Search Filters

Search Results

Found 1 results

510(k) Data Aggregation

    K Number
    K994221

    Validate with FDA (Live)

    Device Name
    BAYER PSA ASSAY
    Manufacturer
    Date Cleared
    2000-04-13

    (120 days)

    Product Code
    Regulation Number
    866.6010
    Age Range
    All
    Reference & Predicate Devices
    N/A
    Predicate For
    N/A
    AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticPediatricDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
    Intended Use

    The Bayer Diagnostics PSA Immunoassay is for the quantitative serial determination of prostate specific antigen in human serum and to aid in the managenent (monitoring) of patients with prostate cancer

    Device Description

    The ACS:180 and ADVIA Centaur PSA assays are a two-site sandwich immunoassay using direct chemiluminometric technology, which uses constant amounts of two antibodies. The direct offermilammomotio to a polyclonal anti-goat antibody labeled with acridinium not ankibody, in the Solid Phase, is a monoclonal anti-mouse antibody, which is covalently coupled to paramagnetic particles.
    A direct relationship exists between the amount of PSA present in the patient sample and the amount of relative light units (RLUs) detected by the system

    AI/ML Overview

    Here's an analysis of the provided text, focusing on acceptance criteria and the study data, organized by your requested points:

    Bayer Diagnostics ACS: 180/ADVIA Centaur PSA Assay Performance Data

    This document describes the performance of the Bayer Diagnostics ACS: 180 and ADVIA Centaur PSA assays, which are immunoassays for the quantitative determination of prostate-specific antigen (PSA) in human serum.

    1. Table of Acceptance Criteria and Reported Device Performance

    The document does not explicitly state "acceptance criteria" as pass/fail thresholds for specific metrics. However, it presents various performance characteristics intended to demonstrate the device's analytical capabilities and equivalence to a predicate device. For this table, I will present the reported performance data for key metrics.

    Performance MetricAcceptance Criteria (Implied/Reference)Reported Device Performance (ACS: 180)Reported Device Performance (ADVIA Centaur)
    Analytical SensitivityNot explicitly stated but expected to be precise at low concentrations0.06 ng/mL (minimum detectable concentration) at 2 SD above mean zero standard0.06 ng/mL (minimum detectable concentration) at 2 SD above mean zero standard
    Assay RangeNot explicitly stated (standard for PSA assays)Up to 100 ng/mLUp to 100 ng/mL
    Accuracy (Method Comparison with Alternate Method)Strong correlation (e.g., r > 0.95, slope ~1, intercept ~0)ACS: 180 PSA = 0.98 (alternate method) + 0.0118 ng/mL; r = 0.986 (for 629 samples, 0.06-100 ng/mL)Not directly compared to "alternate method" in this section
    Accuracy (Method Comparison with ACS: 180)Strong correlation (e.g., r > 0.95, slope ~1, intercept ~0)N/AADVIA Centaur PSA = 0.99 (ACS: 180 PSA) - 0.09 ng/mL; r = 0.990 (for 661 samples, 0.06-100 ng/mL)
    Precision (Total % CV) at specific PSA levelsNot explicitly stated (typically <10% for clinical assays, often lower for R&D/manufacturing targets)See table belowSee table below
    Examples (ACS: 180)0.70 ng/mL: 5.9% CV0.44 ng/mL: 5.97% CV
    1.83 ng/mL: 5.0% CV1.831 ng/mL: 5.12% CV
    76.25 ng/mL: 6.3% CV17.706 ng/mL: 3.31% CV

    ACS: 180 Precision Summary Table (Total % CV)

    Mean (ng/mL)Total % CV
    0.705.9
    0.915.3
    1.835.0
    17.554.2
    18.234.6
    29.735.1
    54.345.3
    76.256.3

    ADVIA Centaur Precision Summary Table (Total % CV)

    Mean (ng/mL)Total % CV
    0.445.97
    0.7083.71
    1.8315.12
    1.9342.60
    11.3084.68
    17.7063.31

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

    • Accuracy and Method Comparison:

      • ACS: 180 vs. Alternate Method: 629 samples
      • ADVIA Centaur vs. ACS: 180: 661 samples
      • Data Provenance: Not specified in the provided text. It does not mention country of origin or if the data was retrospective or prospective.
    • Expected Results (Distribution of PSA by Diagnostic Category): This can be considered a test set for assessing the clinical utility of the assay against expected population distributions.

