FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

The CAPILLARYS HEMOGLOBIN(E) kit is designed for the separation of the normal hemoglobins (A, A2 and F) in human blood samples, and for the detection of the major hemoglobin variants (S, C, E and D), by capillary electrophoresis in alkaline buffer (pH 9.4) with the SEBIA CAPILLARYS 2 FLEX-PIERCING instrument. The CAPILLARYS HEMOGLOBIN(E) kit is designed for laboratory use.

The CAPILLARYS 2 FLEX-PIERCING instrument is an automated analyzer which performs a complete hemoglobin profile for the quantitative analysis of the normal hemoglobin fractions A, A2 and F and for the detection of major hemoglobin variants S, C, E and D. The assay is performed on the hemolysate of whole blood samples collected in tubes containing K₂EDTA or K3EDTA as anticoagulant.

For In Vitro Diagnostic Use.

Device Description

AI/ML Overview

This is a 510(k) premarket notification for a medical device, which typically focuses on demonstrating substantial equivalence to a legally marketed predicate device rather than providing extensive details about acceptance criteria and comprehensive study reports as would be found in a Premarket Approval (PMA) application or a detailed clinical study publication.

Based on the provided document, here's an analysis of the information available regarding acceptance criteria and the study:

The document is an FDA 510(k) clearance letter for the Sebia CAPILLARYS HEMOGLOBIN(E) using the CAPILLARYS 2 FLEX-PIERCING instrument. This letter confirms that the FDA has reviewed the submission and determined that the device is "substantially equivalent" to legally marketed predicate devices. This means the device performs at least as well as existing devices on the market for its intended use.

Important Note: The provided document is the FDA's clearance letter and the "Indications for Use" page. It does not include the detailed study reports or the full 510(k) submission where the specific acceptance criteria and detailed study methodology would be found. Therefore, much of the requested information cannot be directly extracted from this document.

However, based on the nature of a 510(k) for an in vitro diagnostic (IVD) device like an abnormal hemoglobin assay, we can infer some general characteristics and make assumptions where specific details are not given.

1. Table of Acceptance Criteria and Reported Device Performance

This information is not provided in the given document. A 510(k) submission would typically include detailed performance data (e.g., accuracy, precision, analytical specificity, analytical sensitivity, linearity, limits of detection/quantification) compared against pre-defined acceptance criteria. The clearance letter only states that the device was found "substantially equivalent."

For an abnormal hemoglobin assay like this, typical acceptance criteria would involve:

Accuracy/Agreement: Correlation or concordance with a reference method (e.g., HPLC, CE, or a predicate device) for identifying and quantifying hemoglobin fractions (A, A2, F) and variants (S, C, E, D). This might be expressed as a percentage agreement, mean bias, or confidence intervals.
Precision: Reproducibility and repeatability of measurements (e.g., CV% for quantitative measurements) over various levels of hemoglobin.
Interference: Lack of significant interference from common substances in blood.
Analytical Sensitivity and Specificity: Ability to correctly detect and differentiate variants from normal and other variants.

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

This information is not explicitly stated in the provided document.

Sample Size: While not mentioned, IVD 510(k) studies typically involve a few hundred to a few thousand samples for method comparison, precision, and other analytical performance evaluations. These samples would ideally cover a range of hemoglobin types (normal, various variants) and concentrations.
Data Provenance: The document does not specify the country of origin. Given Sebia Inc. is in Georgia, USA, and the FDA is a US regulatory body, some data would likely be from the US, but it could also include international study sites. It's highly probable the study was retrospective for clinical samples used in method comparisons and prospective for collecting data during validation studies (e.g., precision, linearity).

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

This information is not provided in the given document.

For an IVD device like this, ground truth is typically established using:

An established reference method (e.g., High-Performance Liquid Chromatography (HPLC) or another capillary electrophoresis system that is considered a gold standard).
In some cases, samples might be characterized by multiple orthogonal methods to confirm the presence and type of hemoglobin variants.
Less commonly, individual "experts" are used to establish ground truth for quantitative assays unless the ground truth is a subjective clinical interpretation (which is not the case for hemoglobin fractions). If expert review was part of a discrepant analysis, it would typically involve clinical laboratory scientists or pathologists experienced in hemoglobinopathy diagnostics.

