(568 days)
The Cholesterol2 assay is used for the quantitation of cholesterol in human serum or plasma on the ARCHITECT c System. The Cholesterol2 assay is to be used as an aid in the diagnosis and treatment of disorders involving excess cholesterol in the blood and lipid and lipoprotein metabolism disorders.
The Cholesterol2 assay is an automated clinical chemistry assay for the quantitation of cholesterol in human serum or plasma on the ARCHITECT c System. Cholesterol esters are enzymatically hydrolyzed by cholesterol esterase to cholesterol and free fatty acids. Free cholesterol, including that originally present, is then oxidized by cholesterol oxidase to cholest-4-ene-3-one and hydrogen peroxide. The hydrogen peroxide oxidatively couples with N,N-Bis(4-sulfobutyl)-3-methylaniline (TODB) and 4-aminoantipyrine to form a chromophore (quinoneimine dye) which is quantitated at 604 nm.
The provided text describes the Abbott Cholesterol2 assay, an in vitro diagnostic device for quantifying cholesterol in human serum or plasma.
Here's an analysis of the acceptance criteria and study data:
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly state pre-defined acceptance criteria for each performance characteristic. Instead, it presents the results of various studies as proof of device performance. The table below summarizes the reported performance, which implicitly serves as the "met" criteria.
| Performance Characteristic | Reported Device Performance (Cholesterol2) |
|---|---|
| Analytical Measuring Interval (AMI) | 5-748 mg/dL |
| Extended Measuring Interval (EMI) | 748-2992 mg/dL |
| Reportable Interval | 2-2992 mg/dL |
| Precision | |
| Control Level 1 (251 mg/dL) | SD: 1.9 mg/dL (Within-Run), 2.6-3.1 mg/dL (Within-Laboratory); %CV: 0.7% (Within-Run), 1.0-1.2% (Within-Laboratory) |
| Control Level 2 (106 mg/dL) | SD: 1.0 mg/dL (Within-Run), 1.3-1.7 mg/dL (Within-Laboratory); %CV: 1.0% (Within-Run), 1.2-1.6% (Within-Laboratory) |
| Panel A (21 mg/dL) | SD: 0.6 mg/dL (Within-Run), 0.7-0.8 mg/dL (Within-Laboratory); %CV: 3.0% (Within-Run), 3.2-4.1% (Within-Laboratory) |
| Panel B (237 mg/dL) | SD: 2.8 mg/dL (Within-Run), 3.7-4.9 mg/dL (Within-Laboratory); %CV: 1.2% (Within-Run), 1.5-2.0% (Within-Laboratory) |
| Panel C (718 mg/dL) | SD: 6.4 mg/dL (Within-Run), 4.6-6.9 mg/dL (Within-Laboratory); %CV: 0.9% (Within-Run), 0.7-1.0% (Within-Laboratory) |
| Limit of Blank (LoB) | 0 mg/dL |
| Limit of Detection (LoD) | 2 mg/dL |
| Limit of Quantitation (LoQ) | 5 mg/dL (at 20% CV maximum allowable precision) |
| Linearity | Linear across the analytical measuring interval of 5 to 748 mg/dL |
| Interference (Endogenous) | |
| Conjugated Bilirubin (7 mg/dL) | No significant interference (within ± 10%) |
| Unconjugated Bilirubin (11 mg/dL) | No significant interference (within ± 10%) |
| Hemoglobin (1000 mg/dL) | No significant interference (within ± 10%) |
| Total Protein (15 g/dL) | No significant interference (within ± 10%) |
| Conjugated Bilirubin (40 mg/dL) | Interference: -39% at 150 mg/dL analyte, -31% at 220 mg/dL analyte |
| Unconjugated Bilirubin (16 mg/dL) | Interference: -11% at 150 mg/dL analyte |
| Interference (Exogenous) | |
| Acetaminophen (160 mg/L) | No significant interference |
| Acetylcysteine (150 mg/L) | No significant interference |
| Acetylsalicylic acid (30 mg/L) | No significant interference |
| Aminoantipyrine (40 mg/L) | No significant interference |
| Ampicillin-Na (80 mg/L) | No significant interference |
| Biotin (4250 ng/mL) | No significant interference |
| Ca-dobesilate (60 mg/L) | No significant interference |
| Cefotaxime (53 mg/dL) | No significant interference |
| Cefoxitin (6600 mg/L) | No significant interference |
| Cyclosporine (2 mg/L) | No significant interference |
| Desacetylcefotaxime (6 mg/dL) | No significant interference |
| Dipyrone (100 mg/L) | No significant interference |
| Dobutamine (0.2 mg/dL) | No significant interference |
| Doxycycline (20 mg/L) | No significant interference |
| Ibuprofen (220 mg/L) | No significant interference |
| Intralipid (1050 mg/dL) | No significant interference |
| Levodopa (8 mg/L) | No significant interference |
| Methotrexate (140 mg/dL) | No significant interference |
| Metronidazole (130 mg/L) | No significant interference |
| Methylaminoantipyrine (40 mg/L) | No significant interference |
| Methyldopa (20 mg/L) | No significant interference |
| N-Acetyl-p-benzoquinone (NAPQI) (20 mg/L) | No significant interference |
| Phenylbutazone (330 mg/L) | No significant interference |
| Phenytoin (6 mg/dL) | No significant interference |
| Rifampicin (50 mg/L) | No significant interference |
| Sodium heparin (4 U/mL) | No significant interference |
| Sulpiride (15 mg/L) | No significant interference |
| Theophylline (60 mg/L) | No significant interference |
| Ascorbic acid (60 mg/L) | Interference: -10% at 150 mg/dL analyte |
| Intralipid (2000 mg/dL) | Interference: -27% at 150 mg/dL analyte, -22% at 220 mg/dL analyte |
| Methyldopa (30 mg/L) | Interference: -14% at 150 mg/dL analyte |
| Method Comparison (vs. Predicate) | |
| Serum (n=138) | Correlation Coefficient: 1.00; Intercept: 0.41; Slope: 0.98 (Range: 7-684 mg/dL) |
| Tube Type Suitability | Acceptable for Serum, Serum separator, Lithium heparin, Lithium heparin separator, Sodium heparin tubes. |
| Dilution Verification | Automated dilution protocol (1:5.97) and manual dilution procedure (1:4) evaluated. (Performance details not provided in summary). |
| Traceability | Traceable to National Reference System for Cholesterol (Abell-Kendall reference method in a CDC-Certified CRMLN). |
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
- Precision Study: 2 controls and 3 human serum panels were tested. Each sample was tested in duplicate, twice per day for 20 days. This means 80 measurements per sample (2 duplicates x 2 times/day x 20 days).
