Search Results

ImmunoCAP Rapid Reader, part of ImmunoCAP Rapid System, is an instrument including software to be used for reading and scoring ImmunoCAP Rapid assay results. It is intended for in vitro diagnostic use and is to be used in clinical laboratories, licensed under CLIA to perform non-waived assays.

The ImmunoCAP® Rapid Reader is part of ImmunoCAP® Rapid System, which is a combination of lateral flow immunoassay reagents and instrument/software for semi-quantitative determination of IgE antibodies in human capillary whole blood, heparinized venous whole blood or heparinized plasma. The Rapid Reader is a stand alone instrument to be used with ImmunoCAP Rapid Assay Device. It consists of the Rapid Reader instrument and associated software. The user interface consists of a LCD display with a touch screen and a slot for insertion of the ImmunoCAP Rapid Assay Device. The Reader includes functions for photometric reading and scoring of the test results. The Assay Device is illuminated with white color spectrum LED's. A sensor records an image of the Assay Device including Test and Control Windows, and reads the color saturation (Color Units).

Here's a breakdown of the acceptance criteria and study detailed in the provided 510(k) decision summary:

Acceptance Criteria and Device Performance

The core acceptance criterion for the ImmunoCAP Rapid Reader is that its performance in reading IgE antibody levels is in agreement with the predicate device (ImmunoCAP Rapid Reader, K081830). This agreement is quantitatively measured by Overall Agreement Percentage, Negative Percent Agreement (NPA), and Positive Percent Agreement (PPA) between the new and predicate device readings.

Table of Acceptance Criteria and Reported Device Performance

Explanation of Performance Metrics:

• Overall Agreement %: The percentage of cases where the new device and the predicate device yield the same semi-quantitative class score for a given allergen.
• Negative Percent Agreement (NPA) %: The percentage of cases where both devices correctly identify the absence of IgE antibodies (i.e., Class 0 or below a certain threshold, corresponding to a "negative" result in a semi-quantitative context).
• Positive Percent Agreement (PPA) %: The percentage of cases where both devices correctly identify the presence of IgE antibodies (i.e., any positive class like Class 1, 2, or 3).
• Precision (Total CV%): This measures the reproducibility and repeatability of measurements. A lower CV% indicates higher precision.
• Limit of Blank (LoB): The highest measurement result that is likely to be observed for a blank sample (one containing no analyte).
• Limit of Detection (LoD): The lowest quantity of analyte that can be distinguished from the absence of that analyte with a stated confidence level.

Study Details

The primary study conducted was a Method Comparison study to demonstrate agreement with the predicate device.

2. Sample Sizes Used for the Test Set and Data Provenance:

• Whole Blood Comparison Study:
  • Sample Size: 134 donors (resulting in 1340 measurements across 10 allergens).
  • Data Provenance: Heparinized capillary whole blood collected from 134 donors. The country of origin is not explicitly stated, but "in-house plasma samples" in the next section suggests it was likely internal to the manufacturer's region or facilities. The study design (collecting new samples and testing on both devices) indicates this was a prospective comparison for the purpose of this submission, although the samples themselves were subject to immediate testing.
• Plasma Comparison Study:
  • Sample Size: 239 sensitized donors (resulting in 2390 measurements across 10 allergens).
  • Data Provenance: "In-house plasma samples from 239 sensitized donors." The retrospective or prospective nature of these samples is not explicitly stated beyond being "in-house," but the testing on both new and predicate devices for comparison marks this as a prospective comparison study for the device.

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

• Not applicable. The "ground truth" for the method comparison study was the predicate device's measurement. There were no human experts establishing a separate ground truth for the semi-quantitative classification of IgE levels; the comparison was directly between the new and predicate instrument readings.

4. Adjudication Method for the Test Set:

• Not applicable. Since the ground truth was the predicate device's reading, no human adjudication was performed or described. The comparison was a direct measurement comparison between two instruments.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done:

• No. This was a method comparison study between two instruments (new device vs. predicate device), not a study comparing human reader performance with and without AI assistance. The device itself is an automated reader, eliminating a human-in-the-loop component for the reading process.
• A precision study was performed involving 3 different Readers reading the same pre-prepared Assay Device 30 times each. This assessed inter-reader variability but wasn't an MRMC study in the context of comparative effectiveness with human interpretation.

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

• Yes. The primary performance data (method comparison, precision, LoB/LoD) are all standalone performance of the ImmunoCAP Rapid Reader. The device is designed to automatically read and score the assay results without human intervention in the interpretation process.

7. The Type of Ground Truth Used:

• The "ground truth" for the method comparison study was the semi-quantitative class scores generated by the predicate device (ImmunoCAP Rapid Reader, K081830).

8. The Sample Size for the Training Set:

• The document does not explicitly describe a separate "training set" in the context of a machine learning algorithm. The ImmunoCAP Rapid Reader's software performs calculations to convert Color Units to Class scores. This process likely involves pre-defined thresholds and algorithms rather than a dynamically trained machine learning model in the contemporary sense. Therefore, information on a training set size for an AI/ML model is not provided and likely not applicable in the way it would be for modern AI-driven image analysis. The reader is calibrated at manufacturing.

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

• As a training set for a machine learning algorithm is not explicitly mentioned, the method for establishing its "ground truth" is not applicable/provided. The operational principle relies on photometric reading and conversion of Color Units to Class scores using built-in software, likely based on established reference values or a lookup table, rather than a learned truth from a labeled training dataset.

Ask a specific question about this device