Search Results

CRYOcheck Factor VIII Deficient Plasma with VWF is for clinical laboratory use as a deficient substrate in the quantitative determination of Factor VIII activity in 3.2% citrated human plasma based on the activated partial thromboplastin time (APTT) assay. It is intended to be used in identifying factor VIII deficiency and as an aid in the management of hemophilia A in individuals aged 2 years and older. For in vitro diagnostic use

CRYOcheck Factor VIII Deficient Plasma with VWF is normal human citrated plasma which has been immunodepleted of factor VIII and to which an exogenous source of human von Willebrand Factor (vWF) has been added and buffered with HEPES. Factor VIII has been assayed at less than 1% of normal activity levels and vWF antigen and activity are >50%. It will be provided to users frozen in small-volume aliquots (25 vials of 1.0 mL, and 25 vials of 1.5 mL). Vials will be packaged into boxes; these will be frozen during the manufacturing process and will be shipped and stored frozen until use to preserve the integrity of the components

The provided FDA 510(k) summary for the "CRYOcheck Factor VIII Deficient Plasma with VWF" describes various performance studies and their results. Based on the document, here's a structured breakdown of the acceptance criteria and the study details:

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

The document does not explicitly present a table of pre-defined acceptance criteria for each performance characteristic in a pass/fail format. Instead, it presents the results of various studies and implies that these results met internal or regulatory expectations for substantial equivalence to the predicate device. However, we can infer performance targets based on the documented results and common regulatory expectations for in vitro diagnostic devices.

Here's a table summarizing the reported device performance, with inferred acceptance "criteria" based on the conclusions drawn in the report:

2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

• Multi-Reagent Lot Precision: Not explicitly stated as a "test set" sample size for patient plasma, but control samples were tested 80 replicates per sample per lot across 3 lots.
• Multi-Reagent Lot Site to Site Reproducibility: 270 replicates per sample (control or plasma) for the pooled data, across 3 sites and 3 lots. Patient samples included high, low, and very low FVIII plasma.
• Linearity/Assay Reportable Range: Fifteen sample dilutions, each tested in quadruplicate.
• Reference Interval: 136 normal, ostensibly healthy individuals.
• Shelf-Life Stability: Five replicates of one normal and two abnormal reference controls and two patient plasma samples per timepoint, across 3 lots.
• In-Use Stability: Five replicates of one normal and two abnormal reference controls and two patient plasma samples per timepoint, across 3 lots.
• Interferences: 10 replicates of spiked samples and 10 replicates of control for each possible interferent.
• Recovery of FVIII Replacements: Congenital FVIII deficient plasma spiked with six FVIII replacement therapies at seven concentrations (total 42 conditions).
• Method Comparison: 366 human plasma samples (from normal ostensibly healthy individuals, patients with congenital or acquired hemophilia A, patients with hemophilia B, patients with von Willebrand disease, hemophilia A patients on FVIII replacement therapies, and patients with other factor deficiencies). These samples were distributed across four sites.
• Sample Integrity: 94 plasma samples.

Data Provenance: The document does not specify the country of origin for any of the patient samples. All studies appear to be prospective in design, as they describe the testing of samples under controlled conditions to evaluate device performance characteristics.

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 in vitro diagnostic (IVD) reagent for determining Factor VIII activity in human plasma, which is a quantitative measurement, not an interpretative one like image analysis. Therefore, the concept of "experts establishing ground truth" in the way it applies to diagnostic algorithms (e.g., radiologists for imaging) is not directly relevant here.

The "ground truth" for the performance studies is established by:

• The known concentrations/characteristics of reference controls.
• The actual FVIII activity levels as measured by a legally marketed predicate device (HemosIL Factor VIII Deficient Plasma) in the method comparison study.
• The carefully prepared spiked samples for linearity and recovery studies.
• The clinical diagnosis categories of patient samples (e.g., congenital hemophilia A).

There is no mention of human experts (e.g., pathologists, hematologists) adjudicating "ground truth" for individual cases in these studies beyond ensuring appropriate sample selection and handling. The performance is assessed against quantitative benchmarks and comparison with a known predicate.

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

As mentioned above, the device performs a quantitative measurement, not an interpretative diagnosis requiring adjudication. Therefore, no adjudication method (like 2+1 or 3+1 for expert consensus on image interpretation) was used or is applicable for this type of IVD device testing. The "ground truth" is analytical and quantitative.

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 is not applicable as this is an in vitro diagnostic reagent, not an AI-powered diagnostic tool that assists human readers. The device itself performs the measurement, and its performance is evaluated directly through analytical studies, not through its impact on human reader performance.

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

This refers to the performance of the device itself (the reagent in combination with instrumentation) without human interpretation affecting the measurement result. All the analytical performance studies (precision, reproducibility, linearity, stability, interferences, recovery, method comparison, sample integrity) represent the standalone performance of the CRYOcheck Factor VIII Deficient Plasma with VWF in conjunction with the specified automated APTT assay systems. The "human-in-the-loop" here refers to the user performing the test, but the performance characteristics relate to the analytical accuracy and reliability of the automated test method itself.

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

The ground truth used in these studies is primarily analytical and quantitative:

• Known concentrations/values of commercially available reference controls.
• Known FVIII activity levels measured by a predicate device for method comparison.
• Expected FVIII activity levels in carefully fabricated samples (e.g., highly concentrated plasma diluted with congenital hemophilia A plasma for linearity, spiked samples for recovery).
• Clinical classification of patient samples (e.g., "patients with congenital or acquired hemophilia A") as the basis for sample selection in the method comparison and reference interval studies, though specific individual diagnoses are not the "ground truth" for each quantitative measurement, rather the context.

8. The sample size for the training set

This document describes a premarket notification (510(k)) for an in vitro diagnostic reagent. IVD reagents are not typically "trained" in the machine learning sense. Therefore, there is no training set in the context of an AI/ML algorithm. The studies detailed are analytical verification and validation studies to demonstrate performance and establish substantial equivalence to a predicate device.

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

Since there is no training set for this type of device, this question is not applicable.

Ask a specific question about this device