(267 days)
EliA anti-TG Immunoassay is intended for the in vitro quantitative measurement of IgG antibodies directed to thyroglobulin (TG) in human serum and plasma (Li-heparin, EDTA) as an aid in the clinical diagnosis of autoimmune thyroiditis and Graves' disease in conjunction with other laboratory and clinical findings. EliA anti-TG uses the EliA IgG method on the instrument Phadia 250.
EliA anti-TG Immunoassay is intended for the in vitro quantitative measurement of IgG antibodies directed to thyroglobulin (TG) in human serum and plasma (Li-heparin, EDTA) as an aid in the clinical diagnosis of autoimmune thyroiditis and Graves' disease in conjunction with other laboratory and clinical findings. EliA anti-TG uses the EliA IgG method on the instrument Phadia 2500/5000.
EliA anti-TPO Immunoassay is intended for the in vitro quantitative measurement of IgG antibodies directed to thyroid peroxidase (TPO) in human serum and plasma (Li-heparin, EDTA) as an aid in the clinical diagnosis of autoimmune thyroiditis and Graves' disease in conjunction with other laboratory and clinical findings. EliA anti-TPO uses the EliA IgG method on the instrument Phadia 250.
EliA anti-TPO Immunoassay is intended for the in vitro quantitative measurement of IgG antibodies directed to thyroid peroxidase (TPO) in human serum and plasma (Li-heparin, EDTA) as an aid in the clinical diagnosis of autoimmune thyroiditis and Graves' disease in conjunction with other laboratory and clinical findings. EliA anti-TPO uses the EliA IgG method on the instrument Phadia 2500/5000.
EliA Thyroid Positive Control 250 is intended for laboratory use in monitoring the performance of in vitro measurement of antibodies against thyroid peroxidase (TPO) and thyroglobulin (TG) with Phadia 250 using the EliA IgG method.
EliA Thyroid Positive Control 2500/5000 is intended for laboratory use in monitoring the performance of in vitro measurement of antibodies against thyroid peroxidase (TPO) and thyroglobulin (TG) with Phadia 2500/5000 using the EliA IgG method.
The method specific reagents on Phadia® 250 and Phadia® 2500/5000 are identical: they are only filled in different containers. Each device consists of:
- EliA anti-TG wells are coated with a human thyroglobulin antigen 4 carriers (16 wells each), ready to use; or EliA anti-TPO wells are coated with a human recombinant thyroid peroxidase
- antigen 4 carriers (16 wells each), ready to use; EliA Sample Diluent: PBS containing BSA, detergent and 0.095% sodium azide - 6 bottles, 48 mL each, ready to use; or 6 bottles, 400 mL each, ready to use;
- -EliA IgG Conjugate 50 or 200: ß-Galactosidase labeled anti-IgG (mouse monoclonal antibodies) in PBS containing BSA and 0.06% sodium azide - 6 wedge shaped bottles, 5 mL each, ready to use; or 6 wedge shaped bottles, 19 mL each, ready to use
- -EliA Thyroid Positive Control 250 or 2500/5000: Human serum containing IqG antibodies to TG and TPO in PBS containing BSA, detergent and 0.095% sodium azide - 6 single use vials, 0.3 mL each, ready to use;
- EliA Negative Control 250 or 2500/5000: Human sera from healthy donors in -PBS containing BSA, detergent and 0.095% sodium azide - 6 single-use vials, 0.3 mL each, ready to use;
- -EliA IgG Calibrator Strips: Human IqG (0, 4, 10, 20, 100, 600 µq/L) in PBS containing BSA, detergent and 0.095% sodium azide - 5 strips, 6 single-use vials per strip, 0.3 mL each, ready to use;
- -EliA IgG Curve Control Strips: Human IgG (20 µg/L) in PBS containing BSA, detergent and 0.095% sodium azide - 5 strips, 6 single-use vials per strip, 0.3 mL each, ready to use;
- -EliA IqG Calibrator Well: Coated with mouse monoclonal antibodies - 4 carriers (12 wells each), ready to use.
The Phadia EliA Immunodiagnostic System is an automated system for immunodiagnostic testing. The EliA reagents are available as modular packages, each purchased separately. All packages except the positive and negative controls are required to carry out an EliA anti-TG or anti-TPO test.
Here's an analysis of the acceptance criteria and the study proving the device meets them, based on the provided text.
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria are not explicitly stated as a formal table with pass/fail values. Instead, the document presents performance characteristics and demonstrates that the device meets these characteristics, implying they are the de facto acceptance criteria. I will infer these criteria from the studies conducted.
