(241 days)
cobas pure integrated solutions is an automated analyzer, intended for running qualitative, semiguantitative and quantitative clinical chemistry and immunochemistry assays as well as ion selective measurements.
Glucose HK Gen.3 is an in vitro test for the quantitative determination of glucose in human serum, plasma, urine and CSF on Roche/Hitachi cobas c systems. Glucose measurements are used in the diagnosis and treatment of carbohydrate metabolism disorders including diabetes mellitus, neonatal hypoqlycemia, idiopathic hypoglycemia and pancreatic islet cell tumors.
The ISE analytical unit of the Roche/Hitachi cobas c systems is intended for the quantitative determination of sodium in serum, plasma or urine using ion-selective electrodes. Sodium measurements are used in the diagnosis and treatment of aldosteronism (excessive secretion of the hormone aldosterone), diabetes insipidus (chronic excretion of large amounts of dilute urine, accompanied by extreme thirst), adrenal hypertension, Addison's disease (caused by destruction of the adrenal glands), dehydration, inappropriate antidiuretic hormone secretion, or other diseases involving electrolyte imbalance.
Methadone II (MDN2) is an in vitro diagnostic test for the qualitative and semiquantitative detection of methadone in human urine on Roche/Hitachi cobas c systems at a cutoff concentration of 300 ng/mL. Semiquantitative test results may be obtained that permit laboratories to assess assay performance as part of a quality control program. Semiquantitative assays are intended to determine an appropriate dilution of the specimen for confirmation by a confirmatory method such as gas chromatography/mass spectrometry (GC-MS).
Elecsys TSH is an immunoassay for the in vitro quantitative determination of thyrotropin in human serum and plasma. Measurements of TSH are used in the diagnosis of thyroid and pituitary disorders. The electrochemiluminescence immunoassay "ECLIA" is intended for use on cobas e immunoassay analyzers.
The cobas pure integrated solutions is a fully automated, random-access, software controlled system intended for in vitro quantitative and qualitative analysis of analytes in body fluids. It will typically be used in low to mid throughput clinical laboratories. The system consolidates clinical chemistry, homogenous immunoassays as well as electrolyte testing within one workplace. The cobas pure integrated solutions consists of a clinical chemistry analytical unit (cobas c 303) with an integrated ISE analytical unit, an immunoassay analytical unit (cobas e 402) and a core unit.
Glucose is phosphorylated by hexokinase (HK) in the presence of adenosine triphosphate (ATP) and magnesium ions to produce glucose-6-phosphate (G-6-P) and adenosine diphosphate (ADP). Glucose-6-phosphate dehydrogenase (G-6-PDH) specifically oxidizes G-6-P to 6-phosphogluconate with the concurrent reduction of nicotinamide adenine dinucleotide (NAD) to nicotinamide adenine dinucleotide reduced (NADH). One micromole of NADH is produced for each micromole of glucose consumed. The NADH produced absorbs light at 340 nm and can be detected spectrophotometrically as an increased absorbance.
The ISE analytical unit for Na+ employs ion-selective membrane to develop an electrical potential (electromotive force, EMF) for the measurements of ions in solution. Selective membrane is in contact with both the test solution and an internal filling solution. Due to the selectivity of the membrane, only the ions to be EMF. The membrane EMF is determined by the difference in concentration of the test ion in the test solution and the internal filling solution.
The ISE analytical unit of the Roche/Hitachi cobas c systems is intended for the quantitative determination of sodium in serum, plasma or urine using ion-selective electrodes. Sodium is the major extracellular cation and functions to maintain fluid distribution and osmotic pressure. Some causes of decreased levels of sodium include prolonged vomiting or diarrhed reabsorption in the kidney and excessive fluid retention. Common causes of include excessive fluid loss, high salt intake and increased kidney reabsorption.
The Methadone assay is based on the kinetic interaction of microparticles in a solution (KIMS) as measured by changes in light transmission. In the absence of sample drug conjugates bind to antibody-bound microparticles, causing the formation of particle aggregates. As the aggregation reaction proceeds in the absence of sample drug, the absorbance increases.
