scientific principles and validated through extensive performance testing as detailed in the document, against established reference methods.
§ 862.1373 Hemoglobin A1c test system.
(a)
Identification. A hemoglobin A1c test system is a device used to measure the percentage concentration of hemoglobin A1c in blood. Measurement of hemoglobin A1c is used as an aid in the diagnosis of diabetes mellitus and as an aid in the identification of patients at risk for developing diabetes mellitus.(b)
Classification. Class II (special controls). The special controls for this device are:(1) The device must have initial and annual standardization verification by a certifying glycohemoglobin standardization organization deemed acceptable by FDA.
(2) The premarket notification submission must include performance testing to evaluate precision, accuracy, linearity, and interference, including the following:
(i) Performance testing of device precision must, at a minimum, use blood samples with concentrations near 5.0 percent, 6.5 percent, 8.0 percent, and 12 percent hemoglobin A1c. This testing must evaluate precision over a minimum of 20 days using at least three lots of the device and three instruments, as applicable.
(ii) Performance testing of device accuracy must include a minimum of 120 blood samples that span the measuring interval of the device and compare results of the new device to results of a standardized test method. Results must demonstrate little or no bias versus the standardized method.
(iii) Total error of the new device must be evaluated using single measurements by the new device compared to results of the standardized test method, and this evaluation must demonstrate a total error less than or equal to 6 percent.
(iv) Performance testing must demonstrate that there is little to no interference from common hemoglobin variants, including Hemoglobin C, Hemoglobin D, Hemoglobin E, Hemoglobin A2, and Hemoglobin S.
(3) When assay interference from Hemoglobin F or interference with other hemoglobin variants with low frequency in the population is observed, a warning statement must be placed in a black box and must appear in all labeling material for these devices describing the interference and any affected populations.