(173 days)
The ADVIA® Chemistry Enzymatic Hemoglobin A1c (A1c E) assay is an in vitro diagnostic assay for the quantitative determination of mmol/mol HbA1c (IFCC) and % HbA1c (DCCT/NGSP) in human anticoagulated venous whole blood and hemolysate for use on the ADVIA® Chemistry systems. Measurement of Hemoglobin A1c is used as an aid in the diagnosis and monitoring of long-term blood glucose control in patients with diabetes mellitus, and as an aid in the identification of patients at risk for developing diabetes mellitus.
The ADVIA® Chemistry Enzymatic Hemoglobin A1c (A1c_E) assay measures hemoglobin A1c in human anticoagulated whole blood and hemolysate. The assay consists of three reagents (R1, R2, and Pretreatment). These reagents are liquid and are ready to use. The assay offers both an automated and a manual application. The automated application (A1c_E) lyses the anticoagulated whole blood specimen on the system for the automated application (A1c_E). Samples may also be lysed manually using the ADVIA® Chemistry A1c_E pretreatment solution to obtain hemolysate for the manual application (A1c EM). The two applications yield the same results.
Here's a breakdown of the acceptance criteria and study details for the ADVIA Chemistry Enzymatic Hemoglobin A1c (A1c E) Assay, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
The document defines satisfactory performance based on multiple aspects rather than a single set of explicit "acceptance criteria" presented in a table format with specific numerical targets. Instead, the performance evaluations demonstrate that the device meets clinically acceptable metrics as per established guidelines (e.g., CLSI documents) and is comparable to a predicate device. For the purpose of this request, I will extract relevant performance metrics and indicate what constitutes acceptable performance from the context.
| Performance Characteristic | Acceptance Criteria (Implicit from Context / Guidelines) | Reported Device Performance |
|---|---|---|
| Method Comparison | High correlation (r), slopes and y-intercepts close to 1 and 0 respectively, and clinically acceptable bias when compared to reference method. | Automated (Whole Blood) vs. NGSP Reference: |
- Passing-Bablok: r = 0.99, Slope = 1.019 (95% CI: 1.000 to 1.037), y-int = -0.110 (95% CI: -0.248 to 0.010)
- Deming: r = 0.99, Slope = 1.020 (95% CI: 1.004 to 1.036), y-int = -0.120 (95% CI: -0.265 to -0.006)
Manual (Hemolysate) vs. NGSP Reference: - Passing-Bablok: r = 1.00, Slope = 1.022 (95% CI: 1.004 to 1.041), y-int = -0.132 (95% CI: -0.280 to -0.019)
- Deming: r = 1.00, Slope = 1.027 (95% CI: 1.012 to 1.042), y-int = -0.176 (95% CI: -0.280 to -0.072) |
| Precision | Low coefficients of variation (CV) across different levels, runs, days, and instruments, demonstrating reproducibility. | Automated (Whole Blood) NGSP Units: Total CVs ranged from 1.0% to 1.6%.
Automated (Whole Blood) IFCC Units: Total CVs ranged from 1.2% to 3.1%.
Manual (Hemolysate) NGSP Units: Total CVs ranged from 0.9% to 1.6%.
Manual (Hemolysate) IFCC Units: Total CVs ranged from 1.2% to 3.0%. |
| Total Error (TE) | Clinically acceptable total error at decision levels. | Passing-Bablok: TE ranged from 1.97% to 3.06% across decision levels (5-12% HbA1c).
Deming: TE ranged from 1.81% to 3.22% across decision levels (5-12% HbA1c). |
| Endogenous Interference | No significant interference (bias greater than ± 5.0%). | No significant interference was observed for the tested endogenous interferents (Ascorbic Acid, Bilirubin, Total Protein, Triglycerides, Urea, Vitamin E) at specified concentrations and HbA1c levels. |
| Exogenous Interference | No significant interference (bias greater than ± 5.0%). | No significant interference was observed for the tested exogenous interferents (e.g., Acarbose, Acetaminophen, Insulin, Metformin, etc.) at specified concentrations and HbA1c levels. |
| Hemoglobin Derivative Interference | No significant interference. | No significant interference observed for HbA0, HbA1b, Acetylated Hemoglobin, Carbamylated Hemoglobin, and Labile Hemoglobin at tested levels. |
| Hemoglobin Variant Interference | No significant interference (bias greater than ± 5.0%). | No significant interference observed for HbC, HbD, HbE, and HbA2. Significant interference bias was observed for HbF. |
| Linearity | Deviations from linearity not observed within the assay range. | No deviations from linearity observed for results from 2.77 to 14.60% HbA1c. Regression equation: Observed %HbA1c = 1.0088 x Expected %HbA1c = 0.1110, R² = 0.9998. |
| Limit of Blank (LoB) | LoB value established. | 3.18% HbA1c |
| Limit of Detection (LoD) | LoD value established. | 3.60% HbA1c |
| Anticoagulant Comparison | Equivalence demonstrated for tested anticoagulants compared to K2-EDTA. | High correlation (r ≥ 0.9989) and slopes near 1 (0.998 to 1.033) for K3-EDTA, Na Fluoride/Na2-EDTA, and Lithium Heparin compared to K2-EDTA. |
2. Sample Size Used for the Test Set and Data Provenance
- Method Comparison: 163 human whole blood samples.