EliA anti-TG Immunoassay
| Acceptance Criterion (Inferred) | Reported Device Performance (Phadia 250) | Reported Device Performance (Phadia 2500/5000) |
|---|---|---|
| Precision (Total Imprecision CV%) | ||
| Low Concentration (30.6 IU/mL) | 11.9% | 17.8% |
| Mid-range Concentration (184.8 IU/mL) | 3.1% | 5.6% |
| High Concentration (4147.2 IU/mL) | 4.7% | 10.2% |
| Linearity/Reportable Range | 12 – 4794 IU/mL | 12 – 4794 IU/mL |
| Hook Effect | No hook effect observed up to 27.7 times above upper limit | No hook effect observed up to 27.7 times above upper limit |
| Limit of Detection (LoD) | 12.0 IU/mL (single shared LoD) | 12.0 IU/mL (single shared LoD) |
| Interference (Bilirubin F/C, Hemoglobin, Lipemic factor, Rheumatoid factor, Thyroxine, Iodide) | No interference observed up to specified concentrations (ratios of blank/spiked 0.94 – 1.09) | No interference observed up to specified concentrations (ratios of blank/spiked 0.94 – 1.09) |
| Reference Sera Evaluation (CAP, NEQAS) | All targets hit ("OK") | All targets hit ("OK") |
| Carry-over | Negligible effect | None (disposable tips) |
| Method Comparison (vs. Predicate VarelisA TG) | ||
| Positive Percent Agreement (equivocal as negative) | 89.7% (95% CI: 83.3% -94.3%) | Not separately reported, assumed similar |
| Negative Percent Agreement (equivocal as negative) | 88.7% (95% CI: 84.6% -92.1%) | Not separately reported, assumed similar |
| Total Percent Agreement (equivocal as negative) | 89.0% (95% CI: 85.7% -91.8%) | Not separately reported, assumed similar |
| Positive Percent Agreement (equivocal as positive) | 86.3% (95% CI: 80.7% - 90.8%) | Not separately reported, assumed similar |
| Negative Percent Agreement (equivocal as positive) | 91.3% (95% CI: 87.0% -94.5%) | Not separately reported, assumed similar |
| Total Percent Agreement (equivocal as positive) | 89.0% (95% CI: 85.7% - 91.8%) | Not separately reported, assumed similar |
| Matrix Comparison (Serum vs. EDTA Plasma) | Slope: 1.00 (0.97 to 1.03), R2: 1.00 | Not separately reported, assumed similar |
| Matrix Comparison (Serum vs. Li-heparin Plasma) | Slope: 1.00 (0.97 - 1.03), R2: 1.00 | Not separately reported, assumed similar |
| Instrument Comparison (Phadia 250 vs. Phadia 2500/5000) | Slope: 0.96 (0.93 - 0.97), Intercept: 1.7 (0.7 - 3.8) | Not separately reported, assumed compared to Phadia 250 |
| Clinical Sensitivity (AI Thyroiditis) | 55.8% (95% CI: 48.9% - 62.6%) | Not separately reported, assumed similar |
| Clinical Specificity (AI Thyroiditis) | 88.9% (95% CI: 85.8% - 91.5%) | Not separately reported, assumed similar |
EliA anti-TPO Immunoassay
| Acceptance Criterion (Inferred) | Reported Device Performance (Phadia 250) | Reported Device Performance (Phadia 2500/5000) |
|---|---|---|
| Precision (Total Imprecision CV%) | ||
| Low Concentration (15.7 IU/mL) | 8.7% | 13.1% |
| Mid-range Concentration (66.7 IU/mL) | 4.5% | 7.0% |
| High Concentration (1212.6 IU/mL) | 6.2% | 9.5% |
| Linearity/Reportable Range | 4 – 1542 IU/mL | 4 – 1542 IU/mL |
| Hook Effect | No hook effect observed up to 13.4 times above upper limit | No hook effect observed up to 13.4 times above upper limit |
| Limit of Detection (LoD) | 4.0 IU/mL (single shared LoD) | 4.0 IU/mL (single shared LoD) |
| Interference (Bilirubin F/C, Hemoglobin, Lipemic factor, Rheumatoid factor, Thyroxine, Iodide) | No interference observed up to specified concentrations (ratios of blank/spiked 0.93 – 1.05) | No interference observed up to specified concentrations (ratios of blank/spiked 0.93 – 1.05) |
| Reference Sera Evaluation (CAP, NEQAS) | All targets hit ("OK") | All targets hit ("OK") |
| Carry-over | Negligible effect | None (disposable tips) |
| Method Comparison (vs. Predicate VarelisA TPO) | ||
| Positive Percent Agreement (equivocal as negative) | 100% (95% CI: 97.8% – 100.0%) | Not separately reported, assumed similar |
