Search Results
Found 7 results
510(k) Data Aggregation
(30 days)
In vitro test for the quantitative determination of glucose in serum, plasma, urine and cerebrospinal fluid (CSF) on COBAS INTEGRA systems.
Glucose measurements are used in diagnosis and treatment of carbohydrate metabolism disorders including diabetes mellitus, neonatal hypoglycemia, and idiopathic hypoglycemia, and pancreatic islet cell tumors.
The cassette COBAS INTEGRA Glucose HK Gen. 3 contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA SYSTEMS for the quantitative determination of glucose in serum, plasma, urine, and cerebrospinal fluid (CSF). The test principle is an enzymatic reference method with hexokinase.
The provided text is a 510(k) Summary for the COBAS INTEGRA Glucose HK Gen. 3 device, which describes an in vitro diagnostic reagent system for the quantitative determination of glucose. This type of submission focuses on demonstrating substantial equivalence to a legally marketed predicate device rather than fulfilling specific acceptance criteria in the way a new, novel device might. Therefore, the information provided focuses on comparative performance rather than predefined acceptance thresholds.
Based on the provided document, here's a breakdown of the requested information:
1. A table of acceptance criteria and the reported device performance
For a 510(k) submission seeking substantial equivalence, the "acceptance criteria" are generally that the modified device performs as well as, or better than, the predicate device. The performance characteristics of the COBAS INTEGRA Glucose HK Gen. 3 (modified device) are compared with its predicate device (COBAS INTEGRA Glucose HK Liquid, K972250).
| Feature | Acceptance Criteria (Implied by Predicate Performance) | Reported Device Performance (COBAS INTEGRA Glucose HK Gen. 3) |
|---|---|---|
| Measuring Range | 0-40 mmol/L (0-720 mg/dL) | 0.12-40 mmol/L (0.12-720 mg/dL) |
| Extended Measuring Range | 0-400 mmol/L (0-7200 mg/dL) | 0.12-400 mmol/L (2.16-7200 mg/dL) |
| Precision (Within run CV%) | Serum & Plasma: 1.7% @ 5.3 mmol/L; 0.72% @ 33.2 mmol/L | Serum & Plasma: 0.41% @ 4.48 mmol/L; 0.47% @ 12.48 mmol/L |
| Urine: 1.7% @ 1.7 mmol/L; 1.8% @ 37.1 mmol/L | Urine: 1.35% @ 0.83 mmol/L; 0.64% @ 2.42 mmol/L | |
| CSF: 1.6% @ 1.7 mmol/L; 1.8% @ 3.3 mmol/L | CSF: 1.13% @ 3.20 mmol/L; 1.49% @ 9.31 mmol/L | |
| Precision (Between run/day CV%) | Serum & Plasma: 2.6% @ 5.3 mmol/L; 1.5% @ 33.2 mmol/L | Serum & Plasma: 1.09% @ 4.44 mmol/L; 0.90% @ 12.46 mmol/L |
| Urine: 4.3% @ 1.7 mmol/L; 2.9% @ 37.1 mmol/L | Urine: 0.75% @ 0.84 mmol/L; 0.83% @ 2.43 mmol/L | |
| CSF: 2.3% @ 1.7 mmol/L; 1.9% @ 3.3 mmol/L | (Not explicitly stated for between day in CSF, only within run) | |
| Linearity | 0-40 mmol/L (before dilution) | 0.12-40 mmol/L (before dilution) |
| Lower Detection Limit | Serum & Plasma: 0.033 mmol/L | Serum, Plasma, Urine & CSF: 0.12 mmol/L |
| Urine: 0.22 mmol/L | ||
| CSF: 0.023 mmol/L | ||
| Endogenous Interferences | Hemolysis, Icterus, Lipemia - no significant interferences for predicate | Hemolysis: up to 1200 H Index; Icterus: up to 60 I Index; Lipemia: up to 1900 L Index (quantitative limits provided) |
| Exogenous Interferences | Falsely low results by elevated pyruvates; Gammopathy may cause unreliable results | Tetracyclin can cause falsely low results in urine; Gammopathy may cause unreliable results |
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
The document primarily describes comparative performance data. Specific sample sizes for "test sets" or the provenance of the data (country of origin, retrospective/prospective) are not explicitly stated. The data presented such as precision data, linearity, and interference studies imply internal validation studies were performed, but details on the samples used (e.g., number of patient samples, type of samples beyond serum/plasma/urine/CSF) are absent in this summary.
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 is an analytical performance study for an in vitro diagnostic device, not an imaging device or a device requiring human interpretation for "ground truth". Therefore, there are no "experts" in the sense of clinical specialists establishing ground truth for the test set. Ground truth for glucose measurements is typically established through a reference method or validated calibrators and controls.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
Not applicable for this type of analytical performance study. Adjudication methods are relevant for human interpretation tasks, such as reading medical images.
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 an in vitro diagnostic device measuring glucose, not an AI-assisted diagnostic tool that aids human readers.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
This is an automated in vitro diagnostic test system. The reported performance characteristics (precision, linearity, lower detection limit, interferences) are inherent to the device and reagent system itself, operating without human intervention for the measurement process once the sample is loaded. Thus, the performance data provided can be considered standalone performance.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
The ground truth for glucose measurements is established by:
- Reference method: The test principle is an "enzymatic reference method with hexokinase."
- Traceability: It is "Standardized against Isotope Dilution Mass Spectrometry," which is a highly accurate and precise reference method for measuring glucose.
8. The sample size for the training set
This document does not specify a "training set" in the context of machine learning or AI models. It refers to the development and validation of an analytical measurement system. Data used for method development (e.g., reagent formulation optimization) would be part of standard product development but is not explicitly defined as a "training set" in this summary.
9. How the ground truth for the training set was established
Not applicable as there is no mention of a "training set" for an AI/ML algorithm. For the analytical method development, ground truth would be established through highly accurate reference methods and calibrators, similar to how the device's performance is validated.
Ask a specific question about this device
The cassette COBAS INTEGRA LDL Direct (LDL-D) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA 700 for the quantitative determination of LDL-cholesterol direct concentration in serum and plasma. Low density lipoprotein cholesterol measurement, in conjunction with other lipid determinations, has been shown to be useful in assessing the risk of coronary heart disease.
The Roche Calibrator LDL Direct is intended for use as calibrator in quantitative Low Density Lipoprotein cholesterol assays. It is recommended for use with LDL Direct reagents on COBAS® chemistry systems. A calibrator is a device intended for medical purposes for use in a test system to establish points of reference that are used in the determination of values in the measurement of substances in human specimens.
The COBAS INTEGRA test applications contained in this submission are intended for use with the COBAS INTEGRA Analyzer, which is also known as the COBAS INTEGRA 700. The COBAS INTEGRA Analyzer and COBAS INTEGRA Reagent cassettes together provide an integrated system for in vitro diagnostic testing. The COBAS INTEGRA Analyzer utilizes three measuring principles, i.e., absorbance, fluorescence polarization and ion-selective electrodes. The analyzer has a throughput of up to 600 tests per hour with STAT samples prioritized and tested immediately. Random sample access, robotics and a user interface optimize time management and streamline workflow. The COBAS INTEGRA can store up to 68 COBAS INTEGRA Reagent Cassettes on board, 24 hours a day at 2-8°C. The COBAS INTEGRA Reagent Cassettes are compact and preparation-free with the added convenience of long term on-board stability. Barcode readers are used to identify newly loaded reagent cassettes, samples for patient identification, and rack inserts and to read calibration and control data from the cassette label. COBAS INTEGRA tests include chemistry, drugs of abuse, immunology, ion selective electrodes, therapeutic drug monitoring, and hematology reagents.
