(111 days)
The Tina-Quant Hemoglobin A1c Gen. test is in vitro diagnostic reagent system intended for use on the COBAS INTEGRA 800 analyzers for the quantitative determination of mmol/mol hemoglobin A1c (IFCC) and % hemoglobin A1c ((DCCT/NGSP) in hemolysate or whole blood on Roche clinical chemistry analyzers. HbA1c determinations are useful for monitoring of long-term blood glucose control in individuals with diabetes mellitus.
With the Tina-Quant Hemoglobin A1c Gen. 3 test system. the anticoagulated whole blood specimen is hemolyzed prior to determination of HbA1c by an turbidimetric inhibition immunoassay (TINIA). Liberated hemoglobin (Hb) in the hemolyzed sample is converted to a derivative having a characteristic absorption spectrum and measured bichromatically. The instrument calculates the % HbA1c from the HbA1c/ Hb ratio according to a user selected protocol.
The Roche Tina-quant HbA1c Gen. 3 Assay is an in vitro diagnostic reagent system for the quantitative determination of HbA1c in hemolysate or whole blood. It is intended for monitoring long-term blood glucose control in individuals with diabetes mellitus.
Here's an analysis of its acceptance criteria and the study that proves it:
1. Table of Acceptance Criteria and Reported Device Performance
The provided document does not explicitly state "acceptance criteria" as a single, clearly defined set of thresholds that the device must meet in a comparative study. Instead, it presents performance characteristics of the new HbA1c Gen. 3 assay alongside those of its predicate device, HbA1c Gen. 2 (K072714). The implicit acceptance criteria are that the Gen. 3 assay performs equivalently or better than the Gen. 2 assay, or meets established clinical and analytical standards for HbA1c testing.
Based on the provided information, we can extract and compare the reported performance:
| Feature | HbA1c Gen. 2 (Predicate) | HbA1c Gen. 3 (New Device) |
|---|---|---|
| Intended Use | Quantitative determination of percent hemoglobin A1c [HbA1c (%)] in whole blood/hemolysate on Roche clinical chemistry analyzers. | Quantitative determination of mmol/mol hemoglobin A1c (IFCC) and % hemoglobin A1c (DCCT/NGSP) in whole blood/hemolysate on Roche clinical chemistry analyzers. (Adds IFCC units) |
| Measuring Range | Integra 400/400 plus: Hb: 4 – 35 g/dL, HbA1c: 0.3 – 2.6 g/dL | |
| Integra 800: Hb: 4 – 35 g/dL, HbA1c: 0.3 – 3.4 g/dL | Integra 400/400 plus: Hb: 4 – 40 g/dL, HbA1c: same (0.3 – 2.6 g/dL) | |
| Integra 800: Hb: 4 – 40 g/dL, HbA1c: 0.3 – 2.6 g/dL (Hb range increased, HbA1c range slightly narrower than Gen 2 on Integra 800 but still within clinical relevance) | ||
| Precision (Whole Blood) | Within-run: | |
| 0.8% @ 5.4% HbA1c | ||
| 0.9% @ 10.2% HbA1c | ||
| Between day: | ||
| 1.3% @ 5.3 % HbA1c | ||
| 1.0% @ 10.3 % HbA1c | Repeatability (Within-run): | |
| sample % CV %HbA1c | ||
| Control N 1.1 5.6 | ||
| Control P 0.8% 10.3 | ||
| human sample 1 0.9% 4.7 | ||
| human sample 2 0.8 5.8 | ||
| human sample 3: 0.7 8.7 | ||
| human sample 4 1.5 12.4 | ||
| Intermediate precision (Between day): | ||
| sample % CV %HbA1c | ||
| Control N 1.3 5.6 | ||
| Control P 0.9 10.3 | ||
| human sample 1 1.1 4.7 | ||
| human sample 2 1.1 5.8 | ||
| human sample 3: 0.8 8.7 | ||
| human sample 4 1.6 12.4 | ||
| (Generally comparable or slightly better performance; reported in more detail) | ||
| Precision (Hemolysate) | Within-run: | |
| 1.0% @ 5.5 % HbA1c | ||