| Negative Percent Agreement (equivocal as negative) | 89.7% (95% CI: 85.9% -92.8%) | Not separately reported, assumed similar |
| Total Percent Agreement (equivocal as negative) | 93.2% (95% CI: 90.6% -95.2%) | Not separately reported, assumed similar |
| Positive Percent Agreement (equivocal as positive) | 100% (95% CI: 98.3% – 100.0%) | Not separately reported, assumed similar |
| Negative Percent Agreement (equivocal as positive) | 73.7% (95% CI: 68.2% -78.7%) | Not separately reported, assumed similar |
| Total Percent Agreement (equivocal as positive) | 84.9% (95% CI: 81.5% -88.8%) | Not separately reported, assumed similar |
| Matrix Comparison (Serum vs. EDTA Plasma) | Slope: 1.00 (0.90 to 1.10), R2: 0.99 | Not separately reported, assumed similar |
| Matrix Comparison (Serum vs. Li-heparin Plasma) | Slope: 0.99 (0.95 - 1.03), R2: 0.99 | Not separately reported, assumed similar |
| Instrument Comparison (Phadia 250 vs. Phadia 2500/5000) | Slope: 0.98 (0.97 – 0.99), Intercept: 0.2 (-0.4 – 0.8) | Not separately reported, assumed compared to Phadia 250 |
| Clinical Sensitivity (AI Thyroiditis) | 82.3% (95% CI: 76.6% – 87.2%) | Not separately reported, assumed similar |
| Clinical Specificity (AI Thyroiditis) | 90.5% (95% CI: 87.5% – 92.9%) | Not separately reported, assumed similar |
2. Sample Size Used for the Test Set and Data Provenance
- Precision/Reproducibility:
- EliA anti-TG/TPO on Phadia 250/2500/5000: 8 samples, each with 252 replicate determinations (21 runs x 3 instruments x 7 runs each over 7 days).
- Data Provenance: Not explicitly stated, implied to be internal lab studies.
- Linearity/Assay Reportable Range:
- EliA anti-TG/TPO on Phadia 250/2500/5000: 6-7 patient serum samples.
- Data Provenance: Not explicitly stated, implied to be internal lab studies.
- Detection Limit (LoB/LoD):
- EliA anti-TG/TPO on Phadia 250/2500/5000: Four analyte-free samples and four low antibody concentration blood donor samples. Each measured in 36 replicates (6 replicates x 6 runs). Total of 8 samples x 36 replicates = 288 measurements.
- Data Provenance: Not explicitly stated, implied to be internal lab studies using blood donors.
- Endogenous Interference:
- EliA anti-TG/TPO: Three serum samples (one negative, one near cut-off, one high positive). Each spiked with different interfering substances or blanks and analyzed in triplicates. Runs repeated twice.
- Data Provenance: Not explicitly stated, implied to be internal lab studies.
- Reference Sera:
- EliA anti-TG: 8 samples from CAP.
- EliA anti-TPO: 8 samples from CAP and 12 samples from UK-NEQAS.
- Data Provenance: External proficiency testing programs (CAP, UK-NEQAS).
- Carry-over:
- Phadia 250: A serum sample (diluted 1:2 and 1:20). Tested with EliA Ro (another assay) to demonstrate instrument's general carry-over performance.
- Data Provenance: Not explicitly stated, implied to be internal lab studies.
- Assay Cut-off / Expected Values:
- EliA anti-TG/TPO: 604 apparently healthy blood donor samples.
- Data Provenance: Caucasian, African American, Hispanic, and Asian individuals, almost equally distributed by sex and age. (Source country not specified but "blood donor samples" are typically collected prospectively for such studies).
- Method Comparison and Clinical Sensitivity/Specificity:
- EliA anti-TG/TPO: 718 serum samples from patients with various thyroid and autoimmune conditions (Graves' disease, autoimmune thyroiditis, non-AI thyroid disease, connective tissue disease, Crohn's disease, ulcerative colitis, primary biliary cirrhosis, HIV, HCV, HBV, other infection, cancer, rheumatoid arthritis, hypergamma-globulinemia, systemic lupus erythematosus, Sjögren's syndrome, celiac disease, type 1 diabetes mellitus, type II diabetes mellitus, pregnant women of all trimesters, pre-eclampsia, miscarriage, thyroid cancer, myasthenia gravis, pernicious anemia, chronic lymphocytic thyroiditis, sub-acute thyroiditis, multi-nodular goiter).