      • Total Patients: 100 (Healthy Female) + 283 (Healthy Male) + 191 (Prostate Cancer) + 152 (BPH) + 50 (GU) + 18 (Prostatitis) + 5 (Rheumatoid Factor) + 10 (Breast Cancer) + 6 (Renal Cancer) + 10 (Pulmonary Cancer) + 39 (Misc. GU) + 12 (Gastrointestinal) + 18 (Other) = 814 samples
      • Data Provenance: Not specified. It refers to "serum samples from healthy subjects and patients with various malignant diseases." No mention of country or retrospective/prospective nature.
    • Precision:

      • ACS: 180: 8 samples, each assayed 3 times in 6 assays, on each of 4 systems (n = 72 for each sample). This implies 8 x 72 = 576 data points for calculation, but the samples are typically controls or pooled patient samples run multiple times, not unique patient samples.
      • ADVIA Centaur: 6 samples, each assayed 3 times in 8 runs, on each of 4 systems, over 3 days (n = 24 for each sample). This implies 6 x 24 = 144 data points for calculation.
      • Data Provenance: Not specified.

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

    The provided text describes analytical performance studies and clinical distribution studies for a PSA immunoassay. For immunoassay performance studies (like accuracy and precision), ground truth is typically established by the reference measurement procedure or the assigned value of calibrators/control materials, not by experts adjudicating images or clinical cases.

    For the "Expected Results" section, which stratifies PSA levels by patient diagnosis, the diagnoses (Prostate Cancer, BPH, etc.) would represent the "ground truth." The document does not specify the number or qualifications of experts used to establish these diagnoses for the patients whose samples were included in the study. These diagnoses would have been established by standard clinical practices (e.g., biopsy results, imaging, clinical examination, and physician judgment) at the time of sample collection.

    4. Adjudication method for the test set

    Not applicable. As noted above, this study evaluates an immunoassay's analytical performance and the distribution of a biomarker in various clinical populations. There is no multi-reader or image interpretation component requiring an adjudication method like 2+1 or 3+1. The "ground truth" for patient diagnoses (in the "Expected Results" section) would have been based on established clinical diagnostic criteria, not expert adjudication in the context of this document.

    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 an in vitro diagnostic (IVD) immunoassay for measuring PSA levels in blood. It is not an AI-assisted diagnostic device, nor does it involve human readers interpreting cases (like imaging diagnostics). Therefore, an MRMC comparative effectiveness study involving AI assistance is not relevant to this submission.

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

    Yes, the studies presented are standalone performance studies of the immunoassay devices themselves. The devices (ACS: 180 and ADVIA Centaur) are automated systems that measure PSA levels. Their performance is described analytically (sensitivity, range, accuracy, precision) as the algorithm/device only, without human interpretation of the assay results as a primary variable in these specific performance metrics. The end-user (clinician) then interprets the quantitative PSA result in the context of a patient's clinical picture.

    7. The type of ground truth used

    • Analytical Sensitivity, Assay Range, Precision: The ground truth for these metrics usually comes from reference materials, certified calibrators, and control samples with known, assigned values. Analytical sensitivity is based on statistical deviation from a zero standard.
    • Accuracy (Method Comparison): The "ground truth" is typically the result from a methodologically equivalent or predicate device (e.g., "alternate method" or the ACS: 180 assay when evaluating the ADVIA Centaur). This establishes agreement between the new device and an already accepted method.
    • Expected Results (Distribution of PSA by Diagnostic Category): The ground truth for these categories (e.g., "Prostate Cancer," "Benign Prostatic Hypertrophy") is the clinical diagnosis of the patients. This diagnosis would be based on a combination of clinical findings, imaging, and often, histopathology (pathology results) from biopsies for cancer diagnoses. "Outcomes data" could also contribute to confirming disease status over time.

    8. The sample size for the training set

    The document does not explicitly describe a "training set" in the context of machine learning or AI. This is a submission for an immunoassay, which is a chemical and biological measurement system, not a software algorithm that is "trained" in the conventional AI sense.

    The development of such an immunoassay involves extensive research and development, including:

    • Selection and optimization of antibodies and reagents.
    • Optimization of assay protocols (incubation times, temperatures).
    • Calibration curve development.
    • Internal validation studies to refine the assay before formal performance testing.

    The samples used during these developmental and optimization phases would informally serve a "training" function for the assay design, but they are not referred to as a "training set" in this type of regulatory submission. The performance data presented are from validation studies intended to demonstrate the final product's performance.

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

    As there is no "training set" in the AI sense for this immunoassay, this question is not directly applicable. For the developmental and optimization phases, the "ground truth" would be established through a combination of:

    • Known concentrations of PSA in spiked samples.
    • Characterized reference materials and patient samples whose PSA levels have been reliably measured by established reference methods or predicate devices.
    • Clinical diagnoses (including pathology) for patient samples used to understand the relationship between PSA levels and disease states during method development.
    Ask a Question

    Ask a specific question about this device

    Page 1 of 1