The cassette COBAS INTEGRA LDL Direct (LDL-D) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA 700 for the quantitative determination of LDL-cholesterol direct concentration in serum and plasma.
The Roche Calibrator LDL Direct is intended for use as calibrator in quantitative Low Density Lipoprotein cholesterol assays. It is recommended for use with LDL Direct reagents on COBAS® chemistry systems.
Here's a breakdown of the acceptance criteria and study information for the Roche COBAS® INTEGRA LDL Direct and Roche Calibrator LDL Direct, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
The document doesn't explicitly state "acceptance criteria" as a separate, pre-defined set of thresholds. Instead, it presents performance characteristics of the new device and compares them to those of predicate devices to demonstrate substantial equivalence. The implication is that if the new device's performance aligns with or is comparable to the predicate devices, it meets the "acceptance criteria" for substantial equivalence.
| Performance Characteristic | Acceptance Criteria (Implied by Predicate Performance) | Reported Device Performance (COBAS INTEGRA LDL Direct) |
|---|---|---|
| Intended Use (Reagent) | Quantitative determination of LDL-cholesterol direct concentration in serum or plasma, useful in assessing coronary heart disease risk. | Quantitative determination of LDL-cholesterol direct concentration in serum and plasma, useful in assessing coronary heart disease risk. |
| Intended Use (Calibrator) | Calibrator for quantitative Low Density Lipoprotein cholesterol assays. | Calibrator for quantitative Low Density Lipoprotein cholesterol assays on COBAS chemistry systems. |
| Matrix (Calibrator) | Human serum | Human serum |
| Approx. Value (Calibrator) | 51.4 mg/dL | 2.85 mmol/L (110 mg/dL) |
| Precision (Level 1 Mean) | Not explicitly stated (predicate data not available for direct comparison) | 2.72 mmol/L (105 mg/dL) |
| Precision (Level 1 Within Run CV) | Not explicitly stated | 1.4 |
| Precision (Level 1 Total CV) | Not explicitly stated | 1.9 |
| Precision (Level 2 Mean) | Not explicitly stated | 5.12 mmol/L (198 mg/dL) |
| Precision (Level 2 Within Run CV) | Not explicitly stated | 1.8 |
| Precision (Level 2 Total CV) | Not explicitly stated | 2.1 |
| Linearity | 500 mg/dL | 14.0 mmol/L (540 mg/dL) |
| Accuracy (Correlation Coefficient vs. COBAS MIRA) | Not explicitly stated (predicate comparison is to Beta-quantification or Friedewald) | 0.964 |
| Accuracy (Linear Regression vs. COBAS MIRA) | Not explicitly stated | y = 0.85x + 0.7 mmol/L |
Note: For accuracy, the new device is compared to COBAS MIRA, Beta-quantification, and Friedewald formula. The predicate device's accuracy is provided against Beta-quantification. The document implies that a strong correlation (e.g., r > 0.95) and a linear regression close to y=x would be acceptable for accuracy.
2. Sample Size Used for the Test Set and Data Provenance
The document refers to "clinical and nonclinical studies."
-
Test Set Sample Size for Accuracy:
- Against COBAS MIRA: n = 276
- Against Beta-quantification: n = 150
- Against Friedewald formula: n = 276
-
Data Provenance: The document does not explicitly state the country of origin or whether the data was retrospective or prospective.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications of Those Experts
This type of information is not provided in the document. For an in vitro diagnostic device like this, ground truth is typically established through reference methods or established laboratory procedures, not by human experts in the way it would be for an imaging AI.
4. Adjudication Method for the Test Set
This information is not applicable and therefore not provided. Adjudication methods (like 2+1, 3+1) are common in clinical trials involving subjective interpretations (e.g., imaging reads) to resolve discrepancies. For a quantitative diagnostic test like LDL-cholesterol, discrepancies would typically be resolved by retesting, using a definitive reference method, or investigating pre-analytical/analytical errors.
5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size of Human Readers Improvement with AI vs. Without AI Assistance
This information is not applicable and therefore not provided. MRMC studies are relevant for imaging devices or AI tools where human readers are interpreting data, and the AI's role is to assist or augment their performance. The Roche COBAS INTEGRA LDL Direct is a laboratory diagnostic assay, not one that involves human interpretation of "cases" in an MRMC context.
6. If a Standalone Study (Algorithm Only Without Human-in-the-Loop Performance) Was Done
Yes, this is essentially a standalone study. The document describes the performance of the COBAS INTEGRA LDL Direct system itself (the reagent cassette + analyzer) in measuring LDL-cholesterol directly. There isn't a human-in-the-loop component described for the function of this diagnostic device.
7. The Type of Ground Truth Used
The ground truth for the accuracy studies was established by comparing the COBAS INTEGRA LDL Direct results against:
- COBAS MIRA: Another automated chemistry analyzer, likely acting as a comparative reference method.
- Beta-quantification: A established reference method for lipoprotein analysis, considered highly accurate for LDL-cholesterol.
- Friedewald formula: A calculated estimate of LDL-cholesterol based on other lipid measurements (total cholesterol, HDL-cholesterol, triglycerides).
8. The Sample Size for the Training Set
The document does not mention a "training set". This is an in vitro diagnostic device, not an AI/machine learning algorithm in the typical sense that would require a distinct training set for model development. The development process for an IVD involves formulation, optimization, and verification, rather than "training" with a dataset for an algorithm.
9. How the Ground Truth for the Training Set Was Established
Since there is no "training set" mentioned or implied for this type of device, this information is not applicable.
Ask a specific question about this device
(82 days)
The cassette COBAS INTEGRA x-1-Antitrypsin (AAT) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA 700 for the quantitative immunological determination of human x-1-antitrypsin in serum and plasma. The measurements aid in the diagnosis of several conditions including juvenile and adult cirrhosis of the liver. In addition, a-1-antitrypsin deficiency has been associated with pulmonary emphysema.
The cassette COBAS INTEGRA Immunoglobulin A (IGA/IGAP) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA 700 for the quantitative immunological determination of human immunoglobulin A in serum and plasma. addition to the standard application (IGA), the sensitive application (IGAP) is designed for the quantitative determination of low IgA concentrations in e.g. pediatric samples. Measurement of this immunoglobulin aids in the diagnosis of abnormal protein metabolism and the body's lack of ability to resist infectious agents.
The cassette COBAS INTEGRA Immunoglobulin M (IGM/IGMP) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA 700 for the quantitative immunological determination of human immunoglobulin M in serum and plasma. In addition to the standard application (IGM), the sensitive application (IGMP) is designed for the quantitative determination of low IgM concentrations in e.g. pediatric samples. Measurement of this immunoglobulin aids in the diagnosis of abnormal protein metabolism and the body's lack of ability to resist infectious agents.