When a urine sample contains the drug in question, this drug derivative conjugate for microparticle-bound antibody. Antibody bound to sample drug is no longer available to promote particle aggregation, and subsequent particle lattice formation is inhibited. The presence of sample drug diminishes the increasing absorbance in proportion to the concentration of drug in the sample. Sample drug content is determined relative to the value obtained for a known cutoff concentration of drug.
The Elecsys TSH immunoassay makes use of a sandwich test principle using monoclonal antibodies specifically directed against human TSH. The antibodies labeled with ruthenium complex) construct from human and mouse specific components. Elecsys TSH immunoassay 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 and pituitary disorders. It is intended for use on the cobas e immunoassay analyzers.
The provided text describes several in vitro diagnostic (IVD) devices and their performance characteristics. It outlines how Roche Diagnostics established the substantial equivalence of these devices (Glucose HK Gen.3, ISE indirect Na for Gen.2, ONLINE DAT Methadone II, Elecsys TSH, and the cobas pure integrated solutions analyzer) to their predicate devices through various non-clinical tests.
However, the document does not provide a table of acceptance criteria and reported device performance in a format that easily allows for direct comparison against specific numeric targets. Instead, it describes general study types and states that the "analytical performance data for all representative assays meet specifications and support the substantial equivalence."
Furthermore, it does not explicitly detail the following requested information:
- Sample sizes used for the test set and data provenance: The document mentions "human serum, plasma, urine and CSF" as sample types but not explicit test set sample numbers or their country of origin for most tests. It generally refers to "human samples" or "pooled human plasma and serum samples."
- Number of experts used to establish the ground truth for the test set and their qualifications: This information is not present for any of the described tests. The ground truth for these IVDs is typically established through reference methods or established control materials, not expert consensus on individual cases.
- Adjudication method: Not applicable/provided as the tests are analytical and do not involve human interpretation requiring adjudication.
- Multi Reader Multi Case (MRMC) comparative effectiveness study: Not applicable, as these are IVD assays, not AI-assisted reader studies.
- Standalone performance: The entire document describes the standalone performance of the algorithms/assays.
- Type of ground truth used: For quantitative assays (Glucose, Sodium, TSH), the ground truth is implicitly based on established reference methods or known concentrations in control materials. For qualitative/semi-quantitative assays like Methadone, it's based on known concentrations relative to a cutoff or confirmed by methods like GC-MS.
- Sample size for the training set: Not explicitly stated, as these are typically not machine learning models in the sense of needing a distinct "training set" for classification, but rather reagents and analytical systems whose performance is validated. The mentioned "study" refers to validation studies, not AI model training.
- How the ground truth for the training set was established: Not applicable, as explained above.
Given the nature of the document, which focuses on device-specific analytical performance claims rather than AI model validation, several of the requested categories are not directly addressed or are not relevant.
Below is an attempt to structure the available information, noting the limitations.
1. Table of Acceptance Criteria and Reported Device Performance
As specific numerical acceptance criteria (e.g., "sensitivity ≥ 95%") are not explicitly provided in the text, this table will summarize what types of performance were evaluated and that they met specifications as stated in the document.
