(114 days)
The HbA 1c (Hemoglobin A 1c) Advanced assay on the Beckman Coulter DxC700 AU Clinical Chemistry Analyzer, is intended for the quantitative determination of mmol/mol HbA1c (DCCT/NGSP) concentration in human venous whole blood. The determination of HbA1c is used as an aid in the diagnosis of diabetes mellitus, for the monitoring of long-term glucose control in individuals with diabetes mellitus and identifying patients who may be at risk for developing diabetes mellitus. For in vitro diagnostic use only.
The HbA1c Advanced reagent kit is in a liquid format and is ready to use. It contains four reagents HbA1c R1 and HbA1c R2, Total Hemoglobin R1 and Hemolyzing reagent R1. The HbA1c calibrator is supplied with the reagent, in a liquid, ready to use format and contains 5 x 2mL calibrator levels. The sample hemolysis is automated on the DxC700 AU Clinical Chemistry analyzer. Sample handling is performed as follows: 200 uL of hemolyzing reagent is aspirated from the Hemolyzing Reagent R1and dispensed into a cuvette. Tetradecyltrimethylammonium bromide (TTAB) in the hemolyzing reagent eliminates interference from leukocytes. 2 µL of whole blood sample is then aspirated from the patient sample and added to the hemolyzing reagent in the cuvette. This hemolyzed whole blood is then added to the THb assay cuvette and HbA1c assay cuvette as per the assay parameters. The concentrations of both HbA1c and Total Hemoglobin are determined. The HbA1c/Total Hemoglobin ratio is expressed either as mmol/mol (IFCC) or %HbA1c (DCCT/NGSP). Total Hemoglobin Reagent is used to measure total hemoglobin concentration by a colorimetric method. Change in absorbance is measured at 570/660 nm. HbA1c reagent is used to measure hemoglobin A1c concentration by a turbidimetric immunoinhibition method. In the reaction, hemoglobin A1c antibodies combine with HbA1c from the sample to form soluble antigen-antibody complexes. Polyhaptens from the reagent then bind with the excess antibodies and the resulting agglutinated complex is measured turbidimetrically. Change in absorbance is measured at 340/700 nm.
Here's an analysis of the acceptance criteria and study details for the HbA1c Advanced device, based on the provided document:
Acceptance Criteria and Device Performance
| Parameter | Acceptance Criteria | Reported Device Performance (NGSP Units) | Pass/Fail |
|---|---|---|---|
| Precision | |||
| Repeatability (Within-run) | CV ≤1.5% or SD ≤0.1% HbA1c | Human Whole Blood 1 (5.06%): 0.88% CV, 0.04 SD | |
| Human Whole Blood 2 (6.72%): 1.01% CV, 0.07 SD | |||
| Human Whole Blood 3 (8.06%): 0.77% CV, 0.06 SD | |||
| Human Whole Blood 4 (11.70%): 0.79% CV, 0.09 SD | |||
| Spiked Human Whole Blood (14.02%): 0.74% CV, 0.10 SD | |||
| Whole Blood Control 1 (5.32%): 1.19% CV, 0.06 SD | |||
| Whole Blood Control 2 (9.88%): 0.77% CV, 0.08 SD | Pass | ||
| Total Precision | CV ≤2% or SD ≤0.13% HbA1c | Human Whole Blood 1 (5.06%): 1.63% CV, 0.08 SD | |
| Human Whole Blood 2 (6.72%): 1.64% CV, 0.11 SD | |||
| Human Whole Blood 3 (8.06%): 1.57% CV, 0.13 SD | |||
| Human Whole Blood 4 (11.70%): 1.26% CV, 0.15 SD | |||
| Spiked Human Whole Blood (14.02%): 1.19% CV, 0.17 SD | |||
| Whole Blood Control 1 (5.32%): 2.08% CV, 0.11 SD | |||
| Whole Blood Control 2 (9.88%): 1.54% CV, 0.15 SD | Pass | ||
| Linearity (NGSP) | |||
| Linear Range | 4-15% HbA1c | 3.94% HbA1c to 15.37% HbA1c | Pass |
| Regression Parameters | Slope: 1.0 ± 0.05; Intercept: ≤ ± 0.5 % HbA1c; R: ≥ 0.990; N: ≥ 9 | Slope: 1.0 ± 0.05; Intercept: ≤ ± 0.5 % HbA1c; R: ≥ 0.990; N: ≥ 9 (All met) | Pass |
| Method Comparison (NGSP) | |||
| Slope (Weighted Deming) | 1.0 ± 0.05 | 0.990 (0.978; 1.002) | Pass |
| Intercept (Weighted Deming) | ≤ ± 0.5% HbA1c | 0.010 %HbA1c (-0.070; 0.089) %HbA1c | Pass |
| R (Weighted Deming) | ≥ 0.975 | 0.998 | Pass |
| Slope (Passing-Bablok) | 1.0 ± 0.5 | 0.980 (0.964; 0.992) | Pass |
| Intercept (Passing-Bablok) | ≤ ± 0.5% HbA1c | 0.090 %HbA1c (-0.006; 0.187) % HbA1c | Pass |
| R (Passing-Bablok) | ≥ 0.975 | 0.998 | Pass |
| Total Error | ≤6% | 5.0% HbA1c: 4.3% | |
| 6.5% HbA1c: 4.2% | |||
| 8.0% HbA1c: 4.3% | |||
| 12.0% HbA1c: 3.3% | Pass | ||
| Analytical Specificity | No Significant Interference (recovery within 7% of initial value) | Endogenous Interference: No significant interference up to stated concentrations for Conjugated Bilirubin (60 mg/dL), Unconjugated Bilirubin (60 mg/dL), Lipemia (500 mg/dL), Ascorbic Acid (300 mg/dL), RF (1000 IU/ml), Total Protein (21 g/dL), Glucose (2000 mg/dL). | |
| Drug Interference: No significant interference up to stated concentrations for numerous drugs (e.g., Glyburide 0.12 mg/dL, Salicylic Acid 4.76 mg/dL, Acetaminophen 26 mg/dL, etc.). | |||
| Hemoglobin Derivative and Cross Reactants: No significant interference up to stated concentrations for Labile Hemoglobin (2000 mg/dL), Acetylated Hemoglobin (0.5 mg/mL), Carbamylated Hemoglobin (1.5 mg/mL), Glycated Albumin (5mg/mL), HbA0 (12 mg/mL), HbA1a + 1b (0.16 mg/mL). | Pass | ||