4. Adjudication Method for the Test Set

This information is not provided in the given document.

Adjudication methods (like 2+1 or 3+1) are typically used when the ground truth itself is subject to inter-rater variability, often in image-based diagnostic systems or clinical assessments where experts interpret data. For a quantitative assay like hemoglobin electrophoresis, the "adjudication" would more likely involve:

Statistical analysis of agreement between the device and the reference method.
Discrepant analysis: Cases where the device result differs from the reference method would be re-tested by both methods, potentially using an additional confirmatory method, to understand the source of the discrepancy. This is more of a technical investigation than an expert adjudication.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and the Effect Size

No, an MRMC comparative effectiveness study was almost certainly NOT done for this device.

Reasoning: MRMC studies are primarily relevant for diagnostic imaging devices or other systems where human readers interpret complex data, and the AI system is intended to assist or replace human interpretation.
This device, the Sebia CAPILLARYS HEMOGLOBIN(E) system, is an automated in vitro diagnostic (IVD) assay. It performs analytical separation and quantification of hemoglobin fractions. While a human might interpret the final report, the core function of the device is automated analysis, not "reading" in the sense of a radiologist interpreting an image.
Therefore, there's no "human readers improve with AI vs without AI assistance" scenario applicable to evaluate using an MRMC study for this type of device.

6. If a Standalone (Algorithm Only Without Human-in-the-Loop Performance) Was Done

Yes, the primary performance evaluation for this device would be a standalone (algorithm only) performance study.

Reasoning: As an automated IVD assay, the device's performance (accuracy, precision, etc.) is assessed based on its ability to correctly identify and quantify hemoglobin fractions and variants directly from the blood sample using its internal algorithms and methods.
The "performance" is the output of the instrument, not dependent on human-in-the-loop interpretation during the measurement process. A human reviews the results generated by the device, but the device's analytical performance is tested independently of that human review. This is the standard for IVD assays.

7. The Type of Ground Truth Used

Based on the device type (automated hemoglobin assay), the ground truth would most likely be established by:

Reference Methods: Such as a "gold standard" High-Performance Liquid Chromatography (HPLC) system or another well-established and validated capillary electrophoresis (CE) method from a reputable manufacturer, or a combination of methods.
Potentially, molecular diagnostics (DNA sequencing) for definitive identification of specific hemoglobin variants where biochemical methods might be ambiguous.
Expert Consensus (less likely as primary ground truth): While a lab director or clinical pathologist would oversee the process, the "ground truth" for the quantitative values and variant identification is typically derived from objective analytical methods rather than expert visual interpretation for this specific type of assay.

8. The Sample Size for the Training Set

This information is not provided in the given document.

Reasoning: For an automated IVD device like this, which likely uses established biochemical principles and algorithms for peak detection and quantification, a "training set" in the machine learning sense might not be explicitly defined or might be integrated into the algorithm development and validation process rather than being a distinct "training set" for a separate machine learning model from a clinical dataset.
If any machine learning or signal processing algorithms were fine-tuned, the training data size would depend on the complexity of the algorithm and the variability of the input signal. This data would typically consist of electropherogram patterns with known hemoglobin compositions.

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

This information is not provided in the given document.

Reasoning: Similar to the training set size, if there was a "training set" for internal algorithm development, the ground truth would be established by analyzing samples with known hemoglobin compositions using established reference methods (HPLC, confirmed CE, perhaps genetic testing). This allows the algorithm to learn to accurately identify and quantify the various peaks corresponding to normal hemoglobins and common variants.

Summary

Regulation Number and Section

§ 864.7415 Abnormal hemoglobin assay.

(a)
Identification. An abnormal hemoglobin assay is a device consisting of the reagents, apparatus, instrumentation, and controls necessary to isolate and identify abnormal genetically determined hemoglobin types.(b)
Classification. Class II (special controls). A control intended for use with an abnormal hemoglobin assay is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 864.9.