(294 days)
The cobas c 513 clinical chemistry analyzer is a fully automated, standalone clinical chemistry analyzer intended for the in-vitro quantitative determination of analytes in body fluids.
The Tina-quant Hemoglobin A1cDx Gen.3 assay is intended for use as an aid in diagnosis of diabetes and as an aid in identifying patients who may be at risk for developing diabetes. It is an in vitro diagnostics reagent system intended for quantitative determination of mmol/mol hemoglobin A1c (IFCC) and % hemoglobin A1c (DCCT/NGSP) in hemolysate or whole blood on the Roche/Hitachi cobas c 513 clinical chemistry analyzer. HbA1c determinations are useful for monitoring of long-term blood glucose control in individuals with diabetes mellitus.
The cobas c 513 clinical chemistry analyzer is a fully automated, software controlled analyzer system for in vitro quantitative determination of analytes in human body fluids. The cobas c 513 analyzer system includes a control unit and an analyzer with a closed tube sampling functionality.
The cobas c 513 control unit includes a computer (PC) located in the sampler unit, a touchscreen monitor, soft-keyboard, a mouse (optional use), and a printer. The control unit is used to perform tasks on the analyzer and the PC runs the software that controls the analyzer.
The software manages all instrument functions, all system functions, and all information related to orders and results. The software offers a graphical user interface (GUI) to control all functions by the operator. The control unit contains System Control software including interfaces to a Medical Device Data System (cobas Link) and to the customer Laboratory Information System (LIS).
The analyzer is composed of the sampler unit and the analytical unit. The sampler unit is composed of the rack loading/unloading areas, a barcode reader, a rack rotor, a STAT port, and conveyor lines. It is used to load and unload racks, power on the system, access the PC, and manage the conveyance of samples to the analytical unit.
The analytical unit is comprised of the reagent area, the sample area, and the reaction disk. Samples are conveyed to the analytical unit from the sample unit for photometric analysis before being returned to the sample unit.
Anticoagulated whole blood is hemolyzed either manually or automatically prior to determination of HbAlc by a turbidimetric inhibition immunoassay. Liberated hemoglobin (Hb) in the hemolyzed sample is converted to a derivative having a characteristic absorption spectrum and measured bichromatically. The instrument calculates the % HbAlc from the HbAlc/Hb ratio according to a user selected protocol, either IFCC or NGSP protocols.
The assay offers separate applications that are specific to the sample types whole blood and hemolysate. The Whole Blood Application differs from the Hemolysate Application in the hemolyzing step. For the Whole Blood Application, whole blood samples are placed on the analyzer. Hemolysis occurs onboard the analyzer. For the Hemolysate Application, hemolyzed samples are placed on the analyzer. Hemolysis occurs manually before placing the samples onboard the analyzer. The two applications yield the same results.
This document, a 510(k) premarket notification, describes the performance evaluation of the cobas c 513 Analyzer and cobas c 513 Tina-quant HbA1cDx Gen.3 Assay. It does not describe an AI/ML powered device, but rather an in-vitro diagnostic system for measuring Hemoglobin A1c. Therefore, many of the typical acceptance criteria and study designs relevant to AI/ML based devices (e.g., expert consensus, MRMC studies, training set details) are not applicable or detailed in the same way.
However, I can extract and present the information as per your request based on the provided document, interpreting "acceptance criteria" as the performance criteria the device was tested against and "study" as the non-clinical performance evaluation conducted.
Device Name: cobas c 513 Tina-quant HbA1cDx Gen.3 Assay
Device Type: In-vitro diagnostic reagent system for quantitative determination of HbA1c.
1. Table of Acceptance Criteria and Reported Device Performance
Given that this is a 510(k) for an in-vitro diagnostic (IVD) device, the "acceptance criteria" are not explicitly listed in a single table with pass/fail remarks as might be seen for an AI/ML algorithm. Instead, the document details various performance studies (e.g., precision, linearity, method comparison, interference) and their results, implicitly demonstrating that the device meets an acceptable level of performance for its intended use based on standard IVD guidelines.
The table below summarizes the key performance metrics and their reported values. The "Acceptance Criteria" column reflects typical expected performance ranges or goals for such IVDs based on general industry standards and the context provided by the results themselves. The document states "All acceptance criteria for method comparison were met" and "All acceptance criteria for drug interferences were met," indicating that internal predefined criteria were satisfied.
| Performance Metric | Acceptance Criteria (Implicit/Inferred from Context) | Reported Device Performance (cobas c 513 Tina-quant HbA1cDx Gen.3 Assay) |
|---|---|---|
| Limit of Blank (LoB) | Expected to be very low, indicating minimal signal in the absence of analyte. | Hb: 0.50 g/dL (0.31 mmol/L) |
| HbA1c: 0.19 g/dL (0.12 mmol/L) | ||
| Limit of Detection (LoD) | Expected to be low, indicating the lowest concentration reliably detected. | Hb: 1.00 g/dL (0.62 mmol/L) |
| HbA1c: 0.29 g/dL (0.18 mmol/L) | ||
| Precision (Total %CV) (Hemolysate) | Expected to be low across the measuring range (e.g., typically 7% deviation in HbA1c) at specified maximum concentrations. | Met criteria; specific maximum concentrations without interference provided for: Lipemia, Bilirubin, Ditaurobilirubin, Glucose, Rheumatoid Factor |
| Drug Interferences | No significant interference (defined as > ±7% deviation) at specified drug concentrations. | Met criteria for all 16 tested drugs. |
| Cross-Reactivity | No significant interference due to common hemoglobin variants and other related substances. | No significant interference with tested Hb variants (HbS, HbC, HbE, HbD, HbA2) unless otherwise specified. HbF > 7% may yield lower than expected values. |
2. Sample Sizes Used for the Test Set and Data Provenance
- Limit of Blank (LoB): One analyte-free sample measured 10-fold with three lots, over six runs distributed over >3 days on one c 513 analyzer. Total 60 measurements per lot.
