(126 days)
cobas e 801 immunoassay analyzer is intended for the in-vitro determination of analytes in body fluids.
Elecsys TSH immunoasay is intended for the in vitro quantitative determination of thyrotropin in human serum and plasma. Measurements of TSH are used in the diagnosis of thyroid or pituitary disorders. The Elecsys TSH immunoassay is an electrochemiluminescence immunoasay 'ECLIA', which is intended for use on the cobas e immunoassay analyzers.
The cobas e 801 immunoassay analyzer is a fully automated, software controlled analyzer system for in vitro determination of analytes in human body fluids. It is part of the cobas 8000 modular analyzer series cleared under K100853. It uses electrochemiluminescent technology for signal generation and measurement.
Here's a breakdown of the acceptance criteria and study information for the Elecsys TSH assay on the Cobas e 801 immunoassay Analyzer, based on the provided document:
1. Table of Acceptance Criteria and Reported Device Performance
The document doesn't explicitly state "acceptance criteria" for each performance characteristic as pass/fail thresholds. Instead, it presents the results of various validation studies. I will present the reported performance, and where applicable, infer the implied acceptance based on the presentation of the results as successful.
| Performance Characteristic | Acceptance Criteria (Implied) | Reported Device Performance |
|---|---|---|
| Repeatability (CV%) | Low CV values, generally 0.99) and slope close to 1, intercept close to 0 across the measuring range | Serum 1: Pearson's r = 0.9994, slope = 0.963, intercept = -0.00155 |
| Serum 2: Pearson's r = 0.9992, slope = 0.958, intercept = -0.00193 | ||
| Serum 3: Pearson's r = 0.9986, slope = 0.952, intercept = -0.00272 | ||
| Limit of Blank (LoB) | Low value, typically indicative of assay's ability to distinguish analyte-free samples from those with very low levels. | 0.0025 µIU/mL |
| Limit of Detection (LoD) | Low value, indicating sensitivity to low analyte concentrations. Specific 95% probability is a common criterion. | 0.005 µIU/mL (detected with 95% probability) |
| Limit of Quantitation (LoQ) | Low value with acceptable precision (e.g., CV ≤ 20%) | 0.005 µIU/mL at a CV ≤ 20% |
| Endogenous Interferences | No significant interference at specified levels | No interference observed up to the indicated levels for Intralipid (2000 mg/dL), Biotin (56.0 ng/mL), Bilirubin (66.0 mg/dL), Hemoglobin (1000 mg/dL), Rheumatic Factor (1500 IU/mL), human IgG (2.80 g/dL), human IgM (0.500 g/dL). |
| Exogenous Interferences (Anticoagulants) | Values obtained from different sample types (serum, plasma with various anticoagulants) should be comparable. | Data supported the use of Serum, Li-Heparin, K2-EDTA, and K3-EDTA plasma tubes, evaluated using Passing/Bablok regression analysis comparing serum/plasma pairs. |
| Exogenous Interferences (Drugs) | No significant interference at specified drug concentrations | No interference found with 16 commonly used drugs and several special drugs (Amiodarone, Carbimazole, Fluocortolone, Hydrocortisone, Iodide, Levotyroxine, Liothyronine, Methimazole, Octreotide, Prednisolone, Propanolol, Propylthiouracil, Perchlorate) at tested concentrations. |
| Method Comparison (Correlation) | Strong correlation (e.g., Pearson's r > 0.98) and agreement (slope close to 1, intercept close to 0) with predicate device | N = 130 samples |
| Passing/Bablok: Slope = 0.936, Intercept = -0.003, Kendall ($\tau$) = 0.989 | ||
| Linear Regression: Slope = 0.958, Intercept = -0.052, Pearson (r) = 0.999 |
2. Sample Size and Data Provenance for Test Set
- Repeatability and Intermediate Precision:
- Sample Size:
- 7 human serum samples (5 pooled, 5 pooled spiked) and 2 control samples (PC Universal, PreciControl TS).
- Each sample tested in 2 replicates per run, 2 runs per day for 21 days (total of 84 replicate measurements per sample type).