- Data Provenance: Not explicitly stated (e.g., country of origin, retrospective/prospective).
- Matrix Comparison:
- EliA anti-TG/TPO: 57 patients from whom serum, lithium heparin plasma, and EDTA plasma were collected.
- Data Provenance: Not explicitly stated, implied to be patient samples.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document does not specify the use of "experts" in the context of diagnostic performance studies. For immunoassay devices like these, ground truth is typically established by:
- Reference Methods: Comparison to a legally marketed predicate device (VarelisA TG Antibodies, VarelisA TPO Antibodies) as seen in the method comparison study.
- Clinical Diagnosis: For sensitivity and specificity, samples are categorized based on "diagnosis" (e.g., Graves' disease, autoimmune thyroiditis). This implies a clinical ground truth, but the details of how these diagnoses were established (e.g., by how many clinicians, their specialties, or experience levels) are not provided.
- External Reference Sera: For reference sera evaluation, targets are established by external institutions like CAP and UK-NEQAS, which rely on peer groups or provider definitions.
Therefore, there's no mention of a specific number of experts or their qualifications establishing ground truth in the way one might see for image-based diagnostics.
4. Adjudication Method for the Test Set
The document does not describe an explicit adjudication method for establishing ground truth, such as 2+1 or 3+1 expert consensus. Instead, it relies on established clinical diagnoses for sensitivity/specificity studies and predicate device results for method comparison.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance
No MRMC comparative effectiveness study was done. This device is an in vitro diagnostic (IVD) immunoassay, not an AI-assisted diagnostic tool for human readers. Its performance is evaluated intrinsically and in comparison to a predicate device, not in terms of human reader improvement.
6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was done
The device is a standalone immunoassay system (EliA anti-TG Immunoassay and EliA anti-TPO Immunoassay run on Phadia 250/2500/5000 instruments). Its performance is inherently "standalone" in the sense of being an automated laboratory test without direct human interpretive input at the point of result generation. Results are quantitative values that clinicians interpret. Therefore, the entire analytical and clinical performance section describes the device's standalone performance.
7. The Type of Ground Truth Used
- Clinical Diagnosis: For clinical sensitivity and specificity studies, samples were linked to known diagnoses such as "Graves' Disease" and "Autoimmune Thyroiditis." The basis for these diagnoses (e.g., pathology, clinical outcomes, or expert opinion) is not detailed beyond being "clinically defined."
- Predicate Device: For method comparison, the reference standard was the legally marketed predicate device (VarelisA TG Antibodies and VarelisA TPO Antibodies).
- External Program Targets: For reference sera, the ground truth was the targets set by CAP (College of American Pathologists) and UK-NEQAS (United Kingdom National External Quality Assessment Service). These targets are often established by peer consensus or by the program providers.
8. The Sample Size for the Training Set
The document describes studies for validation of performance characteristics but does not explicitly mention a "training set" in the context of machine learning or AI development. Since these are immunoassays, they are based on biochemical reactions and calibration rather than a machine learning model that requires training data. The calibration process uses specific calibrator materials.
9. How the Ground Truth for the Training Set Was Established
As this is an immunoassay and not an AI/ML device, the concept of a "training set" with established ground truth doesn't apply in the same way. The device is calibrated using:
- EliA IgG Calibrator Strips: Human IgG at defined concentrations (0, 4, 10, 20, 100, 600 µq/L). These calibrators are traceable to the International Reference Preparation (IRP) 67/86 of Human Serum Immunoglobulins A, G and M from WHO. New batches are compared to a secondary standard or the IRP directly to ensure correct concentration. This traceability constitutes the "ground truth" for calibration.
- EliA IgG Curve Control Strips: Human IgG (20 µg/L) for monitoring calibration curve performance.
The "ground truth" for the calibrators is therefore established through established international reference standards and their traceability.
§ 866.5870 Thyroid autoantibody immunological test system.
(a)
Identification. A thyroid autoantibody immunological test system is a device that consists of the reagents used to measure by immunochemical techniques the thyroid autoantibodies (antibodies produced against the body's own tissues). Measurement of thyroid autoantibodies may aid in the diagnosis of certain thyroid disorders, such as Hashimoto's disease (chronic lymphocytic thyroiditis), nontoxic goiter (enlargement of thyroid gland), Grave's disease (enlargement of the thyroid gland with protrusion of the eyeballs), and cancer of the thyroid.(b)
Classification. Class II (performance standards).