(170 days)
The DENV Detect IgM Capture ELISA is for the qualitative detection of IgM antibodies to DEN recombinant antigens (DENRA) in serum for the presumptive clinical laboratory diagnosis of Dengue virus infection. The assay is intended for use only in patients with clinical symptoms consistent with either dengue fever or dengue hemorrhagic fever. Positive results must be confirmed by Plaque Reduction Neutralization Test (PRNT), or by using the current CDC guidelines for diagnosis of this disease.
The DENV Detect™ IgM Capture ELISA is a sandwich-type immunoassay. The test kit includes microtiter wells coated with anti-human IgM antibodies, DEN IgM Negative, and IgM Positive controls, DENV Sample Dilution Buffer, Dengue-derived recombinant antigens (DENRA) and normal cell antigens (NCA). The test kit also contains a HRPlabeled DEN-specific monoclonal antibody and tetramethylbenzidine (TMB) substrate which are used to detect DEN IgM antibodies in the wells.
Here's an analysis of the provided text, extracting the requested information about acceptance criteria and the study proving the device meets them:
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria are not explicitly stated as numerical targets in the provided document. However, the performance metrics observed in the clinical studies can be considered as the reported device performance against implicit expectations for a diagnostic test.
| Performance Metric | Implied Acceptance Criteria (Based on context of device evaluation) | Reported Device Performance (Range from studies) |
|---|---|---|
| Positive Percent Agreement (PPA) / Sensitivity | High for detecting recent infection, especially later in disease course. | Early Infection (2-3 days post-onset): 28.6% |
| Later Infection (11-15 days post-onset): 91.7% | ||
| Overall Study 2 (including equivocal as false neg): 84.0% | ||
| Study 5 (Visit 1): 33.3% | ||
| Study 5 (Visit 2): 79.5% | ||
| Negative Percent Agreement (NPA) / Specificity | High for correctly identifying individuals without dengue infection, especially in non-endemic areas. | Study 1: 78.8% - 100.0% |
| Overall Study 2 (including equivocal as false pos): 88.3% | ||
| Study 3 (No other disease): 99.0% | ||
| Study 3 (Other diseases, no WNV): 100.0% | ||
| Study 3 (WNV positive): 58.8% | ||
| Study 4: 97.0% | ||
| Study 5: 93.8% | ||
| Cross-Reactivity (Specificity) | Minimal with other common infectious agents, especially those with similar symptoms. | West Nile Virus: 50% cross-reactivity (Study 3: 58.8% NPA for WNV positive samples, meaning 41.2% false positives/equivocals) |
| Other diseases listed (EEE, JE, SLE, Hepatitis B/C, EBV, CMV, ANA, VZV, Lyme, Leptospirosis): 0% cross-reactivity (except for one leptospirosis sample confirmed dengue positive by PRNT) | ||
| Reproducibility (Within-run, Day-to-Day, Operator-to-Operator, Site-to-Site) | Consistency and reliability of results across different runs, days, operators, and sites. | (Quantitative values for SD and %CV are provided in the table but no explicit acceptance criteria are given. The presence of the data implies it was evaluated for acceptable manufacturing/performance consistency.) |
| Assay Cut-off Performance | Optimal balance between sensitivity and specificity at the chosen cut-off. | Optimal cut-off for ISR values weighing sensitivity and specificity equally was found with a specificity of 99% and sensitivity of 91%. Equivocal range established for specificity of 94% and sensitivity of 96%. |
2. Sample Size for the Test Set and Data Provenance
The "test set" in this context refers to the samples used in the clinical studies to evaluate the DENV Detect IgM Capture ELISA's performance.
- Study Site 1: 197 subjects (394 total samples collected 1-2 weeks apart). Retrospective. Southeast Asia (reference laboratory).
- Study Site 2: 212 serially collected archived samples. Retrospective. Western United States (reference lab). Majority from Caribbean, southern/southeastern US; minority from Africa/Asia.
- Study Site 3: 289 lab archived samples. Retrospective. Public health lab (country not explicitly stated, but implies US due to specific mention of a US state without dengue outbreaks).
- Study Site 4: 199 archived samples. Retrospective. State Dept. of Health in Southern US.
- Study Site 5: 55 symptomatic subjects (samples collected at presentation and 4-14 days later, so 110 total samples). Prospective. Dengue endemic region in South America.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document does not specify the explicit number or qualifications of experts involved in establishing the ground truth for most studies. It refers to "reference laboratory," "public health laboratory," and "CDC" for confirmatory testing.
- Study 1: Ground truth determined by a "reference laboratory using a diagnostic algorithm (validated in-house IgM test result and/or PCR result, and/or a rising IgG titer, and/or a four-fold rise of HAI titer between acute and convalescent blood draw)." No specific expert qualifications are given beyond the lab's general diagnostic capabilities.