- Limit of Detection (LoD): Five samples with low-analyte concentration measured 2-fold with three lots, over six runs distributed over >3 days on one c 513 analyzer. Total 60 measurements per lot.
- Precision: Eight samples (two controls, seven human samples with varying HbA1c concentrations) measured in two aliquots per sample, once each in two runs per day for 21 days, on three cobas c 513 analyzers using 3 reagent lots per system.
- Method Comparison: 155 samples (Hemolysate) and 154 samples (Whole Blood) from the secondary NGSP reference laboratory. Samples were tested over a 3-day period with one lot of reagent on one cobas c 513 analyzer.
- Linearity: Two separate dilution series consisting of at least 11 levels, using human hemolysate sample pools. Samples measured in triplicate.
- Matrix Comparison: At least 40 samples of each sample type (anticoagulant and fill volume) and at least 40 half-filled tubes of each sample type. Overall sample count not explicitly summed but implies a substantial number.
- Endogenous Interference: 12 spiked samples (2 HbA1c levels x 6 interferents). Each treated pool (spiked and control) had a dilution series of at least 10 levels. Dilution series tested 10-fold.
- Drug Interferences: Native patient samples at 2 different HbA1c levels (approx. 6% and 8% HbA1c) were spiked with 16 common drugs at two defined concentrations. Samples measured in 10-fold.
- Cross-reactivity: Multiple experiments for each cross-reactant (HbA0, HbA1a+b, Acetylated Hb, Carbamylated Hb, Glycated Albumin, Labile HbA1c). Each involved a series of dilutions or concentration variations, with 10 replicates per sample.
- Hemoglobin Variants: 20 samples each for HbS, HbC, HbE, HbD, Elevated F. 13 samples for HbA2. Each tested once in at least one run.
Data Provenance: The document does not explicitly state the country of origin for the human samples used in these studies. The studies were non-clinical performance evaluations (laboratory-based testing). The samples for method comparison were obtained from a secondary NGSP (National Glycohemoglobin Standardization Program) reference laboratory. This implies a standardized and presumably well-characterized dataset. The studies are prospective in the sense that they are designed performance evaluations conducted specifically for device submission, rather than analysis of existing stored data.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications
This document describes a laboratory diagnostic device, not an AI/ML algorithm that interprets medical images or other complex data requiring human expert consensus for ground truth.
- Ground truth for HbA1c values: For the method comparison study, the ground truth was established by a secondary NGSP reference laboratory using a Tosoh HPLC system. This is a recognized and standardized reference method for measuring HbA1c. The document mentions "Traceability: This method has been standardized against the approved IFCC reference method for the measurement of HbA1c in human blood and can be transferred to results traceable to DCCT/NGSP by calculation." This indicates that the ground truth is established via an internationally recognized and highly accurate analytical method, rather than through expert human assessment.
Therefore, the concept of "number of experts" and their "qualifications" for establishing ground truth in the context of an AI/ML diagnostic for image interpretation is not directly applicable here.
4. Adjudication Method for the Test Set
Not applicable. As described above, the ground truth for this IVD device is established through a laboratory reference method (Tosoh HPLC system) traceable to international standards (IFCC, NGSP), not through human expert assessment needing adjudication.
5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study Was Done
No. This document describes the performance of an in-vitro diagnostic instrument and assay, not an AI/ML application intended to assist human readers (e.g., radiologists) in interpreting medical images. Therefore, an MRMC study and the concept of "how much human readers improve with AI vs. without AI assistance" are not relevant to this device.
6. If a Standalone (i.e. algorithm only without human-in-the loop performance) Was Done
Yes, in essence. The performance studies detailed in this document (precision, method comparison, linearity, interference, etc.) directly evaluate the cobas c 513 Tina-quant HbA1cDx Gen.3 Assay and cobas c 513 Analyzer as a standalone system for quantitative determination of HbA1c in patient samples. The results are generated entirely by the instrument and reagents, independent of human interpretive input beyond standard laboratory procedures for operating the device and reporting results. There is no "human-in-the-loop" interpretive component that would require a separate "standalone" versus "human-in-the-loop" performance evaluation as might be seen for an AI algorithm.
7. The Type of Ground Truth Used
The ground truth for the HbA1c values, particularly in the method comparison study, was established using a Tosoh HPLC system operated by a secondary NGSP (National Glycohemoglobin Standardization Program) reference laboratory. This method is traceable to the approved IFCC (International Federation of Clinical Chemistry and Laboratory Medicine) reference method and linked to DCCT/NGSP standards. This represents a highly standardized and analytically robust reference method/pathology/analytical outcome data rather than expert consensus on subjective findings.
8. The Sample Size for the Training Set
This document does not describe the development of an AI/ML algorithm that would have separate "training" and "test" sets in the conventional sense. This is an IVD device where performance characteristics are assessed on "test sets" (i.e., various sample groups used in the performance studies) to demonstrate its analytical capability. Therefore, the concept of a "training set" is not applicable here.
9. How the Ground Truth for the Training Set Was Established
Not applicable, as there is no "training set" in the context of an AI/ML algorithm.
§ 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.