(65 days)
The ADVIA Centaur® Tnl-Ultra (Tnl-Ultra) method is for in vitro diagnosic use in the quantitative determination of cardiac troponin I in serum and heparinized and EDTA plasma. Cardiac troponin I determinations aid in the diagnosis of acute myocardial infarction and in the risk stratification of patients with non-ST segment elevation actie coronaly syndromes with respective to relative risk mortality, myocardian Tiffanction or increased probability of ischemic events requiring urgent revascularization procedures.
The ADVIA Centaur® Tnl-Ultra Calibrator is for the in vitro diagnostic use in the calibration of the Tnl-Ultra assay on the ADVIA Centaur® system.
Not Found
This response describes the acceptance criteria and study proving the device meets said acceptance criteria.
1. Table of Acceptance Criteria and Reported Device Performance
| Performance Metric | Acceptance Criteria | Reported Device Performance (ADVIA Centaur® TnI-Ultra) |
|---|---|---|
| Imprecision | Not explicitly stated as criteria, but compared to predicate devices. | Total CV: 2.7% - 4.6% (across various levels) |
| Within-run CV: 1.5% - 3.2% (across various levels) | ||
| Total CV 0.9) with predicate devices is implied. | Vs. Bayer ADVIA Centaur® cTnI | |
Linear Regression (346 samples): 1.26X - 0.52, r = 0.975 | ||
| Vs. ACS:180 cTnI | ||
Linear Regression (334 samples): 1.4X - 0.51, r = 0.967 | ||
| **Correlation (Clinically Significant Range 0.9) with predicate devices is implied. | Vs. Bayer ADVIA Centaur® cTnI | |
Linear Regression (229 samples): 1.04X + 0.04, r = 0.946 | ||
| Vs. ACS:180 cTnI | ||
Linear Regression (230 samples): 1.13X + 0.068, r = 0.943 | ||
| Clinical Utility Comparision (Concordance with 0.9 ng/mL cutoff) | Not explicitly stated. | Vs. ADVIA Centaur cTnI: 97.6% Concordance |
| Vs. ACS:180 cTnI: 93.9% Concordance | ||
| Interference | Not explicitly stated as criteria, but recovery within an acceptable deviation (e.g., 30 ng/mL):** |
* 346 serum samples for comparison with Bayer ADVIA Centaur® cTnI.
* 334 serum samples for comparison with ACS:180 cTnI.
* **Clinically Significant Range (0.1 to 4.0 ng/mL):**
* 229 serum samples for comparison with Bayer ADVIA Centaur® cTnI.
* 230 serum samples for comparison with ACS:180 cTnI.
- Clinical Utility Comparison (0.9 ng/mL cutoff): The sample sizes align with the correlation studies: 371 samples for ADVIA Centaur cTnI comparison, and 359 samples for ACS:180 cTnI comparison.
- Interfering Substances: The number of samples for interfering substances testing is not specified, but it typically involves spiking known concentrations of interferents into samples.
- ** AMI Cutoff and Clinical Performance (Sensitivity/Specificity):**
- 112 patients for sensitivity calculation.
- 166 patients for specificity calculation.
- 99th Percentile Distribution: 648 serum samples from apparently healthy donors.
Data Provenance: The document does not explicitly state the country of origin. It indicates the data is from internal studies (e.g., "data from instructions for use"). Given the manufacturer is Bayer, a multinational corporation, the samples could be from various locations. The studies appear to be prospective in the sense that they were designed to evaluate the performance of the new device against predicate devices and establish performance characteristics.
3. Number and Qualifications of Experts for Ground Truth
For a device like a Troponin I assay, "experts" in the traditional sense of clinicians or radiologists is not applicable for establishing ground truth for the assay's performance metrics (imprecision, correlation, analytical range, etc.).
- Analytical Performance: The "ground truth" for analytical performance (e.g., imprecision, linearity, interference) is established by using reference materials, certified calibrators, and well-characterized samples with known concentrations measured by highly accurate reference methods or by gravimetric/volumetric preparation. The "experts" in this context are laboratory scientists or biochemists with expertise in assay development, validation, and quality control.
- Clinical Performance (Sensitivity/Specificity for AMI): While the document refers to "patients," the ground truth for "AMI" diagnosis against which the assay's performance is measured is not explicitly detailed. However, in such studies, the ground truth for AMI is typically established by clinical diagnosis based on a combination of:
- Characteristic rise and fall of cardiac biomarker levels (often including troponin)
- Evidence of myocardial ischemia (symptoms, ECG changes)
- Imaging evidence (e.g., echocardiography, MRI).
