Search Results

The Access CEA assay is a paramagnetic particle, chemiluminescent immunoassay for the quantitative determination of Carcinoembryonic Antigen (CEA) levels in human serum, using the Access Immunoassay Systems. CEA measured by the Access Immunoassay Systems is used as an aid in the management of cancer patients in whom changing CEA concentrations have been observed.

The Access CEA assay is a two-site immunoenzymatic "sandwich" assay using two mouse monoclonal anti-CEA antibodies (MAb) which react with different epitopes of CEA. The Access CEA reagent kit is in a liguid ready-to-use format designed for optimal performance on Beckman Coulter's immunoassay analyzers. Each reagent kit contains two reagent packs. Other items needed to run the assay include substrate, calibrators, and wash buffer.

The provided text describes the 510(k) submission for the Beckman Coulter Access CEA assay on the Dxl 9000 Access Immunoassay Analyzer. This document focuses on demonstrating substantial equivalence to a predicate device (Access CEA on the Access Immunoassay Analyzer), primarily through analytical performance studies rather than clinical or AI-assisted diagnostic studies.

Therefore, many of the requested criteria related to AI performance, human reader studies, and expert ground truth are not applicable or cannot be extracted from this document. The information provided is characteristic of a submission for an in vitro diagnostic (IVD) device, which relies heavily on analytical performance characteristics.

Here's the breakdown based on the provided text:

1. Acceptance Criteria and Reported Device Performance

The document provides acceptance criteria and performance for various analytical studies.

2. Sample size used for the test set and the data provenance:

• Method Comparison: A total of 153 serum samples were evaluated.
• Imprecision: Multiple samples (7 distinct concentration levels) were tested with a minimum of three replicates in 2 runs per day for a minimum of 20 days. The "N" column in the table refers to the total number of measurements for each sample level (e.g., 126 for Sample 1, 120 for Sample 2, etc.).
• Data Provenance: The document does not specify the country of origin of the data or whether the samples were retrospective or prospective. This is typical for IVD analytical performance studies.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• This information is not applicable to this type of device and study. The ground truth for analytical performance studies like Method Comparison, Imprecision, Linearity, LoB, LoD, and LoQ is established by the reference measurement method (in this case, the predicate device for method comparison, or precise measurements by the instrument itself for other analytical characteristics) and laboratory standards, not by expert human interpretation like a radiologist reading an image.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set:

• This is not applicable. Adjudication methods are used in studies involving human interpretation or challenging diagnoses, not for analytical performance of an IVD assay measuring a biomarker concentration.

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:

• This is not applicable. This is an in vitro diagnostic device (immunoassay for CEA), not an AI-powered image analysis or diagnostic assist device for human readers. No human interpretation or AI assistance study was performed or required for this type of submission.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• This is not applicable in the context of AI algorithms. The device itself (Access CEA assay on the Dxl 9000 Access Immunoassay Analyzer) is a standalone automated analytical instrument. Its performance is evaluated as an "algorithm only" in the sense that it performs the measurement independently, but it's a chemical and optical measurement process, not an AI algorithm.

7. The type of ground truth used:

• For the Method Comparison study, the "ground truth" or reference method was the predicate device (Access CEA on the Access Immunoassay Analyzer).
• For other analytical studies (Imprecision, Linearity, LoB, LoD, LoQ), the ground truth is established through controlled spiking, dilutions, and repeated measurements according to established analytical validation guidelines (e.g., CLSI EP-05-A3 for imprecision). These are intrinsic analytical properties of the assay and instrument combination, verified against laboratory standards.

8. The sample size for the training set:

• This is not applicable as this is not an AI/machine learning device that requires a "training set" in the conventional sense. The development of an immunoassay involves optimizing reagents, antibodies, and instrument parameters, which is a different process from training a machine learning model.

9. How the ground truth for the training set was established:

• This is not applicable for the reasons stated in point 8.

Ask a specific question about this device

The Access CEA assay is a paramagnetic particle, chemiluminescent immunoassay for the quantitative determination of Carcinoembryonic Antigen (CEA) levels in human serum, using the Access Immunoassay Systems. CEA measured by the Access Immunoassay Systems is used as an aid in the management of cancer patients in whom changing CEA concentrations have been observed.

The Access® CEA reagents consist of reagent packs, calibrators, bi-level controls, substrate and wash buffer.