The cassette COBAS INTEGRA Immunoglobulin G (Turbidimetric) (IGGT/IGGTC) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA 700 for the quantitative immunological determination of human immunoglobulin G in serum, plasma (IGGT) and cerebrospinal fluid (IGGTC). Measurement of this immunoglobulin aids in the diagnosis of abnormal protein metabolism and the body's lack of ability to resist infectious agents.
The COBAS INTEGRA test applications contained in this submission are intended for use with the COBAS INTEGRA Analyzer, which is also known as the COBAS INTEGRA 700. The COBAS INTEGRA Analyzer and COBAS INTEGRA Reagent cassettes together provide an integrated system for in vitro diagnostic testing. The COBAS INTEGRA Analyzer utilizes three measuring principles, i.e., absorbance, fluorescence polarization and ion-selective electrodes. The analyzer has a throughput of up to 600 tests per hour with STAT samples prioritized and tested immediately. Random sample access, robotics and a user interface optimize time management and streamline workflow. The COBAS INTEGRA can store up to 68 COBAS INTEGRA Reagent Cassettes on board, 24 hours a day at 2-8°C. The COBAS INTEGRA Reagent Cassettes are compact and preparation-free with the added convenience of long term on-board stability. Barcode readers are used to identify newly loaded reagent cassettes, samples for patient identification, and rack inserts and to read calibration and control data from the cassette label. COBAS INTEGRA tests include chemistry, drugs of abuse, immunology, ion selective electrodes, therapeutic drug monitoring, and hematology reagents.
Here's a breakdown of the acceptance criteria and study information for the Roche COBAS® INTEGRA Reagent Cassettes for α-1-Antitrypsin (AAT), Immunoglobulin A (IGA/IGAP), and Immunoglobulin M (IGM/IGMP), based on the provided text:
Important Note: The provided document is a 510(k) Summary. This type of summary focuses on demonstrating substantial equivalence to a legally marketed predicate device, rather than proving the device meets specific pre-defined acceptance criteria in the same way a novel device might. The "acceptance criteria" here are inferred from the performance characteristics presented to show equivalence. The studies are primarily comparative studies against predicate devices.
1. Table of Acceptance Criteria and Reported Device Performance
Device: Roche COBAS® INTEGRA Reagent Cassettes (AAT, IGA/IGAP, IGM/IGMP)
| Performance Characteristic | Acceptance Criteria (inferred from predicate/previous version) | Reported Device Performance (Modified COBAS INTEGRA) | Predicate Device Performance (where available) |
|---|---|---|---|
| α-1-Antitrypsin (AAT) | |||
| Accuracy (Corr. Coeff. (r)) | Must be comparable to predicate (e.g., K972640: 0.967 / K954992: 0.930) | 0.995 (vs. BM) | Predicate K972640: 0.967; Cleared K954992: 0.930 |
| Linear Regression | Must be comparable to predicate | 1.30x - 0.31 g/L | Predicate K972640: 0.993x + 9.9 mg/dL; Cleared K954992: 0.90x + 0.06 g/L |
| Immunoglobulin A (IGA/IGAP) | |||
| Precision (IGA) | |||
| Level 1 (Mean ~2.0-2.3 g/L) | Within-run CV comparable to predicate/cleared (e.g., BM: 0.9%, Cleared: 1.4%) | 2.0% | Predicate BM: 0.9%; Cleared K954457: 1.4% |
| Level 2 (Mean ~3.5-6.2 g/L) | Within-run CV comparable to predicate/cleared (e.g., BM: 0.8%, Cleared: 0.81%) | 0.97% | Predicate BM: 0.8%; Cleared K954457: 0.81% |
| Total CV Level 1 | Total CV comparable to predicate/cleared (e.g., BM: 2.2%, Cleared: 2.8%) | 2.3% | Predicate BM: 2.2%; Cleared K954457: 2.8% |
| Total CV Level 2 | Total CV comparable to predicate/cleared (e.g., BM: 1.8%, Cleared: 1.8%) | 1.2% | Predicate BM: 1.8%; Cleared K954457: 1.8% |
| Accuracy (Corr. Coeff. (r)) | Must be comparable to predicate (e.g., BM: 0.99 / Cleared: 0.989) | 0.994 (vs. BM/Hitachi) | Predicate BM: 0.99; Cleared K954457: 0.989 |
| Linear Regression | Must be comparable to predicate | 1.023x - 0.214 g/L | Predicate BM: 0.83x + 20.6 mg/dL; Cleared K954457: 0.97x - 0.05 g/L |
| Assay Range | Must be comparable or improved | 0.45 - 7.3 g/L (std); 0.15 - 98.6 g/L (rerun) | Cleared K954457: 0.79 - 12.6 g/L (std); 0.27 - 30.2 g/L (rerun) |
| Sensitivity | Must be comparable or improved | 0.45 g/L | Cleared K954457: 0.79 g/L |
| Immunoglobulin M (IGM/IGMP) | |||
| Precision (IGM) | |||
| Level 1 (Mean ~0.55-0.6 g/L) | Within-run CV comparable to predicate/cleared (e.g., BM: 0.79%, Cleared: 2.6%) | 2.4% | Predicate BM: 0.79%; Cleared K954457: 2.6% |
| Level 2 (Mean ~1.9-2.0 g/L) | Within-run CV comparable to predicate/cleared (e.g., BM: 0.8%, Cleared: 2.0%) | 1.6% | Predicate BM: 0.8%; Cleared K954457: 2.0% |
| Total CV Level 1 | Total CV comparable to predicate/cleared (e.g., BM: 3.7%, Cleared: 3.1%) | 3.2% | Predicate BM: 3.7%; Cleared K954457: 3.1% |
| Total CV Level 2 | Total CV comparable to predicate/cleared (e.g., BM: 2.4%, Cleared: 2.2%) | 1.9% | Predicate BM: 2.4%; Cleared K954457: 2.2% |
| Accuracy (Corr. Coeff. (r)) | Must be comparable to predicate (e.g., BM: 0.979 / Cleared: 0.994) | 0.994 (vs. BM/Hitachi) | Predicate BM: 0.979; Cleared K954457: 0.994 |
| Linear Regression | Must be comparable to predicate | 1.293x + 0.341 g/L | Predicate BM: 0.81x + 5.9 mg/dL; Cleared K954457: 1.12x - 0.06 g/L |
| Assay Range | Must be comparable or improved | 0.16 - 5.18 g/L (std); 0.05 - 24.35 g/L (rerun) | Cleared K954457: 0.31 - 5.0 g/L (std); 0.11 - 12.1 g/L (rerun) |
| Sensitivity | Must be comparable or improved | 0.16 g/L | Cleared K954457: 0.31 g/L |
2. Sample Sizes Used for the Test Set and Data Provenance
- α-1-Antitrypsin (AAT):
- Sample Size: 288
- Data Provenance: Not explicitly stated but implied to be clinical samples (serum/plasma). Origin (e.g., country) and retrospective/prospective nature are not specified.
- Immunoglobulin A (IGA/IGAP):
- Sample Size: 584
- Data Provenance: Not explicitly stated but implied to be clinical samples (serum/plasma). Origin (e.g., country) and retrospective/prospective nature are not specified.
- Immunoglobulin M (IGM/IGMP):
- Sample Size: 556
- Data Provenance: Not explicitly stated but implied to be clinical samples (serum/plasma). Origin (e.g., country) and retrospective/prospective nature are not specified.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
Not applicable. This device is an in vitro diagnostic reagent, and the 'ground truth' is established by comparative measurements against predicate devices using laboratory methods, not human expert interpretation of images or clinical cases. The comparison is against established laboratory testing methods.