| 0.6% @ 10.6 % HbA1c | ||
| Between day: | ||
| 1.0% @ 5.3 % HbA1c | ||
| 0.8% @ 10.7 % HbA1c | Repeatability (Within-run): | |
| sample % CV %HbA1c | ||
| Control N 0.9 5.6 | ||
| Control P 0.9 10.1 | ||
| human sample 1 1.0 4.7 | ||
| human sample 2 0.9 5.8 | ||
| human sample 3: 0.8 8.7 | ||
| human sample 4 1.2 12.0 | ||
| Intermediate precision (Between day): | ||
| sample % CV %HbAlc | ||
| Control N 1.3 5.6 | ||
| Control P 0.9 10.1 | ||
| human sample 1 1.2 4.7 | ||
| human sample 2 1.0 5.8 | ||
| human sample 3: 0.9 8.7 | ||
| human sample 4 1.5 12.0 | ||
| (Generally comparable or slightly better performance; reported in more detail) | ||
| Analytical Sensitivity | LDL | |
| Hb: 0.5 g/dL | ||
| HbA1c: 0.1 g/dL | LoB (Limit of Blank): | |
| Hb=0.50 g/dL | ||
| HbA1c: 0.19 g/dL | ||
| LoD (Limit of Detection): | ||
| Hb=1.0 g/dL | ||
| HbA1c: 0.29 g/dL | ||
| (LoB/LoD are different metrics than LDL, but overall indicates very low detection limits, aligning with the predicate) | ||
| Analytical Specificity | Labile HbA1c, acetylated Hb, carbamylated Hb do not affect result. HbF (> 10%) may yield lower results. | Same as predicate. |
| Endogenous Interferences | Icterus: no significant interference. | |
| Lipemia: no significant interference up to 600 mg/dL (Integra 400/400 plus) and 800 mg/dL (Integra 800) triglycerides. | ||
| Rheumatoid factors: no significant interference up to 750 IU/mL. | ||
| Glycemia: no significant interference up to 1000 mg/dL. | Icterus: Same. | |
| Lipemia: no significant interference up to 800 mg/dL Intralipid (This is a different metric than triglycerides but covers a wide range). | ||
| Rheumatoid factors: Same. | ||
| Glycemia: Not explicitly stated for Gen 3, but likely similar or covered by other equivalency claims. | ||
| Reporting Units | % HbA1c NGSP / DCCT | mmol/mol IFCC |
| % HbA1c NGSP/DCCT (Adds IFCC units) | ||
| Equation for final HbA1c | Protocol 1 (IFCC): $HbA1c(%)=(HbA1c/Hb)x100$ | |
| Protocol 2 (DCCT/NGSP): $HbA1c(%)=(HbA1c/Hb)x87.6+2.27$ | Protocol 1 (IFCC): $HbA1c(mmol/mol)=(HbA1c/Hb)x1000$ | |
| Protocol 2 (DCCT/NGSP): $HbA1c(%)=(HbA1c/Hb)x91.5+2.15$ (Equations are updated for both protocols, maintaining the dual reporting capability) |
The primary acceptance outcome demonstrated is that the Tina-quant HbA1c Gen. 3 assay maintains similar or improved analytical performance characteristics (precision, linearity, specificity, and interference) compared to its predicate device, the Tina-quant HbA1c Gen 2 assay. The addition of IFCC reporting units and updated calculation protocols, along with an extended Hb measuring range, represent enhancements that are accepted as substantially equivalent for the intended use.
2. Sample Size Used for the Test Set and Data Provenance
The document presents performance data across various studies (e.g., precision, measuring range, analytical sensitivity, interference).
- Precision: The precision studies (repeatability and intermediate precision) utilized:
- HbA1c Control N
- HbA1c Control P
- Four human samples with varying HbA1c levels.
The specific number of replicates or individual patient samples for repeatability and intermediate precision, or the total sample size across all categories for these studies, is not explicitly stated beyond "human sample 1, 2, 3, 4" and the controls.
- Other studies (Analytical Sensitivity, Interferences): The sample sizes for these tests are not explicitly quantified in the provided text.