| Hemoglobin Variants (Bias) | No Significant Interference (recovery within 7% of reference value) | HbC: -2.57% bias (range -4.30% to -1.80%) at ~6.5% HbA1c; -3.19% bias (range -6.48% to 0.41%) at ~9.0% HbA1c. HbD: -0.77% bias (range -4.81% to 2.99%) at ~6.5% HbA1c; -1.22% bias (range -6.30% to -0.22%) at ~9.0% HbA1c. HbE: -1.12% bias (range -9.16% to 2.48%) at ~6.5% HbA1c; 0.47% bias (range -1.76% to 4.21%) at ~9.0% HbA1c. HbS: -1.18% bias (range -2.17% to 3.04%) at ~6.5% HbA1c; -1.04% bias (range -3.33% to 4.41%) at ~9.0% HbA1c. HbA2: 0.48% bias (range -1.92% to 5.60%) at ~6.5% HbA1c; 2.49% bias (range -0.98% to 3.60%) at ~9.0% HbA1c. | |
| HbF: Specimens containing >7% HbF may yield lower than expected HbA1c values (disclaimer). | Pass (with disclaimer for HbF) |
Study Details
-
Sample size used for the test set and the data provenance:
- Precision (EP05-A3): Four levels of HbA1c K2 EDTA human venous whole blood patient samples (approx. 5.0%, 6.5%, 8.0%, 12%, and 14%) and two whole blood controls. Samples were analyzed in duplicate, twice daily, over 20 working days on 3 different instrument lots. (n=2 for each sample/control, 2x daily, 20 days: 80 measurements per sample/control per instrument, total for all samples/controls/instruments: 8073 = 1680 individual measurements for precision)
- Linearity (EP06-A): High and low pools of human whole blood were used to create a linearity series to span the analytical range. The exact number of samples in the linearity series is implied to be at least 9 (N: ≥ 9 for regression parameters).
- Method Comparison (EP09-A3): 138 venous human frozen whole blood specimens (K2 EDTA anticoagulant type). These were patient samples. The provenance is not explicitly stated as country of origin, but they are referred to as "venous human frozen whole blood specimens." The study design is prospective in the sense that the samples were collected and then tested for the study; it is not explicitly called out as retrospective/prospective.
- Analytical Specificity (Interference - EP07): Two % HbA1c concentrations (approx. 6.5% and 8.0% HbA1c) using low and high pools prepared from human whole blood. Interfering substances were tested at a minimum of 5 levels each, with 10 replicates per level. The interference assessment was done on human venous whole blood K2 EDTA samples.
- Hemoglobin Variants: A minimum of 20 samples were tested for each variant (HbC, HbD, HbE, HbF, HbS, HbA2), totaling at least 120 samples. These are patient samples with identified hemoglobin variants.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- For the Method Comparison study, the ground truth was established by an "NGSP Secondary Reference Laboratory (SRL) using a test system (method X - HA8180V HPLC)." The specific number and qualifications of experts at the SRL are not detailed in this document.
- For Hemoglobin Variants, the reference methods used to establish the ground truth were "Trinity Biotech Hb9210 and Ultra2, Menarini HA8181V and TOSOH G8," which are demonstrated to be free from hemoglobin interference. Again, the number and qualifications of the operators of these reference methods are not specified.
-
Adjudication method (e.g. 2+1, 3+1, none) for the test set:
- The document does not describe an adjudication method involving multiple human readers for establishing ground truth. The ground truth for method comparison and hemoglobin variant studies relies on reference methods/laboratories.
-
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 MRMC comparative effectiveness study involving human readers or AI assistance was conducted or described, as this device is an in-vitro diagnostic (IVD) test system (HbA1c assay) and not an imaging AI device that would typically involve human reader interpretation.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Yes, the performance characteristics described are for the "HbA1c Advanced reagent on the DxC 700 AU Clinical Chemistry Analyzer" as a standalone device (algorithm only performance, in a laboratory setting). The entire submission details the analytical performance of the device itself.
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- Precision: Internal validation against the device's own specifications.
- Linearity: Internal validation for demonstrating accurate measurement across specific concentration ranges.
- Method Comparison: Comparison to an NGSP Secondary Reference Laboratory (SRL) standardized method (HA8180V HPLC).
- Analytical Specificity (Interference): Comparison to baseline measurements of samples without interferents, assessed against a 7% recovery criteria.
- Hemoglobin Variants: Comparison to reference methods demonstrated to be free from hemoglobin interference (Trinity Biotech Hb9210 and Ultra2, Menarini HA8181V and TOSOH G8).
-
The sample size for the training set:
- This document describes the analytical validation of a biochemical assay on an automated analyzer. There is no "training set" in the context of machine learning, as the device is a reagent and instrument system, not an AI/ML algorithm that is trained on data. The device's performance is rigorously tested as described above.
-
How the ground truth for the training set was established:
- Not applicable, as this is not an AI/ML device with a training set. The assay's analytical parameters are established through chemical and immunochemical principles and optimized during development.
§ 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.