The provided text describes a 510(k) premarket notification for a modification to the Access® CEA Reagents on the Access® Immunoassay Systems. The modification involves adding a new instrument platform, the Beckman Coulter UniCel™ Dxl 800 Access® Immunoassay System, to the existing family of Access Immunoassay Systems.

The study aimed to demonstrate that the Access CEA assay on the Dxl system is substantially equivalent to the Access CEA assay on the Access 2 system.

Here's a breakdown of the requested information based on the provided text:

1. A table of acceptance criteria and the reported device performance

The document states that the Access CEA assay met established acceptance criteria for "method comparison, precision and analytical sensitivity." However, the specific numerical acceptance criteria for these parameters (e.g., specific ranges for agreement, coefficients of variation, or limits of detection) are not detailed in the provided summary. Similarly, the reported performance values from the studies (e.g., actual method comparison results, precision data, or analytical sensitivity figures) are not provided. The text only offers a general statement of compliance.

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 does not specify the sample sizes used for the method comparison, precision, or analytical sensitivity studies. It also does not mention the data provenance (e.g., country of origin, retrospective or prospective nature of the 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 immunoassay for quantitative determination of Carcinoembryonic Antigen (CEA) levels, which relies on a chemical reaction to produce a numerical result. Therefore, there is no ground truth established by human experts in the same way it would be for an imaging device requiring expert interpretation. The "ground truth" for evaluating this device would be established by reference methods or established analytical standards. The document does not provide details on how this was established for the comparison.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

Since this is an immunoassay seeking substantial equivalence to a predicate device, the "adjudication method" in the context of human expert review is not applicable. The evaluation relies on quantitative analytical comparisons, not human interpretation consensus.

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 performed. This type of study is relevant for medical imaging or diagnostic interpretation tasks where human readers are involved. This device is an automated immunoassay system, and its evaluation focuses on analytical performance metrics rather than human reader improvement.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

Yes, the studies described (method comparison, precision, and analytical sensitivity) inherently represent standalone performance of the algorithm/device. The device itself performs the quantitative determination of CEA levels. Human involvement would be in operating the instrument and interpreting the numerical output, but the performance being evaluated is that of the automated system. The document states that the new system (Dxl) uses the same reagents and calibrators as the predicate (Access 2), implying that the algorithm/assay itself is unchanged, and the evaluation is on the new instrument platform's ability to produce equivalent results.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

The "ground truth" for these types of studies would typically be established by:

• Predicate device results: For method comparison, the results from the legally marketed Access® CEA Reagents on the Access® Immunoassay Analyzer (K981985, K991707) would serve as the comparator or "reference."
• Reference materials/standards: For precision and analytical sensitivity, the device would be tested against known concentrations of CEA or characterized control materials.

The document does not explicitly state the type of ground truth used beyond indicating that method comparison was against the predicate device.

8. The sample size for the training set

The document does not mention a training set. This is because the device is an immunoassay kit/system, not an artificial intelligence or machine learning algorithm that requires a distinct training phase. The "training" of such a system would involve its initial development and validation by the manufacturer, but not in the context of a dataset used to optimize an AI model.

9. How the ground truth for the training set was established

As there is no "training set" in the context of an AI/ML algorithm for this immunoassay device, this question is not applicable.

Ask a specific question about this device

The Access® CEA assay is a paramagnetic particle, chemiluminescent immunoassay for the quantitative determination of Carcinoembryonic antigen (CEA) in human serum using the ACCESS Immunoassay System. CEA measured by the Access CEA Immunoassay, is used as an aid in the management of cancer patients.

The ACCESS CEA Immunoassay Reagents and the ACCESS Immunoassay Analyzer comprise the ACCESS Immunoassay System for the quantitative determination of CEA in human serum.

The provided text describes a 510(k) submission for the ACCESS® CEA Assay, a device intended for the quantitative determination of carcinoembryonic antigen (CEA) in human serum to aid in the management of cancer patients. The submission focuses on demonstrating substantial equivalence to a predicate device, the Abbott IMx CEA Microparticle Enzyme Immunoassay.

Here's an analysis based on the provided document:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state "acceptance criteria" for the ACCESS CEA Immunoassay in numerical thresholds. Instead, it presents summaries of studies demonstrating its performance in comparison to the predicate device and in various analytical tests. The underlying acceptance logic for substantial equivalence would be that the performance of the new device is comparable to or better than the predicate.

