LogoFDA 510(k) Search
Search Filters

Search Results

Found 1 results

510(k) Data Aggregation

    K Number
    K100101
    Device Name
    GSP NEONATALGALT KIT, MODEL 3303-001U
    Manufacturer
    PERKINELMER, INC.
    Date Cleared
    2010-06-11

    (149 days)

    Product Code
    KQP
    Regulation Number
    862.1315
    Type
    Traditional
    Panel
    Clinical Chemistry
    Reference & Predicate Devices
    K950803
    Predicate For
    N/A
    AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
    Intended Use

    The GSP Neonatal GALT kit is intended for the quantitative determination of r ne Sose - 1-phosphate uridyl transferase (GALT) activity in blood specimens dried on filter paper as an aid in screening newborns for classical galactosemia caused by GALT deficiency using the GSP™ instrument.

    Device Description

    The GSPTM Neonatal GALT assay is an adaptation of the quantitative enzymatic assay of Beutler and Baluda. The fluorescence is measured with the GSP Instrument using an excitation wavelength of 355 nm and an emission wavelength of 460 nm. The GSP Neonatal GALT assay uses prompt fluorescence technology.

    AI/ML Overview

    Here's a breakdown of the acceptance criteria and study information for the GSP Neonatal GALT kit, based on the provided 510(k) summary:

    Acceptance Criteria and Device Performance

    The provided document does not explicitly state pre-defined acceptance criteria in terms of specific thresholds for precision, linearity, detection limits, or agreement with the predicate device. Instead, it presents the results of these performance characteristics studies and a comparison with the predicate device.

    However, based on the intent of the comparison study with the predicate device, we can infer some implicit acceptance goals relating to agreement with the established method. The screening performance tables directly compare the classification of samples between the new device and the predicate.

    Inferred Acceptance Criteria & Reported Device Performance:

    Performance MetricInferred Acceptance Criteria (Implicit)Reported Device Performance
    Precision (CV%)Values within acceptable clinical laboratory ranges for quantitative assays, typically demonstrating good reproducibility. (Specific quantitative targets for CV% are not stated, but results are presented to demonstrate low variability across multiple conditions.)Within-Run Variation: CV% range from 3.0% to 12.8%Within-Lot Variation: CV% range from 4.9% to 14.3%Total Variation: CV% range from 5.2% to 15.9%
    LinearityDemonstrated linearity across the expected measuring range for clinical utility.Linear throughout the measuring range of 2.5 U/dL to 25 U/dL. Maximum observed difference between linear and 3rd order regression models is -2.6% for GALT activities >4 U/dL, and max absolute difference of 0.07 U/dL for GALT activities ≤ 4 U/dL.
    Detection Limit (LoQ)Ability to accurately quantify GALT activity at low concentrations relevant for screening GALT deficiency.LoB: 1.6 U/dLLoD: 2.5 U/dLLoQ: 2.5 U/dL (with total CV ≤ 20%)
    Analytical SpecificityNo significant interference from common endogenous substances or therapeutic compounds (e.g., icteric, lipemic samples, ascorbic acid, galactose). Minimal or explainable interference from other relevant substances (e.g., glutathione, GAL-1-P) at clinically relevant concentrations.No interference from icteric (bilirubin ≤ 40 mg/dL), lipemic (Intralipid ≤ 1000 mg/dL), ascorbic acid (≤ 3 mg/dL), or galactose (≤ 50 mg/dL). Glutathione: Interference (decrease up to 63%) above 18.8, 37.5, 56.3 mg/dL at 3, 6, 12 U/dL GALT activity respectively.GAL-1-P: No effect on 3 U/dL samples; interference (decrease up to 37%) on 6 and 12 U/dL samples at 12.5 mg/dL.Total protein (HSA): No effect on 12 U/dL samples; interference (increase up to 30%) on 3 and 6 U/dL samples above 3000 mg/dL.
    Comparison with Predicate - Screening Performance (using 0.5th percentile cut-off)High overall agreement, positive percent agreement, and negative percent agreement with the predicate device (NG-1100/4100), demonstrating comparable screening classification.Overall percent agreement: 99.6% (CI 99.3%-99.9%)Positive percent agreement: 92.9% (CI 83.9%-100%)Negative percent agreement: 99.8% (CI 99.5%-100%)
    Comparison with Predicate - Screening Performance (using 1.0st percentile cut-off)Similarly high agreement metrics at a different cut-off.Overall percent agreement: 99.3% (CI 98.9%-99.7%)Positive percent agreement: 84.9% (CI 74.3%-95.5%)Negative percent agreement: 99.6% (CI 99.3%-99.9%)
    Comparison with Predicate - Screening Performance (using 1.5th percentile cut-off)Similarly high agreement metrics at a different cut-off.Overall percent agreement: 98.9% (CI 98.4%-99.4%)Positive percent agreement: 83.6% (CI 73.5%-93.7%)Negative percent agreement: 99.3% (CI 99.0%-99.7%)