- Lower Limits of Measurement Study: n ≥ 60 replicates for LoB, LoD, and LoQ determinations. They used low-analyte level samples and zero-analyte samples.
- Linearity Study: The number of samples for the linearity study is not explicitly stated, but it covered the range of 5 to 748 mg/dL.
- Interference Studies: Each endogenous substance was tested at 2 analyte levels (approximately 150 mg/dL and 220 mg/dL). Exogenous substances were tested at various specified interferent levels. The number of samples for each interferent is not provided.
- Method Comparison Study: 138 serum samples were used.
- Tube Type Study: Samples were collected from a minimum of 40 donors.
- Dilution Verification: 8 human serum samples.
Data Provenance: The document does not specify the country of origin of the data or whether the studies were retrospective or prospective. Given the context of medical device regulatory submission, these are typically prospective studies conducted in a controlled laboratory environment. The "human serum panels" and "human serum samples" imply human-derived biological samples.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)
This device is an in vitro diagnostic (IVD) chemistry assay. The concept of "experts establishing ground truth" as it applies to image interpretation or clinical diagnosis by medical professionals (like radiologists) does not directly apply here in the same way.
For IVDs like this, the "ground truth" or reference values are established through:
- Reference methods: The Cholesterol2 reagent is certified to be traceable to the National Reference System for cholesterol, against the Abell-Kendall reference method in a CDC-Certified Cholesterol Reference Method Laboratory Network (CRMLN). The Abell-Kendall method is considered the gold standard for cholesterol measurement.
- Analytically Validated Methods: For values outside the AMI but within the EMI, samples were "value assigned using the analytically validated method."
- Known concentrations: For studies like linearity, spiked samples with known concentrations are used.
Therefore, the "experts" are the methodologists and laboratory professionals overseeing and validating the reference methods and the analytical validation processes. No specific number or qualification of clinical experts (e.g., radiologists) is relevant for establishing the ground truth for a quantitative chemistry assay.
This is a standalone performance evaluation of the assay itself, demonstrating its analytical accuracy, precision, and robustness.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
The concept of "adjudication" (e.g., 2+1, 3+1 where multiple human readers agree or have a tie-breaker by an expert) is not applicable to this type of quantitative diagnostic assay. The results are numerical values generated by the automated instrument and reagents. Deviations or discrepancies would be resolved through re-testing, quality control, or investigation into instrument or reagent issues, rather than human expert adjudication of a "diagnosis."
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, an MRMC comparative effectiveness study was not done. This is an in vitro diagnostic assay, not an AI-powered diagnostic imaging device or an AI assistant for human readers. Its output is a quantitative measurement of cholesterol, not an interpretation that requires human "reading" or decision support.
6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done
Yes, a standalone performance evaluation of the device (Cholesterol2 assay on the ARCHITECT c8000 System) was done. The studies described (reportable interval, precision, lower limits of measurement, linearity, interference, method comparison, tube type, dilution verification) all evaluate the analytical performance of the assay and instrument directly, without human interpretation as part of the primary outcome measure. The output is a numerical concentration of cholesterol.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The ground truth used for this quantitative assay primarily relies on:
- Reference Methods: Specifically, the Abell-Kendall reference method, which is considered the gold standard for cholesterol measurement and is used in CDC-Certified Cholesterol Reference Method Laboratory Networks (CRMLN). The device's traceability to this method is explicitly stated.
- Analytically Validated Methods: For verifying values in the extended measuring interval.
- Known Spiked Concentrations: For studies such as linearity and dilution verification, where samples are prepared with precisely known concentrations.
This is an analytical ground truth, not a clinical ground truth derived from pathology or patient outcomes.
8. The sample size for the training set
This document does not describe a typical "training set" in the context of machine learning or AI. This is a chemistry assay that uses reagents and enzymatic reactions, not an algorithm that is "trained" on data. Therefore, the concept of a training set as used in AI development is not applicable here. The assay's analytical characteristics are determined through standard laboratory validation studies.
9. How the ground truth for the training set was established
As explained above, there is no "training set" in the AI sense for this device. The analytical accuracy and reliability are established through comparisons to certified reference methods and known standard concentrations, as described in point 7.
§ 862.1175 Cholesterol (total) test system.
(a)
Identification. A cholesterol (total) test system is a device intended to measure cholesterol in plasma and serum. Cholesterol measurements are used in the diagnosis and treatment of disorders involving excess cholesterol in the blood and lipid and lipoprotein metabolism disorders.(b)
Classification. Class I (general controls). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to § 862.9.