| Device/Assay | Performance Metric Evaluated | Reported Device Performance |
|---|---|---|
| Glucose HK Gen.3 | Precision (Repeatability, Intermediate Precision) | Met specifications (according to CLSI EP05-A3) |
| Linearity | Met specifications (according to CLSI EP06-A) | |
| Limit of Blank (LoB) | Determined; highest observed measurement value for analyte-free samples. | |
| Limit of Detection (LoD) | Determined; lowest amount of analyte detectable with 95% probability. | |
| Limit of Quantitation (LoQ) | Determined; lowest concentration quantifiable with ≤ 20% total error. | |
| Endogenous Interference | Effects determined for hemoglobin, lipemia, albumin, IgG, bilirubin. | |
| Drug Interference | Evaluated for potential interference. | |
| Method Comparison | Performed vs. Glucose HK Gen.3 on cobas c 503. | |
| Matrix Comparison | Supported use of different anticoagulant tube types. | |
| On-board Reagent Stability | Verified for 26 weeks. | |
| Post Dilution Check | Verified automatic rerun function. | |
| Recovery in Controls | Tested. | |
| ISE indirect Na for Gen.2 | Precision (Repeatability, Intermediate Precision) | Met specifications (according to CLSI EP05-A3) |
| Linearity | Demonstrated across claimed measuring range (according to CLSI EP06-A). | |
| Limit of Blank (LoB) | Determined; highest observed measurement value for analyte-free samples. | |
| Limit of Detection (LoD) | Determined; lowest amount of analyte detectable with 95% probability. | |
| Limit of Quantitation (LoQ) | Determined; lowest concentration quantifiable with ≤ 30% total error. | |
| Endogenous Interference | Effects determined for bilirubin, hemolysis, lipemia. | |
| Drug Interference | Evaluated for potential interference. | |
| Method Comparison | Performed vs. ISE indirect Na for Gen2. on cobas pro ISE and flame photometry. | |
| Matrix Comparison | Determined effect of anticoagulants (Li-Heparin Plasma and Serum). | |
| Calibration Frequency | Study verifying performed. | |
| Post Dilution Check | Verified automatic rerun function. | |
| Recovery in Controls | Measured. | |
| ONLINE DAT Methadone II | Precision (Repeatability, Intermediate Precision) | Met specifications (according to CLSI EP05-A3) |
| Endogenous Interference | Tested effects of interfering substances on urine samples with methadone at -25% of cutoff. | |
| Drug Interference | Evaluated effects of various drugs/compounds on results relative to 300 ng/mL cutoff. | |
| Cross Reactivity | Inhibition curves generated; approximate quantity for equivalent reactivity to 300 ng/mL cutoff determined. | |
| Method Comparison | Performed vs. GC-MS and cobas c 503. | |
| Recovery in Controls | Tested. | |
| Elecsys TSH | Precision (Repeatability, Intermediate Precision) | Met specifications (according to CLSI EP05-A3) |
| Linearity | Assessed on cobas e 402 (according to CLSI EP06-A). | |
| Limit of Blank (LoB) | Determined; highest observed measurement values for samples. | |
| Limit of Detection (LoD) | Determined; lowest amount of analyte detectable with 95% probability. | |
| Limit of Quantitation (LoQ) | Determined; lowest concentration quantifiable with ≤ 20% total error. | |
| Endogenous Interfering Substances | Determined effect on quantitation in human serum samples. | |
| Drug Interference | Determined effect on quantitation by comparing spiked vs. unspiked samples. | |
| Cross-reacting Compounds | Determined effect on quantitation in human serum sample pool. | |
| On-board Reagent Stability | Tested on cobas e 402. | |
| Method Comparison | Performed vs. predicate Elecsys TSH on cobas e 801 to assess bias. | |
| Anticoagulants Effect | Determined effect on quantitation in various plasma tubes. | |
| High-dose Hook Effect | Assessed on cobas e 402. |
Overall Conclusion for all devices: "The analytical performance data for all representative assays meet specifications and support the substantial equivalence...to the predicate devices."
2. Sample sizes used for the test set and the data provenance
- Glucose HK Gen.3:
- Precision studies: "Two aliquots per run, two runs per day for ≥ 21 days". Types of samples (e.g., patient, control) not specified, but typically human-derived.
- Linearity, LoB, LoD, LoQ, Interference, Matrix Comparison, Post Dilution Check, Recovery in Controls: Sample numbers or types of samples for these specific studies are not detailed.
- Method Comparison: "all sample types" (e.g., human serum, plasma, urine, CSF) tested between cobas c 303 and cobas c 503.
- Data Provenance: Not specified (e.g., country of origin, retrospective/prospective).