- Sample Size:
- Linearity:
- Sample Size: 3 high analyte serum samples diluted to 12 concentrations. Each concentration assayed in 3-fold determination within a single run.
- Analytical Sensitivity (LoB, LoD, LoQ):
- LoB: Blank sample tested with 60 replicates (10 replicates per run, 6 days).
- LoD: 5 low-level human serum samples tested with 60 replicates (2 replicates per sample per run, 6 days).
- LoQ: 10 low-level TSH samples tested over 5 days, 5 replicates per sample per day.
- Endogenous Interferences:
- Human serum samples. Specific number not provided, but the outcome is qualitative ("No interference observed").
- Exogenous Interferences (Anticoagulants):
- A minimum of 40 serum/plasma pairs per sample material (presumably 40 serum, 40 Li-Heparin plasma, 40 K2-EDTA plasma, 40 K3-EDTA plasma). Tested in singleton for each type.
- Exogenous Interferences (Drugs):
- 16 commonly used drugs and 13 special drugs. Specific number of samples not provided, but the outcome is qualitative ("No interference").
- Method Comparison:
- Sample Size: 130 human serum samples (single donors and serum pools; native, spiked as well as diluted).
- Data Provenance: Not explicitly stated (e.g., country of origin, retrospective/prospective). However, the use of "human serum samples" and "human serum/plasma pairs" indicates biological samples. The studies are prospective in nature, conducted specifically to validate the device.
3. Number of Experts Used to Establish Ground Truth for Test Set and Qualifications
This device is an immunoassay for quantitative determination of TSH. The "ground truth" for the test set is established by the reference measurement method or the quantitative value itself. No human experts are used to establish ground truth in the same way they would be for image interpretation. The accuracy of the quantitative measurements is assessed against known concentrations (e.g., in linearity, LoB/LoD studies) or against a predicate device (method comparison).
4. Adjudication Method
Not applicable for an immunoassay. Adjudication is typically used when human interpretation of results contributes to the ground truth, which is not the case here.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No. This type of study is relevant for diagnostic imaging or other interpretations where human readers are involved. This document describes the performance of an in-vitro diagnostic device (IVD) for quantitative measurement, which operates without direct human interpretive input being part of the core measurement.
6. Standalone Performance
Yes, these studies describe the standalone (algorithm only, in this context meaning the device's automated performance without human intervention after sample loading) performance of the Elecsys TSH assay on the cobas e 801 immunoassay analyzer. The results presented are directly from the instrument's measurements.
7. Type of Ground Truth Used
- Reference Materials/Known Concentrations: For precision, linearity, LoB, LoD, LoQ, endogenous and exogenous interference studies, ground truth implicitly refers to the expected concentration/value based on the preparation of known samples or the absence of an analyte (for blanks).
- Predicate Device Measurements: For the method comparison study, the "ground truth" for comparison is the measurement obtained from the predicate device (Elecsys TSH on Elecsys 2010 analyzer).
- Expected Values: The "Expected Values" section establishes a reference range (0.27-4.20 µIU/mL) based on healthy test subjects. This is a clinical reference range, not a direct ground truth for individual measurements, but rather a benchmark for interpretation.
8. Sample Size for the Training Set
The document describes validation studies for an in vitro diagnostic device (IVD), specifically an immunoassay analyzer and assay. IVDs like this do not typically have a "training set" in the machine learning sense. The device's operational parameters, calibration curves, and algorithms are developed during the product development phase by the manufacturer, which might involve internal data, but generally, the submission focuses on external validation. The document does not provide details on the development data used.
9. How the Ground Truth for the Training Set Was Established
As there is no "training set" described in the context of machine learning, this question about establishing ground truth for it is not applicable here. The device output is a quantitative value based on chemical reactions and detection, not a learned prediction from a ground-truthed dataset in the AI sense.
§ 862.1690 Thyroid stimulating hormone test system.
(a)
Identification. A thyroid stimulating hormone test system is a device intended to measure thyroid stimulating hormone, also known as thyrotrophin and thyrotrophic hormone, in serum and plasma. Measurements of thyroid stimulating hormone produced by the anterior pituitary are used in the diagnosis of thyroid or pituitary disorders.(b)
Classification. Class II.