(77 days)
The MULTIGENT™ Hb A1c assay is used in clinical laboratories for the quantitative in vitro measurement of percent Hb Alc (hemoglobin fraction) in human whole blood on the AEROSET® System and ARCHITECT® c8000™ System. The Hb A1c assay is intended to aid in the monitoring of long-term blood glucose control and compliance in individuals with diabetes mellitus. The MULTIGENT™ Hb A1c assay is not intended for use in diagnosing diabetes mellitus.
The MULTIGENT™ Ho A1c Calibrators are intended for in vitro diagnostic use with the AEROSET® System and the ARCHITECT® c8000™ System for the calibration of the assays in the MULTIGENT™ Hb Alc Reagent Kit.
The MULTIGENT™ Ho Alc Controls are intended for in vitro diagnostic use with the AEROSET® System and the ARCHITECT® c8000™ System for quality control of the assays in the MULTIGENTTM Hb A1c Reagent Kit.
The Device consists of the MULTIGENT™ Hemoglobin A 1 c Reagents, MULTIGENT™ Hemoglobin Alc Calibrators , and MULTIGENT™ Hemoglobin A 1 c Controls, intended for use on AEROSET® System and ARCHITECT® c8000™ System for determination of stable % HbAlc.
The assay consists of two separate concentration measurements, the stable form of glycated hemoglobin (Hb A 1 c) and the total hemoglobin (THb), which are used only to determine the percent Hb Alc. and must not be used individually for diagnostic purposes.
The whole blood specimen is pre-treated to lyse the erythrocytes. The hemoglobin is degraded by the proteolytic enzyme, pepsin, to form a hemolysate. Both the THb and the Hb A1c concentrations are determined from the same hemolysate.
The concentration of total hemoglobin is determined colorimetrically using a wavelength of 604 mm. The sample's measured absorbance is compared to a two-point calibration curve for total hemoglobin.
The concentration of stable Hb Alc is measured immunoturbidimetrically using a microparticle agglutination inhibition method. The Hb A1c antibody reagent (R1) contains specific anti-Hb A 1c mouse monoclonal antibodies coupled to microparticles. The Hb A1c agglutinator reagent (R2) contains several copies of the immunoreactive portion of Hb A1c (hapten), covalently bound to a polymer.
In the absence of Hb Alc in the sample, the hapten in the R2 reagent binds with the antibodycoated microparticles in the R1 antibody reagent and results in an increase in the rate of agglutination and results in an increase in measured absorbance. In the presence of HD A Ic in the sample, the Hb Alc competes with the hapten in the R2 reagent for binding sites on the antibody-coated microparticles in the R1 antibody reagent and will slow the rate of agglutination as it competes with the Hb Alc agglutinator for antibody binding sites.
The increase in concentration of Hb Alc in the sample is inversely proportional to the rate of agglutination and the measured absorbance. The absorbance is measured using a wavelength of 700 nm. The measured absorbance of the sample is compared to the measured absorbance of known Hb Alc concentrations (g/dL) of a six-level calibration curve, and the concentration of the sample is interpolated. The percent Hb A1c is the Hb A1c /THb ratio, calculated automatically by the AEROSET® System and ARCHITECT® c8000™ System, using a conversion factor to correlate the result with an NGSP-certified method.
The calibrators are supplied in liquid form and are ready to use without pretreatment. The controls are supplied in lyophilized form and are to be reconstituted with the supplied reconstitution fluid.
Here's a breakdown of the acceptance criteria and the study details for the Seradyn MULTIGENT™ Hemoglobin A1c device, based on the provided 510(k) summary:
Acceptance Criteria and Reported Device Performance
| Acceptance Criteria Category | Acceptance Criteria (from Predicate Device Label Claims) | Reported Device Performance (MULTIGENT™ HbA1c Assay) |
|---|---|---|
| Linearity (% HbA1c) | 4.2% to 20.8% | 2% to 20% |
| Specificity/Interfering Substances | ||
| Bilirubin | Difference of ≤ 1% Hb A1c (for MULTIGENT) / ± 10% of untreated sample (for Tosoh) | No interference (≤ 1% Hb A1c difference from untreated sample) at 50 mg/dL |
| Triglyceride | Difference of ≤ 1% Hb A1c (for MULTIGENT) / ± 10% of untreated sample (for Tosoh) | No interference (≤ 1% Hb A1c difference from untreated sample) at 1600 mg/dL |
| Rheumatoid Factor | None reported for Tosoh | No interference (≤ 1% Hb A1c difference from untreated sample) at 3100 U/mL |
| Acetyl Salicylate | None reported for Tosoh | No interference (≤ 1% Hb A1c difference from untreated sample) at 50.8 mg/dL |
| Sodium Cyanate | Difference of ≤ 1% Hb A1c (for MULTIGENT) / ± 10% of untreated sample (for Tosoh) | No interference (≤ 1% Hb A1c difference from untreated sample) at 50 mg/dL |
| Ascorbic Acid | None reported for Tosoh | No interference (≤ 1% Hb A1c difference from untreated sample) at 50 mg/dL |
| Urea (Carbamyl GHb) | None reported for Tosoh | No interference (≤ 1% Hb A1c difference from untreated sample) at 667 mg/dL |
| Gamma Globulin | None reported for Tosoh | No interference (≤ 1% Hb A1c difference from untreated sample) at 5 g/dL |
| HAMA Type 1 | None reported for Tosoh | No interference (≤ 1% Hb A1c difference from untreated sample) at 100% replaced plasma |
| HAMA Type 2 | None reported for Tosoh | No interference (≤ 1% Hb A1c difference from untreated sample) at 100% replaced plasma |