4. Adjudication Method for the Test Set
Not applicable. The 'test set' here consists of biological samples measured by the device and compared to results from predicate devices/methods. There is no human adjudication process described.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done
No. MRMC studies are typically for imaging devices or AI tools that assist human readers in interpretation. This is an IVD reagent, and the effectiveness is evaluated through analytical performance characteristics like accuracy, precision, and linearity when compared to established methods.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
Yes, in a sense. The studies summarized are standalone performance evaluations of the reagent cassette on the COBAS INTEGRA Analyzer. The results are generated by the automated system, without "human-in-the-loop" needing to interpret the primary measurement, although human operators load samples and review results. The comparison is between the modified reagent's performance and that of other established laboratory methods (predicate devices).
7. The Type of Ground Truth Used
The ground truth used for these studies is comparative measurement data from legally marketed predicate devices and established laboratory methods. Specifically:
- AAT: Comparison against Boehringer Mannheim α-1-Antitrypsin Reagent (K972640) and Cleared COBAS INTEGRA α-1-Antitrypsin (K954992). The "vs. BM" implies using the Boehringer Mannheim method as a reference.
- IGA/IGAP: Comparison against Boehringer Mannheim Immunoglobulin A Reagent (K955907) and Cleared COBAS INTEGRA Immunoglobulin A (K954457). The "vs. BM/Hitachi" indicates comparison with these laboratory systems.
- IGM/IGMP: Comparison against Boehringer Mannheim Immunoglobulin M Reagent (K955908) and Cleared COBAS INTEGRA Immunoglobulin M (K954457). The "vs. BM/Hitachi" indicates comparison with these laboratory systems.
8. The Sample Size for the Training Set
Not applicable. These are reagent cassettes used with an existing analyzer. There is no mention of machine learning or AI models with "training sets" in the context of this 510(k) submission. The development and optimization of the reagent formulations and assay parameters would be part of a different internal R&D process, not typically described as a "training set" in this context.
9. How the Ground Truth for the Training Set Was Established
Not applicable, as there is no "training set" in the context of this device and submission. The "ground truth" for evaluating the performance of these reagents is established through established analytical chemistry and immunology principles, validated against reference materials, and benchmarked against predicate diagnostic assays.
Ask a specific question about this device
The cassette COBAS INTEGRA Alkaline Phosphatase IFCC liquid (ALPL2 and ALPL6) contain an in vitro diagnostic reagent system intended for use on COBAS INTEGRA for the quantitative determination of the catalytic activity of alkaline phosphatase in serum and plasma. Measurements of alkaline phosphatase or its isoenzymes are used in the diagnosis and treatment of liver, bone, parathyroid, and intestinal diseases.
The cassette COBAS INTEGRA a-Amylase EPS Pancreatic (AMY-P / AMYUP) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA for the quantitative determination of the catalytic activity of pancreatic or-amylase in serum, plasma, and urine. Amylase measurements are used primarily for the diagnosis and treatment of pancreatitis (inflammation of the pancreas).
The cassette COBAS INTEGRA C-reactive Protein (Latex), (CRPLX) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA for the quantitative immunological determination of human C-reactive protein in serum and plasma. Measurements of C-reactive protein aids in evaluation of the amount of injury to tissue.
The COBAS INTEGRA test applications contained in this submission are intended for use with the COBAS INTEGRA Analyzer, which is also known as the COBAS INTEGRA 700. The COBAS INTEGRA Analyzer and COBAS INTEGRA Reagent cassettes together provide an integrated system for in vitro diagnostic testing. The COBAS INTEGRA Analyzer utilizes three measuring principles, i.e., absorbance, fluorescence polarization and ion-selective electrodes. The analyzer has a throughput of up to 600 tests per hour with STAT samples prioritized and tested immediately. Random sample access, robotics and a user interface optimize time management and streamline workflow. The COBAS INTEGRA can store up to 68 COBAS INTEGRA Reagent Cassettes on board, 24 hours a day at 2-8°C. The COBAS INTEGRA Reagent Cassettes are compact and preparation-free with the added convenience of long term on-board stability. Barcode readers are used to identify newly loaded reagent cassettes, samples for patient identification, and rack inserts and to read calibration and control data from the cassette label. COBAS INTEGRA tests include chemistry, drugs of abuse, immunology, ion selective electrodes, therapeutic drug monitoring, and hematology reagents. Through this submission, it is the intention of Roche to gain clearance for three additional COBAS Reagent Cassettes. These are the COBAS INTEGRA Alkaline Phosphatase IFCC liquid (ALPL2 and ALPL6), the COBAS INTEGRA a-Amylase EPS Pancreatic (AMY-P / AMYUP), and the COBAS INTEGRA C-Reactive Protein (Latex), (CRPLX).
The provided text is a 510(k) Summary for Roche COBAS® INTEGRA Reagent Cassettes, detailing the characteristics and performance of three new reagent cassettes (Alkaline Phosphatase IFCC liquid, α-Amylase EPS Pancreatic, and C-Reactive Protein (Latex)) compared to their legally marketed predicate devices.
Here's an analysis of the acceptance criteria and study information:
1. Table of Acceptance Criteria and Reported Device Performance
The documents do not explicitly state "acceptance criteria" as a set of predefined thresholds that the device must meet in the same way modern AI/ML device submissions would. Instead, the "Performance Characteristics" section for each proposed device is presented alongside the performance of its predicate device, implying that equivalence in these characteristics is the underlying acceptance criterion for substantial equivalence.
Since specific acceptance criteria are not called out, I will present the reported performance characteristics for the new devices. The implicit acceptance criterion is "comparable performance to the predicate device."