- Study 2: Ground truth mainly established by "confirmatory PRNT" and "CDC Dengue MAC ELISA at the CDC." Again, specific expert numbers/qualifications are not detailed beyond the implication of CDC diagnostic expertise.
- Study 3: Ground truth based on "All testing and diagnosis was performed at the public health laboratory." and "history of dengue incidences in this general area" for assumed dengue negative status. Also PRNT testing was used. No specific expert numbers/qualifications are detailed.
- Study 4: Ground truth established by "CDC Dengue IgM (MAC) ELISA" and "PRNT testing." No specific expert numbers/qualifications are detailed beyond CDC.
- Study 5: Ground truth established by "Confirmatory PRNT testing." "PRNT changes of 4-fold or greater between visits 1 and 2, indicative of current Dengue infection." No specific expert numbers/qualifications are detailed.
It's implied that these reference laboratories and the CDC possess the necessary expertise for performing these confirmatory tests and making diagnoses.
4. Adjudication Method for the Test Set
The document does not describe a formal adjudication committee or method (like 2+1, 3+1). Instead, the "ground truth" (or "final diagnosis") was established by recognized reference methods:
- Study 1: Diagnostic algorithm by the reference lab (validated in-house IgM, PCR, rising IgG titer, four-fold HAI rise).
- Study 2: Confirmatory PRNT and CDC Dengue MAC ELISA. For indeterminate/equivocal CDC MAC ELISA results, PRNT was used for final classification.
- Study 3: Public health laboratory diagnosis, PRNT testing, and epidemiology (dengue-free area).
- Study 4: CDC Dengue IgM (MAC) ELISA and PRNT.
- Study 5: Confirmatory PRNT testing, specifically a 4-fold or greater rise in PRNT titer between two visits.
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
This is an ELISA diagnostic device, not an AI-powered image analysis or diagnostic support system. Therefore, an MRMC comparative effectiveness study involving human readers and AI assistance is not applicable and was not performed. The device provides a quantitative or qualitative result directly, which is then interpreted by a lab professional.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
The DENV Detect IgM Capture ELISA is a standalone diagnostic assay. Its performance, as described in the clinical studies, represents the algorithm-only performance. The ELISA itself is the "algorithm," producing results based on biochemical reactions and optical density measurements. Human involvement is primarily in performing the assay, interpreting the quantitative output (ISR ratios against cut-offs), and making a final clinical diagnosis based on the test result in conjunction with patient symptoms and other clinical information.
7. The Type of Ground Truth Used
The ground truth used varied slightly across studies but generally involved:
- Plaque Reduction Neutralization Test (PRNT): Widely considered the gold standard for serological confirmation of dengue infection.
- CDC Dengue MAC ELISA: A well-established IgM capture ELISA used by the CDC as a reference method.
- PCR (Polymerase Chain Reaction): For direct detection of viral genetic material, indicating active infection.
- Rising IgG Titer / Four-fold rise of HAI titer: Indicative of a seroconversion or a recent infection.
- Expert Consensus / Diagnostic Algorithm: In Study 1, a reference laboratory's diagnostic algorithm combined multiple methods.
- Outcomes Data / Epidemiological Context: In Study 3, the "dengue non-endemic area" was used as a strong indicator of negative status for many samples.
8. The Sample Size for the Training Set
The document does not explicitly state a separate "training set" in the context of machine learning. For traditional in vitro diagnostic (IVD) devices like this ELISA, the "training" or optimization phase typically involves:
- Assay Cut-off Determination: This involved 109 true Positive and 97 True Negative serum samples. This set serves a similar function to a validation set for determining optimal operating points.
The core assay development and optimization (which would be analogous to "training" for a machine learning model) are inherent to the ELISA kit's design and formulation, not a separate data-driven training process in the modern AI sense.
9. How the Ground Truth for the Training Set was Established
For the cut-off determination:
- 109 true Positive and 97 True Negative serum samples were used. How their "true" status was established is not explicitly detailed in this section, but it is implied they were confirmed positive or negative for dengue infection through reliable reference methods, similar to those used in the clinical studies (PRNT, PCR, other validated assays). The statement "true Positive and true Negative" suggests an established and verified diagnosis for these samples.
§ 866.3945 Dengue virus serological reagents.
(a)
Identification. Dengue virus serological reagents are devices that consist of antigens and antibodies for the detection of dengue virus and dengue antibodies in individuals who have signs and symptoms of dengue fever or dengue hemorrhagic fever. The detection aids in the clinical laboratory diagnosis of dengue fever or dengue hemorrhagic fever caused by dengue virus.(b)
Classification. Class II (special controls). The special control is FDA's guideline entitled “Class II Special Controls Guideline: Dengue Virus Serological Reagents.” For availability of the guideline document, see § 866.1(e).