- Data Provenance: The document does not specify the country of origin of the data or whether the studies were retrospective or prospective. Given that Roche Diagnostics is based in Indianapolis, IN, USA, and the 510(k) is submitted to the FDA, it's reasonable to infer that studies were likely conducted to meet U.S. regulatory standards, but the exact geographical origin of the samples is not stated. The type of studies described (precision, sensitivity, interference) are typically prospective, laboratory-based analytical validation studies.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
This device is an in vitro diagnostic assay, meaning its "ground truth" is established through analytical methods and reference materials, not through expert consensus on medical images or clinical diagnoses.
- Ground Truth for Analytical Performance: The "ground truth" for evaluating the assay's performance would be established by:
- Certified Reference Materials: For HbA1c, this would involve materials traceable to an international reference system, such as the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) reference method for HbA1c.
- Calibrators and Controls: The document mentions "Cfas HbA1c" as a calibrator and "HbA1c Control N" and "HbA1c Control P" as controls. These would have assigned values based on rigorous analytical standards and reference methods.
- Comparative methods: Often, new assays are compared against established, validated methods. While not explicitly detailed as a "ground truth" expert, the predicate device itself serves as a comparative benchmark.
Therefore, there are no "experts" in the sense of clinicians or radiologists establishing ground truth for individual test cases. Instead, the ground truth is analytically derived and traceable to internationally recognized standards for HbA1c measurement.
4. Adjudication Method for the Test Set
Not applicable. As this is an in vitro diagnostic device primarily evaluated on analytical performance characteristics (precision, accuracy against reference methods, linearity, interference), there is no adjudication process involving multiple human reviewers interpreting results in the way it would apply to imaging studies or clinical trials. The evaluation relies on quantitative measurements and statistical analysis.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done
No. An MRMC study is relevant for diagnostic imaging systems where human readers interpret medical images. This device is an automated in vitro diagnostic assay for measuring a biochemical marker. Therefore, an MRMC comparative effectiveness study was not conducted or applicable here.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
Yes, the studies described are standalone performance evaluations of the assay system (reagents + instrument). The "algorithm" in this context refers to the assay's chemical reactions and the instrument's photometric measurements and calculations. The precision, sensitivity, and interference studies directly assess the performance of the automated system without requiring human interpretation as part of the primary measurement. Human involvement would primarily be in sample preparation, loading, and quality control monitoring, not in providing a "human-in-the-loop" interpretation that modifies the directly reported numerical result.
7. The Type of Ground Truth Used
The ground truth used for evaluating the performance of the Tina-quant HbA1c Gen. 3 assay is primarily based on analytical reference methods and certified reference materials.
- Accuracy and calibration would be traceable to recognized HbA1c reference methods (e.g., IFCC reference method) or standards (e.g., NGSP certification).
- Control materials mentioned (Control N, Control P) have established target values.
- Precision is evaluated by repeatedly measuring samples and assessing the variability of the results against the mean measured value.
This is distinct from pathology, expert consensus, or outcomes data, which are more common for other types of medical devices.
8. The Sample Size for the Training Set
The document does not explicitly mention a "training set" in the context of machine learning (e.g., for an AI algorithm). For an in vitro diagnostic assay like this, the assay is developed and optimized (which could be considered analogous to "training") using a range of samples and conditions. However, the exact sample size for this development phase is not specified in the provided summary. The data presented focuses on the validation or test set that demonstrates the final product's performance characteristics.
9. How the Ground Truth for the Training Set was Established
Since the concept of a "training set" for an AI algorithm isn't directly applicable here, the question of how its ground truth was established also doesn't fit in the conventional sense.
However, if we interpret "training" as the assay development and optimization phase, the "ground truth" during this phase would have been established through:
- Careful chemical and biological experimentation: Optimizing reagent concentrations, reaction times, and detection methods to ensure specificity and sensitivity.
- Testing against known reference samples: Using samples with accurately determined HbA1c values (via established reference methods) to fine-tune the assay's performance and calibration.
- Addressing potential interferences: Identifying and mitigating substances or conditions that could lead to inaccurate results.
This iterative process of development relies on fundamental analytical chemistry principles and validation against rigorous scientific standards rather than discrete "ground truth" labels for a set of data points as seen in AI model training.
§ 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).