- Provenance: Not explicitly stated, but "human whole blood samples" usually implies clinically obtained samples. Given the context of a 510(k) submission, these would typically be retrospective or prospectively collected samples from a clinical population relevant to diabetes diagnosis and monitoring. No country of origin is specified.
- Precision: Two commercial quality controls and four whole blood patient pools. The number of individual patient samples in the pools is not specified, but the testing was extensive (20 days, 2 runs/day, 2 independent cups, 3 reagent lots, 3 instruments).
- Provenance: Whole blood pools were aliquoted and frozen, indicating patient samples. No country of origin is specified.
- Interference (Endogenous, Exogenous, Hemoglobin Derivative): Not specified as individual patient samples, but tested at two HbA1c levels using fortified samples or samples with known derivatives.
- Provenance: Not specified.
- Hemoglobin Variant Interference:
- HbC: 45 samples
- HbD: 24 samples
- HbE: 20 samples
- HbS: 25 samples
- HbA2: 20 samples
- HbF: 20 samples
- Provenance: "Anticoagulated human blood samples with known concentrations of hemoglobin variant and HbA1c." Implies clinical samples with confirmed variant types. No country of origin specified.
- Linearity: Dilution series created from high and low HbA1c whole blood pools. The number of original patient samples making up these pools is not specified.
- Provenance: "High HbA1c and low HbA1c whole blood pools," suggesting patient samples. No country of origin specified.
- Limit of Blank (LoB) and Limit of Detection (LoD): 4 blank samples and 4 low samples.
- Provenance: Not specified, likely commercial or laboratory-prepared samples.
- Anticoagulant Comparison: 96-97 samples depending on comparator.
- Provenance: Not specified, but "HbA1c values were measured for each sample," implying patient samples. No country of origin specified.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications
- Ground Truth Establishment: For the Method Comparison study, the device's results were compared to "NGSP reference method testing performed at an NGSP primary reference laboratory."
- Number of Experts: Not specified.
- Qualifications of Experts: The ground truth is based on an NGSP primary reference laboratory method, which implies that the method itself is the "expert" or gold standard, rather than individual human experts interpreting data. NGSP (National Glycohemoglobin Standardization Program) certification ensures traceability to the IFCC reference method, which is the highest standard for HbA1c measurement.
4. Adjudication Method for the Test Set
- Adjudication Method: Not applicable. The ground truth for the method comparison study was established by an NGSP primary reference laboratory method, which is an objective measurement, not a subjective interpretation requiring human adjudication. For other studies (precision, interference, linearity), performance was evaluated against statistical or predefined criteria, not against a human-adjudicated ground truth.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was Done
- MRMC Study: No, an MRMC comparative effectiveness study was not done. This device is an in vitro diagnostic assay (a laboratory test) that provides a quantitative numerical result. It does not involve human readers interpreting images, and therefore, an MRMC study comparing human readers with and without AI assistance is not relevant or applicable.
6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) was Done
- Standalone Performance: Yes, the entire performance evaluation presented in the document represents the standalone performance of the ADVIA Chemistry Enzymatic Hemoglobin A1c (A1c E) Assay. As an automated in vitro diagnostic assay, it operates without human "interpretation" in the loop once the sample is loaded and the assay runs. The results are generated by the assay system itself.
7. The Type of Ground Truth Used
- Ground Truth Type:
- For Method Comparison, the ground truth was established by NGSP primary reference laboratory method (which is directly traceable to the IFCC reference method). This is the gold standard for HbA1c measurement.
- For other studies (Precision, Interference, Linearity, LoB/LoD, Anticoagulant Comparison), the evaluations were performed against established analytical performance goals, spiked samples with known concentrations, or comparisons between different collection methods, rather than an external "ground truth" expert consensus or pathological diagnosis.
8. The Sample Size for the Training Set
- Training Set Sample Size: This information is not provided in the document. The document describes analytical performance studies of a finished in vitro diagnostic device, not the development or training of its underlying "algorithm" (which for a chemical assay is the assay's chemistry and detection principles, rather than a machine learning algorithm in the typical sense). Therefore, a "training set" in the context of AI/machine learning isn't directly applicable here. The assay's parameters would have been optimized during its development and verified through the studies presented.
9. How the Ground Truth for the Training Set Was Established
- Training Set Ground Truth Establishment: As mentioned above, the concept of a "training set" and its associated ground truth in the machine learning sense is not applicable to this type of in vitro diagnostic chemical assay. The "ground truth" for developing such assays typically relates to thoroughly characterized reference materials, known concentrations of analytes, and established analytical methods used during the research and development phase to optimize the assay's chemical reactions, instrumentation parameters, and calibration. The performance studies presented here are for the verification and validation of the final device.
§ 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.