(61 days)
Premier C. difficile GDH is a qualitative enzyme immunoassay screening test to detect Clostridium difficile antigen, glutamate dehydrogenase, in fecal specimens from symptomatic persons suspected of having C. difficile infection (CDI). This test does not distinguish between toxigenic strains of C. difficile. Samples from symptomatic patients that produce positive results with this test must be further tested with an assay designed to detect toxigenic C. difficile strains and assist with the diagnosis of CDI.
Premier C. difficile GDH is a qualitative enzyme immunoassay screening test to detect Clostridium difficile antigen, glutamate dehydrogenase, in fecal specimens from symptomatic persons suspected of having C. difficile infection (CDI). The assay consists of Premier C. difficile GDH Microwells coated with polyclonal antibodies specific to C. difficile GDH, Premier C. difficile GDH Enzyme Conjugate, Premier 20X Wash Buffer II, Premier Substrate I, Premier Stop Solution I, Premier C. difficile GDH Sample Diluent/Negative Control, and Premier C. difficile GDH Positive Control.
Here's a summary of the acceptance criteria and study findings for the Premier C. difficile GDH assay, based on the provided text:
Acceptance Criteria and Device Performance
The document implicitly defines acceptance criteria through the reported performance characteristics. The primary measure of clinical performance is the comparison to bacterial C. difficile culture.
Table 1: Acceptance Criteria and Reported Device Performance (Clinical Study)
| Performance Metric | Acceptance Criteria (Implicit) | Reported Device Performance | Comments |
|---|---|---|---|
| Clinical Sensitivity | High % (e.g., above 85-90%) | 92.3% (95% Cl: 86.0 - 95.9%) | Met the expectation for a high sensitivity screening test. |
| Clinical Specificity | High % (e.g., above 90-95%) | 95.8% (95% Cl: 93.9 – 97.1%) | Met the expectation for a high specificity screening test. |
| Overall Correlation | High % (e.g., above 90%) | 95.2% (93.4 - 96.5%) | Good overall agreement with the reference method. |
| Analytical Sensitivity (LoD) | Defined as detectable at a specific concentration with 95% probability | 8 ng/mL (based on 45 replicates per measurand with 95% probability of positive response) | Clear analytical limit of detection established. |
| Interference Testing | No interference at specified concentrations of common substances | No interference observed for listed substances (e.g., Barium sulfate, Metronidazole, Vancomycin HCl) | Demonstrated robustness against common interfering substances. |
| Cross-Reactivity | No cross-reactivity with common microorganisms, or identified and noted | No cross-reactivity observed with a wide range of bacteria and viruses, except for Staphylococcus aureus (Cowan strain I) and Clostridium sporogenes. | Most common pathogens did not cross-react, but two specific Clostridium species and Staphylococcus aureus were noted as cross-reactive. |
| Strain Reactivity | Positive reactions with a representative panel of C. difficile strains | Positive reactions at 5.7 x 10^7 cells/mL with 30+ strains | Demonstrated ability to detect various C. difficile strains. |
| Reproducibility | 100% agreement for moderate positive, high negative, and negative samples | 100% agreement over 5 non-consecutive days, across 3 sites, 2 operators per site. | Excellent reproducibility across different sites and operators. |
Study Information
-
Sample sizes used for the test set and the data provenance:
- Test Set Sample Size: 733 qualified patient samples.
- Data Provenance: The data was collected prospectively (clinical trials conducted from November 2011) from independent clinical test sites located in the Midwestern and Southwestern regions of the United States. Gender and age ranges were reported (22 days to 99 years).
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The document states that the performance characteristics were determined by comparison to bacterial C. difficile culture. It does not mention the use of human experts to establish ground truth for the clinical test set; rather, the gold standard for diagnosis was a laboratory method (bacterial culture).
-
Adjudication method for the test set:
- Since the ground truth was established by bacterial C. difficile culture, an expert adjudication method (like 2+1, 3+1) was not described or necessary. The comparison was directly against the culture results.
-
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 MRMC comparative effectiveness study was done. This device is an Enzyme Immunoassay (ELISA) kit, which is a laboratory test where results are read spectrophotometrically or visually from a microplate, not an imaging device requiring human interpretation of complex visual data. Therefore, the concept of "human readers improve with AI vs without AI assistance" is not applicable.
-
If a standalone (i.e. algorithm only without human-in-the loop performance) was done:
- Yes, the clinical performance described (sensitivity, specificity, correlation) represents the standalone performance of the Premier C. difficile GDH assay. It is an "algorithm only" in the sense that it is a biochemical assay designed to yield a direct result (positive/negative) based on antigen detection, without human interpretation influencing the diagnostic outcome beyond standard laboratory procedures (e.g., pipetting, reading the spectrophotometer).
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- The ground truth used for the clinical performance evaluation was bacterial C. difficile culture.
-
The sample size for the training set:
- The document does not explicitly state a "training set" in the context of machine learning or AI models. For an ELISA kit, development typically involves analytical studies (sensitivity, specificity, interference, cross-reactivity, strain reactivity) and then clinical validation. The "analytical sensitivity" study used 45 replicates for each measurand. The "reproducibility" panels involved blind-coded samples tested multiple times. These studies contribute to the device's development and validation but are not a "training set" in the common AI sense.
-
How the ground truth for the training set was established:
- Again, the concept of a "training set" in the AI sense is not directly applicable. For the analytical studies, the "ground truth" (e.g., known concentration of C. difficile GDH antigen, presence/absence of interfering substances, known microorganisms) was established through controlled laboratory spiking and preparation of contrived samples. For the broader validation, bacterial C. difficile culture served as the reference standard.
§ 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.