VIDAS C. DIFFICILE TOXIN A II ASSAY by BIOMERIEUX VITEK, INC.

Acceptance Criteria Category	Specific Metric	Acceptance Criteria (Implied/Expected)	Reported Device Performance
False Positive Results	Specificity	High (e.g., >95%)	99.1%
False Negative Results	Sensitivity	High (e.g., >90%)	94.9%
Equivocal Results	Occurrence	Low (e.g., <5%)	2.5%
Invalid Results	Occurrence	Zero	0%
Precision	Intra-assay CV	Low variability	CDA 2: 2.7 - 31.9%CDB: 19.0 - 39.2%
	Inter-assay CV	Low variability	CDA 2: 8.6 - 30.8%CDB: 9.1 - 33.4%
Assay Specificity	Cross-reactivity/Interference	None with normal flora	Cross-reactivity/interference seen with C. sordellii and C. difficile toxin-producing strains at high concentrations.

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

Positive Specimens: 136
Negative Specimens: 1,407
Total Test Set Size: 1,543 specimens (136 + 1,407)
Data Provenance: Not specified in the provided text (e.g., country of origin, retrospective or prospective).

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

The ground truth was established by comparing the VIDAS assay to a commercially available EIA with discrepants resolved by cytotoxicity.
The text does not specify the number of experts involved in resolving discrepancies or their qualifications.

4. Adjudication Method for the Test Set

Discrepancy resolution by cytotoxicity. This implies that for cases where the VIDAS assay and the reference EIA disagreed, an additional, more definitive test (cytotoxicity) was used as the tie-breaker to establish the true positive or negative status. It does not mention whether multiple human readers were involved in interpreting the cytotoxicity results or if an adjudication rule (e.g., 2+1) was used for these.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and the Effect Size of How Much Human Readers Improve with AI vs. Without AI Assistance

No, an MRMC comparative effectiveness study was not done. This study is for an in-vitro diagnostic (IVD) assay, not an AI-powered image analysis tool that assists human readers. Therefore, the concept of human reader improvement with/without AI assistance does not apply.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

Yes, this entire study represents a standalone performance evaluation of the VIDAS C. difficile Toxin A II assay. It measures the performance of the assay itself without human interpretation as part of its primary function (beyond potentially interpreting the reference EIA or cytotoxicity results, which are part of the ground truth establishment, not the device's operational performance).

7. The Type of Ground Truth Used

The primary ground truth for establishing sensitivity and specificity was a commercially available EIA with discrepants resolved by cytotoxicity. Cytotoxicity is a functional assay that directly assesses the biological activity of the C. difficile toxin, making it a strong indicator of true positivity for toxin-associated disease.

8. The Sample Size for the Training Set

The document does not specify a training set size. Immunoassays like the VIDAS C. difficile Toxin A II assay typically do not have "training sets" in the same way machine learning algorithms do. Their performance is inherent in their chemical and biological design. Development and optimization might involve various samples, but these are not usually referred to as a "training set" in the context of regulatory submissions for such devices.

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

As a training set is not explicitly mentioned or applicable in the machine learning sense, this question is not relevant to the provided text. The "training" of such an assay involves laboratory development, reagent optimization, and calibration, for which ground truth would be established through established laboratory methods and reference materials.

Summary

{0}------------------------------------------------

K96 4887

APR - 8 1997

SUMMARY: Safety and Effectiveness Information for the VIDAS C. difficile Toxin A II assay

The VIDAS C. difficile Toxin A II assay detects the presence of C. difficile toxin A. It is substantially equivalent to the Meridian Premier C. difficile Toxin A test as an aid in the diagnosis of Clostridium difficile associated disease (CDAD). Safety and effectiveness issues for qualitative enzyme immunoassay such as the VIDAS C. difficile Toxin A II assay may include the following:

1. False positive results: In studies comparing the VIDAS C. difficile Toxin A II assay to a commercially available EIA with discrepants resolved by cytotoxicity, of the 1.407 negative specimens there were 13 false positive results for a specificity of 99.1%.
1. False negative results: In studies comparing the VIDAS C. difficile Toxin A II assay to a commercially available EIA with discrepants resolved by cytotoxicity, of the 136 positive specimens there were 7 false negative results for a sensitivity of 94.9%.
1. Equivocal results: The occurrence of equivocals with the VIDAS CDA 2 test was 2.5% in the studies done to demonstrate the performance of the assay. All equivocals were resolved with the use of the VIDAS C. difficile Blocking Reagents in the VIDAS CDB test of the VIDAS C. difficile Toxin A II assay.
1. Invalid results: In the studies done to support the VIDAS C. difficile Toxin A II assay performance claims, there were no invalid VIDAS results.
1. Precision: In the studies done to support the VIDAS C. difficile Toxin A II assay performance claims, the VIDAS CDA 2 test intra-assay precision testing showed coefficients of variation for test values ranging from 2.7 to 31.9 % when utilizing stool precision pools. For the VIDAS CDB test intra-assay precision testing showed coefficients of variation for test values ranging from 19.0 to 39.2 %. When utilizing the same stool precision pools, the VIDAS CDA 2 test inter-assay precision testing gave coefficients of variation for the test values ranging from 8.6 to 30.8% and coefficients of variation for the test values ranging from 9.1 to 33.4% for the VIDAS CDB test.
1. Assay Specificity: A panel of normal human colonic flora was tested for both crossreactivity and interference for the VIDAS C. difficile Toxin A II assay. Crossreactivity and interference was seen with the C. sordellii strain (VPI #9048) when tested at a concentration of 3 x 10 organisms/ml and the C. difficile toxin producing strains (ATCC #9689 and VPI # 10463) when tested at a concentration of 1 x 10' organisms/ml.

The VIDAS C. difficile Toxin A II assay must be used according to package insert instructions when testing for the presence of C. difficile toxin A. Additional information and references may be found in the package insert.

Regulation Number and Section

§ 866.2660 Microorganism differentiation and identification device.

(a)
Identification. A microorganism differentiation and identification device is a device intended for medical purposes that consists of one or more components, such as differential culture media, biochemical reagents, and paper discs or paper strips impregnated with test reagents, that are usually contained in individual compartments and used to differentiate and identify selected microorganisms. The device aids in the diagnosis of disease.(b)
Classification. Class I (general controls). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to § 866.9.