| Performance Characteristic | COBAS INTEGRA Alkaline Phosphatase IFCC liquid (ALPL2 & ALPL6) (Proposed) | COBAS INTEGRA α-Amylase EPS Pancreatic (AMY-P / AMYUP) (Proposed) | COBAS INTEGRA C-Reactive Protein (Latex) (CRPLX) (Proposed) | Implicit Acceptance Criteria (Comparable to Predicate) |
|---|---|---|---|---|
| Intended Use | Quantitative determination of catalytic activity of alkaline phosphatase in serum and plasma | Quantitative determination of catalytic activity of pancreatic α-amylase in serum, plasma, and urine | Quantitative immunological determination of human C-reactive protein in serum and plasma | Must align with predicate device's intended use |
| Methodology | Enzymatic colorimetric using 4-Nitrophenylphosphate (IFCC) | Enzymatic colorimetric using substrate 4,6-ethylidene-p-nitrophenyl-α-D-malto-heptaoside | Particle enhanced immunoturbidimetric | Must be similar to or provide equivalent results to predicate device's methodology |
| Sample Type | Serum and Plasma | Serum, plasma, and urine | Serum and Plasma | Must be identical to predicate |
| Calibrator | Roche Calibrator (human) (K942706) | Roche Calibrator (human) (K942706) | CRP T Standard (K951595) | Must be compatible and provide comparable calibration to predicate |
| Controls | Roche Control Serum N and P (human) (K972214) | Roche Control Serum N and P (human) (K972214) | CRP T Control (K954992), CRP T N Control (Exempt) | Must be compatible and provide comparable control performance to predicate |
| Reagents | AMP in vial B (liquid); 4-Nitrophenylphosphate in vial C (liquid) | -enzyme, two monoclonal antibodies (mouse) in vial A (liquid); -substrate in vial C (liquid) | R1 BSA and immunoglobulins (mouse), vials A & B; R2 Latex particles coated with anti-CRP (mouse), vial C (liquid) | Must be chemically and functionally similar to predicate's reagents |
| Assay Range | 0-1500 U/L | Serum/Plasma and Urine: 0 - 1500 U/L | 0 - 160 mg/L | Similar or improved range compared to predicate |
| Sensitivity | 3.7 x 10^-4 ΔA/min per U/L of ALP | Serum/Plasma and Urine: 2.8 x 10^-4 ΔA/min per U/L of pancreatic α-amylase | 0.25 mg/L | Similar or improved sensitivity compared to predicate |
| Precision (Within-run %CV) | Level 1: 2.3; Level 2: 0.55 | Serum Level 1: 1.2; Serum Level 2: 0.91; Urine Level 1: 1.094; Urine Level 2: 0.8796 | Level 1: 1.8; Level 2: 1.5 | |
| (Second Set) Level 1: 2.0; Level 2: 2.4 | %CV values should be within acceptable clinical laboratory limits and comparable to predicate | |||
| Precision (Total %CV) | Level 1: 2.7; Level 2: 1.3 | Serum Level 1: 1.7; Serum Level 2: 1.6; Urine Level 1: 1.2; Urine Level 2: 1.1 | Level 1: 2.9; Level 2: 2.7 | |
| (Second Set) Level 1: 2.5; Level 2: 2.4 | %CV values should be within acceptable clinical laboratory limits and comparable to predicate | |||
| Accuracy (Corr. Coeff. (r)) | 0.999 | Serum: 0.999; Urine: 0.999 | 0.993 | Correlation coefficient should be high (close to 1), indicating strong agreement with predicate or reference method |
| Accuracy (Linear Regression) | 1.00x + 0.15 U/L | Serum: 1.03x + 0.3 U/L; Urine: 1.00x + 1.4 U/L | 1.07x - 6.2 mg/L | Slope should be close to 1 and y-intercept close to 0, indicating minimal bias relative to predicate or reference method |
2. Sample Size Used for the Test Set and Data Provenance
The "test set" in this context refers to the clinical samples used to evaluate the performance characteristics (precision and accuracy) of the new reagent cassettes.
-
COBAS INTEGRA Alkaline Phosphatase IFCC liquid (ALPL2 and ALPL6):
- Accuracy Sample Size (n): 252
- Data Provenance: Not specified, but generally for in vitro diagnostic (IVD) devices, these studies use patient samples collected in a clinical laboratory setting. The origin (country) and retrospective/prospective nature are not detailed in this summary.
-
COBAS INTEGRA α-Amylase EPS Pancreatic (AMY-P / AMYUP):
- Accuracy Sample Size (n):
- Serum: 246
- Urine: 106
- Data Provenance: Not specified, same general assumptions as above.
- Accuracy Sample Size (n):
-
COBAS INTEGRA C-Reactive Protein (Latex) (CRPLX):
- Accuracy Sample Size (n): 244
- Data Provenance: Not specified, same general assumptions as above.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications
This information is not provided in the document. For IVD devices like these, "ground truth" for method comparisons is typically established by running the same samples on a legally marketed predicate device or a highly accurate reference method. It does not involve human expert consensus in the way image-based diagnostics might.
4. Adjudication Method for the Test Set
This is not applicable and therefore not provided. Adjudication methods like 2+1 or 3+1 are used in studies where human readers interpret data, and their disagreements need to be resolved. For quantitative chemical assays, the measurement itself is the "reader," and comparison is made to another assay.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done, and the effect size of how much human readers improve with AI vs without AI assistance
This is not applicable as the device is an in-vitro diagnostic reagent cassette for automated analyzers, not an AI or imaging device involving human readers.
6. If a Standalone (i.e. algorithm only without human-in-the loop performance) was done
Yes, the performance data presented (assay range, sensitivity, precision, accuracy) represent the standalone performance of the reagent cassettes on the COBAS INTEGRA Analyzer. There is no human-in-the-loop component being evaluated for the measurement itself, only for the instrument's operation and result interpretation by laboratory personnel.
7. The Type of Ground Truth Used
For these types of in vitro diagnostic assays, the "ground truth" for the accuracy studies is implicitly established by comparing the results from the new device to those obtained from the legally marketed predicate device (or a recognized reference method). The correlation coefficients and linear regression data are direct comparisons to the predicate. For example, the accuracy section "Corr. Coefficient (r)" and "Linear regression" are direct comparisons between the proposed device and the cleared predicate device.
8. The Sample Size for the Training Set
This information is not provided and is likely not applicable in the same way it would be for an AI/ML device. For chemical reagents, "training" typically refers to the R&D and optimization phase to develop the reagent formulation and instrument application. It's not a discrete "training set" of patient data used to optimize an algorithm parameters that are then locked.
9. How the Ground Truth for the Training Set was Established
Similar to point 8, this is not provided and likely not applicable in the context of an AI/ML device. The "ground truth" during reagent development would involve analytical methods to confirm the chemical properties and performance of the reagents.
Ask a specific question about this device
(27 days)
The cassette COBAS INTEGRA Serum Barbiturates contains an in vitro diagnostic reagent system intended for use on the COBAS INTEGRA analyzer for the detection of barbiturates and their metabolites in human serum or heparinized plasma. This reagent system is intended for use in toxicological screenings where the analytical result is used in the management of barbiturate use or overdose.
The Abuscreen ONLINE Serum Barbiturates Calibrators are designed for calibration of the cassette COBAS INTEGRA Serum Barbiturates on the COBAS INTEGRA chemistry systems. The calibrators are used in the determination of values in the measurement of barbiturates in human specimens.
The cassette COBAS INTEGRA Serum Benzodiazepines contains an in vitro diagnostic reagent system intended for use on the COBAS INTEGRA analyzer for the detection of benzodiazepines and their metabolites in human serum or heparinized plasma. This reagent system is intended for use in toxicological screenings where the analytical result is used in the management of benzodiazepine use or overdose.
The Abuscreen ONLINE Serum Benzodiazepines Calibrators are designed for calibration of the cassette COBAS INTEGRA Serum Benzodiazepines on the COBAS INTEGRA chemistry systems. The calibrators are used in the determination of values in the measurement of benzodiazepines in human specimens.