2. Sample Size Used for the Test Set and Data Providence

• Correlation: 146 samples, ranging from 0 to 1000 ng/ml.
• Expected Range: 234 apparently healthy males and females.
• Monitoring Data: "Longitudinal serum samples from previously diagnosed cancer patients." The exact number of patients or samples is not specified.
• Recovery, Precision, Specificity, Hook Effect, Analytical Sensitivity: Specific sample sizes for these analytical studies are not detailed beyond "human serum samples" or "spiked into serum samples."
• Data Provenance: Not explicitly stated (e.g., country of origin). The studies appear to be quantitative laboratory analyses of human serum samples. The context of a US 510(k) submission generally implies that at least some data would be relevant to the US population, but this is not guaranteed for all sample types. The data is retrospective in the sense that the samples were analyzed after collection, but the studies themselves were conducted prospectively to evaluate the device.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts

• Ground Truth for Correlation, Recovery, Precision, Hook Effect, Analytical Sensitivity: The "ground truth" for these analytical studies is based on the quantitative measurements derived from the reference methods or the expected values of spiked/diluted samples. No human experts are explicitly mentioned for establishing ground truth in these specific analytical performance tests for the test set.
• Ground Truth for Monitoring Data: The "monitoring data" relies on "concurrent clinical assessments." This implies that clinicians or medical professionals were involved in evaluating the patient's condition over time. The number and qualifications of these experts are not explicitly stated, but they would typically include oncologists or other physicians managing cancer patients.
• Ground Truth for Expected Range: This involves classifying individuals as "apparently healthy." This classification is typically done by medical professionals, but details about the number and qualifications are not provided.

4. Adjudication Method for the Test Set

• The document does not describe a formal "adjudication method" in the context of multiple human readers or reviewers for the quantitative analytical studies (correlation, precision, etc.). These are objective quantitative measurements.
• For the "monitoring data" and "clinical assessments," while multiple clinicians might be involved in patient management, a specific adjudication process (like 2+1 or 3+1 consensus) for establishing a definitive ground truth across different clinicians is not described in the summary.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done

No, a multi-reader multi-case (MRMC) comparative effectiveness study was not explicitly described. This type of study typically involves human readers interpreting cases (e.g., medical images) with and without AI assistance to measure reader performance improvement. The ACCESS CEA Immunoassay is a diagnostic assay measuring a biomarker, not an imaging interpretation device. The "monitoring data" involved comparing the assay results with clinical assessments, which is a form of clinical utility assessment, but not an MRMC study.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

Yes, the studies described are primarily standalone performance assessments of the ACCESS CEA Immunoassay. The device itself is an automated immunoassay system (ACCESS Immunoassay Analyzer and ACCESS CEA Immunoassay Reagents). The results it produces (CEA values) are quantitative outputs. The comparisons to the predicate device and the analytical performance tests (precision, recovery, specificity, etc.) evaluate the device's performance inherently in a standalone manner, without human intervention in the generation of the CEA value. While clinicians use these results, the performance evaluation in this 510(k) focuses on the analytical accuracy and reliability of the device's output.

7. The Type of Ground Truth Used

• Correlation: The ground truth is the measurement obtained from the predicate device, the Abbott IMx CEA Immunoassay.
• Expected Range: The ground truth is established by clinical assessment identifying "apparently healthy" individuals, and the measured CEA values in that population.
• Monitoring Data: The ground truth is "concurrent clinical assessments" of cancer patients, implying a physician's overall evaluation of the patient's condition and disease progression.
• Recovery: The ground truth is the known concentration of CEA in diluted or spiked samples.
• Precision, Specificity, Hook Effect, Analytical Sensitivity: The ground truth largely relies on the inherent properties of the samples (e.g., known concentrations, presence/absence of interferents) and accepted laboratory standards for analytical performance.

8. The Sample Size for the Training Set

The document does not explicitly mention a "training set" in the context of developing a machine learning algorithm. This device is an immunoassay, which functions based on chemical and immunological reactions, not a machine learning model that requires a discrete training phase. Therefore, the concept of a training set as understood in AI/ML is not applicable here. The development and validation of the assay would have involved various internal optimization and verification studies, but these would not typically be called a "training set" in this context.

9. How the Ground Truth for the Training Set Was Established

As explained above, the concept of a "training set" for an immunoassay device is not applicable in the same way it is for an AI/ML device. The ground truth for developing and optimizing such an assay would involve using characterized samples with known CEA concentrations, spiked samples, and samples from well-defined patient populations, as well as adherence to established analytical chemistry and immunology principles.

Ask a specific question about this device