- ISE indirect Na for Gen.2:
- Precision studies: "One run per day for ≥ 21 days with two parts, two aliquots per part". Typically human-derived samples.
- Endogenous Interference: "pooled human plasma and serum samples spiked with varying levels of interferent."
- Linearity, LoB, LoD, LoQ, Drug Interference, Calibration Frequency, Post Dilution Check, Recovery in Controls: Sample numbers or specific types of samples not detailed.
- Method Comparison: "all sample types" (e.g., human serum, plasma, urine) tested.
- Data Provenance: Not specified.
- ONLINE DAT Methadone II:
- Precision: "human samples and controls" (n=84 for repeatability, with 2 aliquots per run, 2 runs per day, 21 days for intermediate precision).
- Endogenous Interference: "urine containing methadone" and "pooled human urine".
- Drug Interference: "human urine containing methadone".
- Cross Reactivity: Details not given, but likely spiked drug solutions into human urine.
- Method Comparison: Urine samples compared against GC-MS and cobas c 503.
- Data Provenance: Not specified.
- Elecsys TSH:
- Precision: Likely control materials and potentially patient samples.
- Endogenous Interfering Substances: "human serum samples (native serum pools)".
- Cross-reacting Compounds: "native human serum sample pool".
- Anticoagulants Effect: "native human serum samples, single donors as well as pools" drawn into various tubes.
- Linearity, LoB, LoD, LoQ, Drug Interference, On-board Reagent Stability, High-dose Hook Effect: Sample numbers or types not specified.
- Method Comparison: Human serum and plasma samples compared between cobas e 402 and cobas e 801.
- Data Provenance: Not specified.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
Not applicable. For in vitro diagnostic assays, "ground truth" is established through analytical reference methods, defined concentrations of calibrators/standards, or confirmed by other laboratory methods (e.g., GC-MS for Methadone) rather than expert human interpretation of results.
4. Adjudication method
Not applicable. These are analytical tests performed by automated systems; there is no human interpretation or adjudication involved in generating the primary test result.
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
Not applicable. This is for in vitro diagnostic assays on automated analyzers, not AI-assisted human reader studies. The "AI" would refer to the algorithms within the analytical unit, not a system designed to assist human readers.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
Yes, the entire document focuses on the standalone performance of the analytical systems and assays. The results are generated directly by the device/analyzer without human intervention for interpretation beyond loading samples and performing quality control.
7. The type of ground truth used
- For Glucose, Sodium, TSH (Quantitative assays): The ground truth for these quantitative measurements is based on reference methods, known concentrations in calibrators and controls, or comparative methods (e.g., flame photometry for sodium, or existing cleared assays on predicate devices).
- For Methadone (Qualitative/Semi-quantitative assay): Ground truth is established by known concentrations relative to a cutoff (e.g., 225 ng/mL for negative control, 375 ng/mL for positive control) and confirmed by a "confirmatory method such as gas chromatography/mass spectrometry (GC-MS)".
8. The sample size for the training set
Not explicitly specified. These are not AI/ML models in the typical sense that require a distinct "training set." The development of reagents and analytical platforms involves extensive R&D and optimization, but the validation studies described here are for demonstrating performance and equivalence, not for "training" an algorithm in a machine learning context.
9. How the ground truth for the training set was established
Not applicable, as a distinct "training set" with established ground truth in the context of machine learning model development is not typically associated with the development and validation of these types of in vitro diagnostic reagents and analyzers. The principles are generally based on established biochemical reactions, electrochemical measurements, or immunoassays.
§ 862.1345 Glucose test system.
(a)
Identification. A glucose test system is a device intended to measure glucose quantitatively in blood and other body fluids. Glucose measurements are used in the diagnosis and treatment of carbohydrate metabolism disorders including diabetes mellitus, neonatal hypoglycemia, and idiopathic hypoglycemia, and of pancreatic islet cell carcinoma.(b)
Classification. Class II (special controls). The device, when it is solely intended for use as a drink to test glucose tolerance, is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 862.9.