The COBAS INTEGRA test applications contained in this submission are intended for use with the COBAS INTEGRA Analyzer, which is also known as the COBAS INTEGRA 700. The COBAS INTEGRA Analyzer and COBAS INTEGRA Reagent cassettes together provide an integrated system for in vitro diagnostic testing. The COBAS INTEGRA Analyzer utilizes three measuring principles, i.e., absorbance, fluorescence polarization and ion-selective electrodes. The analyzer has a throughput of up to 600 tests per hour with STAT samples prioritized and tested immediately. Random sample access, robotics and a user interface optimize time management and streamline workflow. The COBAS INTEGRA can store up to 68 COBAS INTEGRA Reagent Cassettes on board, 24 hours a day at 2-8°C. The COBAS INTEGRA Reagent Cassettes are compact and preparation-free with the added convenience of long term on-board stability. Barcode readers are used to identify newly loaded reagent cassettes, samples for patient identification, and rack inserts and to read calibration and control data from the cassette label. COBAS INTEGRA tests include chemistry, drugs of abuse, immunology, ion selective electrodes, therapeutic drug monitoring, and hematology reagents. Through this submission, it is the intention of Roche to gain clearance for two additional COBAS Reagent Cassettes and their associated calibrator sets. These are: the COBAS INTEGRA Serum Barbiturates (SBARB) Reagent Cassette, the Abuscreen ONLINE Serum Barbiturates Calibrators, the COBAS INTEGRA Serum Benzodiazepines (SBENZ) Reagent Cassette, and the Abuscreen ONLINE Serum Benzodiazepines Calibrators.
Here's an analysis of the acceptance criteria and study information for the Roche COBAS® INTEGRA Reagent Cassettes and Calibrators, based on the provided text:
Important Note: The provided document is a 510(k) Summary for a medical device cleared in 1998. It focuses on demonstrating substantial equivalence to predicate devices rather than establishing novel safety and effectiveness through a standalone, comprehensive clinical trial with pre-defined acceptance criteria in the way a new, high-risk device might today. The "acceptance criteria" detailed below are derived from the performance characteristics presented for comparison with the predicate device. The study is primarily a comparative effectiveness study against the predicate, alongside internal analytical validation.
1. Table of Acceptance Criteria and Reported Device Performance
The document doesn't explicitly state "acceptance criteria" in a pass/fail format with specific targets that the device must meet in a prospective study. Instead, it presents performance characteristics of the new device alongside those of its predicate for comparison, implying that the new device's performance should be comparable or substantially equivalent to the predicate. The "reported device performance" refers to the values presented for the COBAS INTEGRA system.
A. COBAS INTEGRA Serum Barbiturates (SBARB)
| Characteristic | Acceptance Criteria (Implied from Predicate) | Reported Device Performance (COBAS INTEGRA SBARB) |
|---|---|---|
| Intended Use | Detection of barbiturates/metabolites in human serum/plasma for toxicological screening, diagnosis, and treatment of use/overdose. | Detection of barbiturates/metabolites in human serum/heparinized plasma for toxicological screening, management of use/overdose. |
| Methodology | Fluorescence polarization | Fluorescence polarization |
| Sample Type | Serum and Plasma | Serum and Plasma |
| Assay Range | 0.70 - 40 µg/mL | 0.03 - 4 µg/mL (0.03 - 80 µg/mL with postdilution) |
| Cutoff Conc. | 2.0 µg/mL | 0.5 µg/mL |
| Sensitivity | 0.07 µg/mL | 0.03 µg/mL |
| Precision (Within-run % CV) | Level 1: 3.18%, Level 2: 3.47%, Level 3: 4.38% | Level 1: 4.1%, Level 2: 4.1%, Level 3: 2.3% |
| Precision (Total % CV) | Level 1: 4.00%, Level 2: 3.92%, Level 3: 4.73% | Level 1: 5.0%, Level 2: 4.2%, Level 3: 2.9% |
| Accuracy (Agreement vs GC/MS) | 79 positive, 14 negative (total 93 cases) | 37 positive, 10 negative (total 47 cases) |
| Reproducibility (Agreement vs Predicate ADX) | 69 positive, 123 negative | 35 positive, 12 negative |
B. COBAS INTEGRA Serum Benzodiazepines (SBENZ)
| Characteristic | Acceptance Criteria (Implied from Predicate) | Reported Device Performance (COBAS INTEGRA SBENZ) |
|---|---|---|
| Intended Use | Detection of benzodiazepines/metabolites in human serum/plasma for toxicological screening, diagnosis, and treatment of use/overdose. | Detection of benzodiazepines/metabolites in human serum/heparinized plasma for toxicological screening, management of use/overdose. |
| Methodology | Fluorescence polarization | Fluorescence polarization |
| Sample Type | Serum and Plasma | Serum and Plasma |
| Assay Range | 12 - 1000 ng/mL | 3 - 200 ng/mL (3 - 2000 ng/mL with postdilution) |
| Cutoff Conc. | 12.0 ng/mL | 3 ng/mL |
| Sensitivity | 12.0 ng/mL | 3 ng/mL |
| Precision (Within-run % CV) | Level 1: 3.19%, Level 2: 1.86%, Level 3: 3.41% | Level 1: 5.5%, Level 2: 1.9%, Level 3: 1.1% |
| Precision (Total % CV) | Level 1: 4.98%, Level 2: 3.73%, Level 3: 5.98% | Level 1: 5.4%, Level 2: 2.7%, Level 3: 2.0% |
| Accuracy (Agreement vs GC/MS) | 76 positive, 38 negative (total 114 cases) | 46 positive, 28 negative (total 74 cases) |
| Reproducibility (Agreement vs Predicate TDX) | 76 positive, 83 negative | 46 positive, 20 negative |
2. Sample Size Used for the Test Set and the Data Provenance
-
COBAS INTEGRA Serum Barbiturates (SBARB):
- Accuracy (compared to GC/MS): 47 samples (37 positive, 10 negative)
- Accuracy (compared to Predicate ADX): 47 samples (35 positive, 12 negative)
- Precision: Not explicitly stated, but typically involves multiple replicates across several runs for each level tested. Given the three levels and %CV calculation, a common practice would be 20 replicates for 2-3 levels over 2-3 days, resulting in at least 40-60 measurements per level. The exact number of individual patient samples used to derive these mean and CV values is not specified, but these are typically control or spiked samples.
- Data Provenance: Not explicitly stated (e.g., country of origin). The studies appear to be retrospective analytical performance evaluations conducted by the manufacturer as part of the submission to demonstrate equivalence.
-
COBAS INTEGRA Serum Benzodiazepines (SBENZ):
- Accuracy (compared to GC/MS): 74 samples (46 positive, 28 negative)
- Accuracy (compared to Predicate TDX): 66 samples (46 positive, 20 negative)
- Precision: Similar to SBARB, not explicitly stated, but common practice would imply multiple replicates for each of the three levels tested.
- Data Provenance: Not explicitly stated. Likely retrospective analytical performance evaluation.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and the Qualifications of Those Experts
- For both assays (Barbiturates and Benzodiazepines), the ground truth for accuracy was established using GC/MS (Gas Chromatography/Mass Spectrometry). GC/MS is considered a "gold standard" analytical method for drug confirmation and quantification in toxicology.
- The document does not specify the number of human experts involved in interpreting the GC/MS results or their qualifications. The interpretation of GC/MS data is typically performed by trained laboratory personnel (e.g., analytical chemists, toxicologists) within a certified laboratory environment.
4. Adjudication Method for the Test Set
- For the accuracy studies using GC/MS as the ground truth, there is no mention of an adjudication method in the traditional sense (e.g., 2+1 physician review). The GC/MS result itself serves as the definitive determination. Discrepancies between the device and GC/MS would be investigated analytically rather than through expert consensus on the clinical classification.