    Study Information

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

      • Precision Study: 8 samples (S1-S8), each tested with 206-216 replicates (n values).
      • Detection Limit Study:
        • LoB: 83 GALT deficient samples
        • LoD: 351 determinations of five low-level samples
        • LoQ: 209 replicates
      • Analytical Specificity Study: Whole blood samples with three different GALT activities (approx. 3, 6, and 12 U/dL) were tested. Hematocrit effect was tested on three whole blood samples (approx. <1, 6, and 15 U/dL). Number of replicates (n) for hematocrit testing ranged from 10 to 12 for each hematocrit level.
      • Comparison Studies (Screening Performance): A total of 2205 infants.
        • Routine screening specimens: 2146 samples.
        • Retrospective low GALT activity specimens: 33 samples.
      • Data Provenance: Not explicitly stated, but the comparison study was performed "in a routine screening laboratory," implying real-world clinical samples. The study does not specify the country of origin, but the submission is from Finland with a US contact, suggesting potential international collaboration or a study done in a US lab, though this is not definitively stated. The use of "routine" and "retrospective" specimens indicates it's a mix of prospective (routine) and retrospective (low GALT activity) data.

    2. 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):

      • The ground truth for the comparison study was established by the predicate device (NG-1100/4100 Neonatal GALT kit). Therefore, no human experts were involved in establishing the ground truth for the comparison study. For the underlying clinical truth, the predicate device's performance would have been validated against clinical outcomes or confirmed diagnoses (e.g., pathology, genetic testing), but this information is not provided for the predicate device within this summary.

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

      • Not applicable. The ground truth was based on the measurement from the predicate device, not on expert adjudication.

    4. 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 a medical device for quantitative determination of a biomarker, not an AI-assisted diagnostic imaging or interpretation system requiring human readers.

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

      • Yes, the performance characteristics (precision, linearity, detection limit, analytical specificity) and the comparison study represent the standalone performance of the GSP Neonatal GALT kit (algorithm only). The device's intended use is for quantitative determination using the GSP instrument, which is an automated process without direct human-in-the-loop interpretation that would alter the quantitative result provided by the device. The "screening performance" discussed refers to the device's ability to classify samples based on its quantitative output and a defined cut-off, not a human-AI interaction.

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

      • For the comparison study, the "ground truth" (or reference standard) was the predicate device (NG-1100/4100 Neonatal GALT kit). The predicate device's classifications were used to assess the agreement of the new device. For the analytical studies, the ground truth was based on known concentrations or properties of the samples used (e.g., GALT deficient samples for LoB, known GALT activities for linearity and specificity).

    7. The sample size for the training set:

      • Not applicable. This device is a diagnostic assay kit, not an AI/machine learning model that typically requires a distinct "training set." The performance characteristics and comparison studies are conducted on test sets to validate the device's analytical performance.

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

      • Not applicable, as there is no training set in the context of an AI/machine learning model. For the assay, the "ground truth" in development would refer to the standards and controls used to establish the assay's biochemical principles and calibrate its measurements.
    Ask a Question

    Ask a specific question about this device

    Page 1 of 1