- This clinical diagnosis would be made by cardiologists and emergency physicians, whose qualifications (e.g., years of experience) are not specified in this document.
4. Adjudication Method for Test Set
The concept of "adjudication method" (e.g., 2+1, 3+1) is primarily relevant for studies where radiologists or other human experts are interpreting images or clinical data, and their disagreements need to be resolved to establish a robust ground truth.
For this in vitro diagnostic device (IVD) assay, adjudication methods as typically understood in imaging or clinical trials are not applicable. The "adjudication" for analytical performance is through rigorous laboratory protocols, use of reference standards, and statistical analysis of quantitative results. For clinical diagnostic performance, the reference standard for AMI diagnosis (the "ground truth") is a complex clinical assessment, not typically a "read" that undergoes adjudication.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No, an MRMC comparative effectiveness study was not done. This type of study is specifically for evaluating the effectiveness of a diagnostic test when interpreted by multiple human readers, often with and without AI assistance, especially in radiology or pathology. The ADVIA Centaur® TnI-Ultra Assay is an automated in vitro diagnostic assay for quantitative measurement of a biomarker in blood, not interpreted by human readers in the same way as an imaging study. Therefore, the concept of "human readers improve with AI vs without AI" does not apply here.
6. Standalone (Algorithm Only) Performance Study
Yes, this entire submission describes a standalone performance study. The ADVIA Centaur® TnI-Ultra Assay is a fully automated system that measures troponin I levels in patient samples. The reported performance metrics (imprecision, analytical range, correlation, sensitivity, specificity, etc.) are the standalone performance of the algorithm and the associated hardware/reagents. There is no "human-in-the-loop" once the sample is loaded and the assay initiated. The output is a quantitative value, not an interpretation requiring human review or modification as part of its primary function.
7. Type of Ground Truth Used
- Analytical Performance (Imprecision, Analytical Range, Interference, Minimum Detectable Concentration): The ground truth for these aspects is intrinsically tied to reference materials, calibrators, and precise quantitative measurements using established laboratory techniques. For example, imprecision is assessed by repeatedly measuring samples with known (or target) concentrations. Interference is determined by spiking known amounts of interfering substances and measuring deviation from expected values.
- Correlation Studies: The "ground truth" here is the measurement result from the predicate devices (ADVIA Centaur® cTnI and ACS:180 cTnI), which are already cleared devices. The study compares the new device's results against these established methods.
- 99th Percentile Distribution: This is established by measuring troponin I levels in a large cohort of apparently healthy donors.
- AMI Cutoff and Clinical Performance (Sensitivity/Specificity): The ground truth for defining "AMI positive" and "AMI negative" in the sensitivity and specificity calculations would be a composite clinical diagnosis data (outcomes data) based on established criteria for acute myocardial infarction, typically involving a combination of clinical symptoms, ECG findings, and other biomarker results. While not explicitly stated, this is the standard for such IVD studies.
8. Sample Size for the Training Set
The document does not specify a separate "training set" sample size. For IVD assays like the Troponin I assay, the development process involves iterative testing and refinement, but a distinct "training set" in the machine learning sense (separate from validation/test sets) is not typically detailed in regulatory submissions for these types of devices. The data presented here is for the validation and verification of the final device, which functions based on established biochemical reaction principles and detection methods, rather than an AI model that learns from large datasets.
9. How the Ground Truth for the Training Set Was Established
As noted above, a distinct "training set" in the machine learning context is not applicable here. The "training" of such an assay involves:
- Chemical and biological research and development: Optimizing reagent formulations, reaction conditions, antibody specificity, detection methods.
- Calibration: Establishing a calibration curve using a series of known concentration standards. This is an iterative process where the assay is "trained" to accurately report concentrations based on its signal output. These standards serve as the "ground truth" for the calibration process.
- Internal validation and verification: Performing numerous internal tests (similar to those described in the acceptance criteria section) during development to ensure the assay performs as expected. The "ground truth" for these steps is based on reference methods, certified standards, and expected physiological ranges.
§ 862.1215 Creatine phosphokinase/creatine kinase or isoenzymes test system.
(a)
Identification. A creatine phosphokinase/creatine kinase or isoenzymes test system is a device intended to measure the activity of the enzyme creatine phosphokinase or its isoenzymes (a group of enzymes with similar biological activity) in plasma and serum. Measurements of creatine phosphokinase and its isoenzymes are used in the diagnosis and treatment of myocardial infarction and muscle diseases such as progressive, Duchenne-type muscular dystrophy.(b)
Classification. Class II.