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 submission concerns in vitro diagnostic (IVD) reagent assays, which are standalone laboratory tests, not imaging devices or AI-assisted diagnostic tools that involve human readers. Therefore, the concept of "human readers improving with AI assistance" is not applicable here.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
- Yes, this entire submission effectively represents a standalone performance evaluation. The COBAS INTEGRA system, with these reagent cassettes, performs automated analysis. Its performance characteristics (precision, accuracy, sensitivity, assay range, cutoff) are measured directly based on the analytical results produced by the instrument and reagents, without immediate human intervention in the result generation itself. Human interaction would occur in sample loading, result review, and clinical interpretation.
7. The Type of Ground Truth Used (expert consensus, pathology, outcomes data, etc.)
- The primary ground truth for the accuracy claims was:
- GC/MS (Gas Chromatography/Mass Spectrometry): This is a highly specific and sensitive analytical method for confirming and quantifying drugs, serving as the gold standard in toxicology for this type of test.
8. The Sample Size for the Training Set
- The document does not specify the sample size for a training set. For IVD assays based on established chemical principles (like fluorescence polarization immunoassay), there isn't a "training set" in the same way there would be for a machine learning or AI algorithm. The assay's parameters (e.g., antibody concentrations, reaction times, calibration curves) are developed and optimized by the manufacturer using internal R&D processes, but these are not explicitly detailed as a "training set" in the context of this 510(k) summary. The performance data presented are for the final, developed product.
9. How the Ground Truth for the Training Set Was Established
- As a "training set" is not explicitly mentioned or applicable in the AI/machine learning sense for this device, the question of how its ground truth was established is not directly answerable from the provided text. The development and optimization of such assays rely on well-characterized samples (e.g., spiked samples, confirmed positive/negative clinical samples, reference materials) to establish robust analytical performance, but these are part of the R&D process rather than a formalized "training set" with ground truth in the context of this regulatory submission.
Ask a specific question about this device
(156 days)
The cassette COBAS INTEGRA Acid / Prostatic Phosphatase contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA for the quantitative determination of the catalytic activity of total and prostatic acid phosphatase in serum.
The cassette Roche COBAS INTEGRA Benzodiazepines contains an in vitro diagnostic reagent system intended for use on the COBAS INTEGRA for the semi-quantitative detection of benzodiazepines in human urine using the enzyme ß-glucuronidase.
The COBAS INTEGRA Analyzer and COBAS INTEGRA Reagent cassettes together provide an integrated system for in vitro diagnostic testing. The COBAS INTEGRA Analyzer utilizes three measuring principles, i.e., absorbance, fluorescence polarization and ion-selective electrodes. The analyzer has a throughput of up to 600 tests per hour with STAT samples prioritized and tested immediately. Random sample access, robotics and a user interface optimize time management and streamline workflow. The COBAS INTEGRA can store up to 68 COBAS INTEGRA Reagent Cassettes on board, 24 hours a day at 2-8°C. The COBAS INTEGRA Reagent Cassettes are compact and preparation-free with the added convenience of long term on-board stability. Barcode readers are used to identify newly loaded reagent cassettes, samples for patient identification, and rack inserts and to read calibration and control data from the cassette label. COBAS INTEGRA tests include chemistry, drugs of abuse, immunology, ion selective electrodes, therapeutic drug monitoring, and hematology reagents.
The provided text describes two in vitro diagnostic reagent systems: COBAS INTEGRA Acid/Prostatic Phosphatase (ACPP) and COBAS INTEGRA Benzodiazepines with β-glucuronidase (BNZGL). The 510(k) summary focuses on demonstrating their substantial equivalence to previously marketed devices rather than establishing novel acceptance criteria or performing a comparative effectiveness study in the typical sense of AI/human reader studies.
Here's an analysis of the provided information against your requested categories, acknowledging that some categories may not be directly applicable to this type of device and submission:
1. A table of acceptance criteria and the reported device performance
The document presents performance characteristics for the new devices and compares them to their predicate devices, implying these are the "acceptance criteria" for demonstrating substantial equivalence. Exact numerical acceptance criteria (e.g., "CV must be 95% confidence | 5.0 ng/mL of nordiazepam at > 95% confidence |
| Accuracy: Positive Samples (INTEGRA vs GC/MS) | 50 (INTEGRA +)/50 (GC/MS +) ; 0 (INTEGRA -)/0 (GC/MS -) from table | 50 (INTEGRA +)/50 (GC/MS +) ; 0 (INTEGRA -)/0 (GC/MS -) from table; i.e., 100% agreement when positive |
2. Sample sizes used for the test set and the data provenance
- ACPP Accuracy Test Set:
- Total Acid Phosphatase: n = 260 samples.
- Prostatic Acid Phosphatase: n = 264 samples.
- Data Provenance: Not explicitly stated (e.g., country of origin). The studies appear to be retrospective comparisons to predicate devices' performance claims.
- BNZGL Accuracy Test Set:
- Positive Samples: n = 50 samples for INTEGRABNZGL vs GC/MS comparison.
- The table indicates accuracy for positive samples where both INTEGRA with and without β-glucuronidase, and GC/MS all show 50 positive samples and 0 negative samples. This implies 50 positive samples were tested, and perhaps an additional number of negative samples that are not detailed in this specific comparison row, but rather in "overall agreement" data that is not fully presented.
- Data Provenance: Not explicitly stated (e.g., country of origin). Appears to be retrospective.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
This information is not provided in the document. For in vitro diagnostic devices, ground truth for quantitative measurements (like acid phosphatase levels) typically comes from reference methods or established laboratory procedures, not from human experts adjudicating images or other subjective interpretations. For benzodiazepine detection, GC/MS is treated as the reference ground truth, which is an analytical method.
4. Adjudication method (e.g., 2+1, 3+1, none) for the test set
Not Applicable for these types of in vitro diagnostic tests, which rely on quantitative measurements compared to reference methods (e.g., GC/MS) or predicate biochemical assays. There is no human adjudication process described.
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 document describes in vitro diagnostic assays, not AI-powered medical image analysis tools or other devices that involve human readers/interpreters. Therefore, no MRMC study or AI assistance effect size is discussed.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
This refers to the performance of the device without human intervention. In the context of these IVD devices, the stated performance characteristics (e.g., precision, accuracy, sensitivity, assay range) are the standalone performance of the reagent system on the COBAS INTEGRA Analyzer. There is no human-in-the-loop component for the analytical part of these tests.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
- COBAS INTEGRA Acid/Prostatic Phosphatase:
- Ground truth for accuracy was established by comparison to the predicate device, "Roche Reagent for Acid Phosphatase" (K831834). This is a method comparison study where the predicate acts as the reference or "ground truth."
- COBAS INTEGRA Benzodiazepines with β-glucuronidase:
- Ground truth for accuracy was established by comparison to Gas Chromatography/Mass Spectrometry (GC/MS) (as indicated in the "Accuracy Positive Samples" table for the predicate device, which is then compared for the new device). GC/MS is a widely accepted confirmatory method for drug detection and serves as the gold standard ground truth in this context.
8. The sample size for the training set
The document does not explicitly identify or specify a "training set" in the context of machine learning. For these diagnostic assays, the development and optimization of the reagent formulation and instrument parameters are analogous to "training" in a broader sense, but there's no data given for this phase. The presented performance data are from validation/verification studies, which would be considered test sets.
9. How the ground truth for the training set was established
As there is no explicitly defined "training set" in the machine learning sense, this question is not applicable. The ground truth for the performance evaluations (test sets) is described in point 7.