| Labile Hb A1c | Separates LA1c (for Tosoh) | No interference (≤ 1% Hb A1c difference from untreated sample) at 14 mg/mL of glucose |
| Ala | Chromatographically Separates out Ala (for Tosoh) | No interference (≤ 1% Hb A1c difference from untreated sample) at 2.85% |
| Alb | Chromatographically Separates out Alb (for Tosoh) | No interference (≤ 1% Hb A1c difference from untreated sample) at 1.25% |
| Precision (Mean ~5.0% HbA1c) | ||
| - Within Run CV | 0.90 %CV | 1.17 %CV |
| - Between Run CV | 0.40 %CV | 0.40 %CV |
| - Total CV | 1.12 %CV | 1.46 %CV |
| Precision (Mean ~10.5% HbA1c) | ||
| - Within Run CV | 0.53 %CV (for 10.9% HbA1c) | 1.07 %CV |
| - Between Run CV | 0.46 %CV (for 10.9% HbA1c) | 0.73 %CV |
| - Total CV | 0.71 %CV (for 10.9% HbA1c) | 1.31 %CV |
| Method Comparison (Correlation) | Implied high correlation (predicate is marketed) | Multiple R (Correlation Coef.): 0.993 (Aeroset) & 0.994 (c8000) |
Study Details
-
Sample Size Used for the Test Set and Data Provenance:
- Specificity and Interfering Substances: Specific concentrations of various interfering substances were used. The number of samples tested per substance is not explicitly stated, but the tests were performed "with the MULTIGENT™ Hb A1c assay." The provenance is not specified but is implicitly from laboratory testing.
- Precision: "Whole blood samples" were used. The sample size for precision studies was n=80 (following NCCLS EP5-A protocol) for each of two concentration levels on both the AEROSET® System and ARCHITECT® c8000™ System. Data provenance is implicitly from laboratory testing, most likely in the USA where Seradyn Inc. is located.
- Method Comparison: 117 "whole blood patient samples" were used. The provenance is not specified but is implicitly from laboratory testing, most likely in the USA.
-
Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts:
- This is a submission for an in-vitro diagnostic device (reagents and controls), not an imaging or diagnostic algorithm that relies on expert interpretation. Therefore, the concept of "experts establishing ground truth" in the traditional sense of clinical interpretations (e.g., radiologists) is not directly applicable.
- The "ground truth" for the comparative studies is established by the predicate device, the Tosoh Medics Inc., G7 Automated HPLC Analyzer: HbA1c Variant Analysis Mode. This predicate device is a legally marketed device and its results are considered the reference for comparison.
-
Adjudication Method for the Test Set:
- Not applicable. The study is a comparative analysis between the new device and a predicate device, and laboratory performance metrics. Human adjudication of results is not described or relevant for this type of device submission.
-
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 type of study is relevant for diagnostic imaging or interpretation by human readers, where AI might assist in improving their performance. This submission is for an in-vitro diagnostic instrument system, which automates the measurement process.
-
If a standalone (i.e. algorithm only without human-in-the loop performance) was done:
- Yes, the performance data presented (linearity, specificity, precision, method comparison) reflects the standalone performance of the MULTIGENT™ Hb A1c assay on the AEROSET® System and ARCHITECT® c8000™ System. While human operators are involved in running the assay and handling samples, the measurement itself is automated by the device, and the reported performance characteristics are inherent to the assay and instrument.
-
The type of ground truth used:
- For the comparative analysis, the results obtained from the legally marketed predicate device (Tosoh G7 Automated HPLC - Hb A1c Variant Analysis Mode) served as the ground truth or "reference method" for demonstrating substantial equivalence.
- For other performance characteristics like linearity and precision, the ground truth is established through internal scientific validation against known standards and established protocols (e.g., NCCLS EP5-A).
-
The sample size for the training set:
- The document does not explicitly mention a "training set" in the context of machine learning. For an in-vitro diagnostic device of this nature, calibration curves are established using known concentrations of HbA1c calibrators, and internal validation is performed. The number of samples for developing these internal parameters is not specified, but the calibrators themselves are a "six-level calibration curve" (page 2).
-
How the ground truth for the training set was established:
- Again, a "training set" in the AI/ML sense is not directly applicable. The "ground truth" for establishing the calibration and internal performance characteristics is based on:
- Known concentrations of stable HbA1c in the MULTIGENT™ Hb A1c Calibrators. These calibrators are provided in liquid form and are used to create the six-level calibration curve for the assay.
- Standardized protocols and materials for internal validation studies, such as the NCCLS EP5-A protocol for precision.
- Again, a "training set" in the AI/ML sense is not directly applicable. The "ground truth" for establishing the calibration and internal performance characteristics is based on:
§ 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).