(142 days)
The Stellar Bio Systems' Indirect Fluorescence Assay (IFA) for Epstein-Barr Early Antigen (EBV-EA) IgG Antibody is intended for the qualitative and semi-quantitative detection of IgG (Immunoglobulin G) antibody to the early antigen complex of the Epstein-Barr Virus in human serum. The test detects IgG antibodies to both the diffuse (D) and restricted (R) components of the EBV early antigen complex. The EBV-EA IFA Test Kit should be used in combination with other Epstein-Barr serologies (Viral Capsid Antigen (VCA) IgG and IgM, Epstein-Barr Nuclear Antigen (EBNA), and heterophile antibody) as an aid in the diagnosis of infectious mononucleosis (IM).
The Epstein-Barr Virus Early Antigen (EBV-EA) IgG IFA is an Indirect Fluorescence Assay (IFA) for the detection of IgG antibodies to EBV-EA in human serum.
Here's an analysis of the acceptance criteria and study details for the Stellar Bio Systems' Indirect Fluorescence Assay (IFA) for Epstein-Barr Virus Early Antigen (EBV-EA) IgG Antibody, based on the provided text:
Acceptance Criteria and Device Performance
The acceptance criteria are not explicitly stated as numerical thresholds for sensitivity and specificity in the provided text. Instead, the study reports the device's performance relative to a commercially available predicate EBV-EA IFA. The implication is that comparable performance to the predicate device would be a key acceptance criterion.
Table 1: Relative Sensitivity and Specificity of the Stellar EBV-EA IFA Kit Relative to Alternate IFA
| Metric | Acceptance Criteria (Implied: Comparable to Alternate IFA) | Reported Device Performance (Stellar EBV-EA IFA) |
|---|---|---|
| Relative Sensitivity | Not explicitly stated, likely expected to be high | 42/43 = 97.7% (with the alternate IFA as reference) |
| Relative Specificity | Not explicitly stated, likely expected to be high | 26/32 = 81.3% (with the alternate IFA as reference) |
| Overall Agreement | Not explicitly stated, likely expected to be high | 68/75 = 90.7% (95% CI: 81.7% - 96.2%) |
| Titer Agreement (Identical) | Not explicitly stated, likely expected to be high | 9/25 = 36% |
| Titer Agreement (+/- one two-fold dilution) | Not explicitly stated, likely expected to be high | 12/25 = 48% |
| Titer Agreement (+/- two two-fold dilutions) | Not explicitly stated, likely expected to be high | 4/25 = 16% |
| Reproducibility (Identical End Point Titer) | Not explicitly stated, likely expected to be high | 131/180 = 72.8% |
| Reproducibility (Within +/- one two-fold dilution) | Not explicitly stated, likely expected to be high | 49/180 = 27.2% |
| Specificity (Cross-reactivity) | No cross-reactivity with CMV, VZV, HSV 1, HSV 2, HHV 6, HHV 8 | Found to be negative for EBV-EA IFA in these cases |
Note on Acceptance Criteria: The document explicitly states: "Please be advised that 'relative' refers to the comparison of this assay's results to that of a similar assay. There was not an attempt to correlate the assay's results with disease presence or absence. No judgment can be made on the comparison assay's accuracy to predict disease." This highlights that the acceptance criteria are based on agreement with a predicate device, not on direct clinical truth or diagnostic accuracy.
Study Details
2. Sample size used for the test set and the data provenance
- Test Set Sample Size for Relative Sensitivity and Specificity: 75 samples (43 positive, 32 negative by the alternate IFA).
- Test Set Sample Size for Titer Agreement: 25 positive sera.
- Test Set Sample Size for Reproducibility: 180 (3 positive sera with various titers, 1 negative sera, each assayed 5 times on 3 different assays at 3 different sites).
- Test Set Sample Size for Specificity (Cross-reactivity): At least 9 samples (CMV+, VZV+, HSV 1&2+, HHV 6+, HHV 8+; note multiple HSV 1&2+ and HHV 6+ samples are listed).
- Data Provenance: The samples were selected from frozen retrospective sera tested at a commercial clinical laboratory. The document doesn't explicitly state the country of origin, but given the applicant's address (Columbia, MD 21046, USA), it's highly probable the data is from the USA.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
The ground truth for the primary relative sensitivity and specificity study was established by the results of a commercially available EBV-EA IFA (the "Alternate IFA"). There is no mention of human experts directly establishing the ground truth for this comparison.
For the categorization of sera by serological status (Acute, Seropositive, Seronegative), the criteria were based on ELISA results for VCA IgM/IgG and EBNA. Again, this points to laboratory assay results, not human expert consensus, as the ground truth.
4. Adjudication method for the test set
No adjudication method involving multiple human readers is mentioned. The ground truth was determined by the results of established laboratory assays (the Alternate IFA and various EBV ELISAs).
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, a multi-reader multi-case (MRMC) comparative effectiveness study was not conducted or reported. This device is a diagnostic assay (IFA), not an AI-assisted interpretation tool for human readers.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
Yes, the performance data presented is for the standalone device (the Stellar EBV-EA IFA kit). The "human-in-the-loop" equivalent for an IFA assay would typically be a lab technician reading the fluorescence under a microscope. The study design tests the performance of the kit itself in producing results compared to other similar kits.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
The ground truth used was primarily the results from a commercially available predicate EBV-EA IFA kit or other serological assays (VCA IgG/IgM ELISA, EBNA ELISA, heterophile antibody assay), rather than expert consensus, pathology, or outcomes data. The study explicitly states there was no attempt to correlate assay results with disease presence or absence.
8. The sample size for the training set
The document does not provide information on a specific training set or its size. As this is a traditional in-vitro diagnostic (IVD) kit based on an established IFA methodology, it's unlikely to have a "training set" in the machine learning sense. The assay development would involve optimizing reagents and protocols during its design phase, but not typically a labeled dataset for algorithm training.
9. How the ground truth for the training set was established
Since no training set for an algorithm is mentioned, the method for establishing its ground truth is not applicable/not provided.
§ 866.3235 Epstein-Barr virus serological reagents.
(a)
Identification. Epstein-Barr virus serological reagents are devices that consist of antigens and antisera used in serological tests to identify antibodies to Epstein-Barr virus in serum. The identification aids in the diagnosis of Epstein-Barr virus infections and provides epidemiological information on diseases caused by these viruses. Epstein-Barr viruses are thought to cause infectious mononucleosis and have been associated with Burkitt's lymphoma (a tumor of the jaw in African children and young adults) and postnasal carcinoma (cancer).(b)
Classification. Class I (general controls).