Ask a specific question about this device
(107 days)
The cassette COBAS INTEGRA HDL - Cholesterol Direct (HDL-D) contains an in vitro diagnostic reagent system intended for use on COBAS INTEGRA for the quantitative determination of HDL - cholesterol direct concentration in serum and plasma.
The COBAS INTEGRA HDL-D reagent cassette is intended for use with the COBAS INTEGRA Analyzer. The COBAS INTEGRA Analyzer and COBAS INTEGRA Reagent cassettes together provide an integrated system for in vitro diagnostic testing. The COBAS INTEGRA Analyzer utilizes three measuring principles, i.e., absorbance, fluorescence polarization and ion-selective electrodes. The analyzer has a throughput of up to 600 tests per hour with STAT samples prioritized and tested immediately. Random sample access, robotics and a user interface optimize time management and streamline workflow. The COBAS INTEGRA can store up to 68 COBAS INTEGRA Reagent Cassettes on board, 24 hours a day at 2-8°C. The COBAS INTEGRA Reagent Cassettes are compact and preparation-free with the added convenience of long term on-board stability. Barcode readers are used to identify newly loaded reagent cassettes, samples for patient identification, and rack inserts and to read calibration and control data from the cassette label. COBAS INTEGRA tests include chemistry, drugs of abuse, immunology, ion selective electrodes, therapeutic drug monitoring, and hematology reagents.
The COBAS INTEGRA HDL-D reagent cassette is an in vitro diagnostic reagent system intended for use on COBAS INTEGRA for the quantitative determination of High Density Lipoprotein cholesterol direct (HDL-D) in serum and plasma.
The principle of the COBAS INTEGRA HDL-D reagent cassette is based on the absorption of synthetic polyanions to the surface of lipoproteins. LDL, VLDL, and chylomicrons are transformed into a detergent-resistant form. whereas HDL is not. Combined action of polyanions and detergent solubilizes cholesterol from HDL, but not from LDL, VLDL, and chylomicrons. Solubilized cholesterol is oxidized by the sequential enzymatic action of cholesterol esterase and cholesterol oxidase. The hydrogen peroxide produced in this reaction is reacted with chromogens to form a colored dye. The increase in absorbance at 552 nm is directly proportional to the HDL cholesterol concentration of the sample.
Here's an analysis of the acceptance criteria and study as presented in the provided text for the Roche COBAS INTEGRA® HDL-D Reagent Cassette:
1. Table of Acceptance Criteria and Reported Device Performance
The document compares the new device (COBAS INTEGRA HDL-D reagent cassette) to a predicate device (COBAS INTEGRA HDL reagent system) rather than explicitly stating acceptance criteria a priori. However, the performance characteristics of the new device serve as its reported performance, and by demonstrating substantial equivalence to the predicate, it implicitly meets the established performance expectations for such devices.
| Performance Characteristic | Acceptance Criteria (Predicate Device Performance) | Reported Device Performance (COBAS INTEGRA HDL-D) |
|---|---|---|
| Assay Range | 0 - 5.0 mmol/L (0 - 193 mg/dL) | 0.01 - 4.0 mmol/L (0.386 - 155 mg/dL) |
| Precision (Level 1) | ||
| Mean (mmol/L) | 0.82 mmol/L (31.7 mg/dL) | 0.60 mmol/L (23.2 mg/dL) |
| % CV (within run) | 1.2% | 1.4% |
| % CV (total) | 2.7% | 2.2% |
| Precision (Level 2) | ||
| Mean (mmol/L) | 1.42 mmol/L (54.9 mg/dL) | 1.41 mmol/L (54.6 mg/dL) |
| % CV (within run) | 0.85% | 1.1% |
| % CV (total) | 5.5% | 2.3% |
| Sensitivity | 6.4 X 10⁻² ΔA per mmol/L (1.7 X 10⁻³ ΔA per mg/dL) | 8.2 X 10⁻² ΔA per mmol/L (2.1 X 10⁻³ ΔA per mg/dL) |
| Accuracy (Sample Size n) | 232 | 258 |
| Corr. Coefficient | 0.998 | 0.901 |
| Linear Regression | 0.99x - 0.05 mmol/L (0.99x - 1.93 mg/dL) | 0.77x + 0.33 mmol/L (0.77x + 12.8 mg/dL) |
| On-board Stability | 8 weeks | 12 weeks |
Note: The document presents the performance of both the "new device" (COBAS INTEGRA HDL-D reagent cassette) and the "predicate device" (COBAS INTEGRA HDL reagent system) side-by-side. The acceptance criteria for the new device are implied to be achieving comparable performance to the predicate, demonstrating substantial equivalence. The table above uses the predicate's performance as the benchmark for "acceptance criteria."
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Accuracy: 258 clinical samples were used for the accuracy study of the COBAS INTEGRA HDL-D reagent cassette.
- Data Provenance: The document does not explicitly state the country of origin or whether the data was retrospective or prospective. Given the nature of medical device submissions, it would typically be prospective clinical data collected under controlled conditions.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications
The document describes an assay for quantitative determination of HDL-Cholesterol Direct. The "ground truth" in this context refers to the true concentration of HDL cholesterol in the samples.
- Experts: Not applicable in the traditional sense for establishing ground truth for a quantitative assay. The ground truth for such assays is established through a reference method or a predicate device that has already been validated and accepted. The predicate device (COBAS INTEGRA HDL reagent system) served as the comparator for accuracy.
- Qualifications of Experts: Not relevant here as ground truth is determined by chemical analysis or comparative measurement, not expert opinion.
4. Adjudication Method for the Test Set
Not applicable. Adjudication is typically used when human interpretation or subjective assessment is part of the ground truth determination (e.g., in imaging studies). For a quantitative chemical assay, the comparison is directly between the result of the new device and the result of the reference/predicate method.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not conducted. MRMC studies are typically used for imaging or diagnostic tools where human readers interpret results, and the study aims to assess the impact of AI assistance on human performance. This submission is for a quantitative in vitro diagnostic reagent, which does not involve human interpretation in the same manner.
6. Standalone (Algorithm Only) Performance Study
Yes, a standalone performance study was conducted. The accuracy, precision, and sensitivity results presented in Table 2 for the COBAS INTEGRA HDL-D reagent cassette represent its standalone performance without human-in-the-loop assistance. The device directly measures HDL-D concentration.
7. Type of Ground Truth Used
The ground truth used for performance evaluation, particularly for accuracy, was based on comparison to a legally marketed predicate device (COBAS INTEGRA HDL reagent system). This predicate device itself quantifies HDL cholesterol using an enzymatic, colorimetric method. In essence, the new device's measurements were compared against the established measurements from the predicate device to show agreement.
8. Sample Size for the Training Set
The document does not provide information regarding a separate training set. For in vitro diagnostic reagents like this, the development process typically involves internal method development and optimization, but the specific term "training set" (common in AI/ML contexts) is not used, nor is a sample size provided for such a set. The presented data appears to be for verification and validation of the final product.
9. How the Ground Truth for the Training Set Was Established
As no training set information is provided, there is no information on how ground truth for a training set (if one existed during development) was established. For chemical assays, the development process would involve optimizing reagents and reaction conditions against known standards or reference methods to ensure accurate measurement across a range of concentrations.
Ask a specific question about this device
Page 1 of 1