Search Results

The Diamedix Is anti-Cardiolipin IgG/IgM Test Kit is an indirect enzyme immunoassay (EIA) for the semi-quantitative measurement of IgG or IgM antibodies to cardiolipin in human serum as an aid in the assessment of the risk of thrombosis in patient with SLE or SLE-like disorders. The assay can be used either manually or in conjunction with the MAGO® Plus Automated EIA Processor.

Device Description

The Is anti-Cardiolipin IgG/IgM Test System is an enzyme-linked immunosorbent assay (ELISA) for the semi-quantitative measurement of IgG or IgM antibodies to cardiolipin in human serum

AI/ML Overview

Here's a breakdown of the acceptance criteria and study details for the Diamedix Is-anti-Cardiolipin IgG/IgM Test System, based on the provided 510(k) summary:

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state pre-defined quantitative acceptance criteria with specific numerical targets. Instead, it describes performance characteristics and then presents the results. The implicit acceptance criteria are that the device demonstrates comparable performance to a commercially available predicate device and satisfactory analytical and clinical performance.

Performance Metric	Indication of Acceptance Criteria (Implicit)	Reported Device Performance
Calibration Equivalence	3-point vs. 6-point calibration should show strong correlation.	IgG: Y = 2.4036 + 0.9996X (r = 0.9861, n=172). IgM: Y = 0.5333 + 0.9411X (r = 0.9925, n=162). (Strong linear correlation demonstrated for both.)
Relative Sensitivity (IgG)	Should be comparable to commercially available ELISA kit.	91.5% (95% CI: 79.6-97.6%)
Relative Specificity (IgG)	Should be comparable to commercially available ELISA kit.	100.0% (95% CI: 96.6-100.0%)
Overall Agreement (IgG)	Should be comparable to commercially available ELISA kit.	97.4% (95% CI: 93.5-99.3%)
Relative Sensitivity (IgM)	Should be comparable to commercially available ELISA kit.	78.4% (95% CI: 67.3-87.1%)
Relative Specificity (IgM)	Should be comparable to commercially available ELISA kit.	100.0% (95% CI: 95.8-100.0%)
Overall Agreement (IgM)	Should be comparable to commercially available ELISA kit.	90.1% (95% CI: 84.4-94.2%)
Clinical Specificity (IgG)	High specificity in normal population, acceptable specificity in other autoimmune diseases.	Normals: 99.5% (213/214), RPR Positive: 73.3% (11/15)
Clinical Specificity (IgM)	High specificity in normal population, acceptable specificity in other autoimmune diseases.	Normals: 98.5% (211/214), RPR Positive: 73.3% (11/15)
Clinical Sensitivity (IgG)	Detectable in patients with APS and SLE.	APS: 82.5% (47/57), SLE: 21.2% (7/33), Other Autoimmune Diseases: 8.6% (3/35)
Clinical Sensitivity (IgM)	Detectable in patients with APS and SLE.	APS: 47.4% (27/57), SLE: 21.2% (7/33), Other Autoimmune Diseases: 14.2% (5/35)
Cross Reactivity	Should show minimal cross-reactivity with other common autoantibodies.	Out of 36 samples tested (SSA/SSB, Scl-70, Jo-1, dsDNA, RF), one sample positive for dsDNA showed a positive result. The remaining 34 samples were negative. (Suggests minimal cross-reactivity with most, but potential with dsDNA.)
Linearity (IgG)	Demonstrates a linear response over a range of concentrations.	Y = -4.9960 + 147.5101 X (r = 0.9957, R² = 0.9915).
Linearity (IgM)	Demonstrates a linear response over a range of concentrations.	Y = 4.0837 + 87.9931 X (r = 0.9849, R² = 0.9701).
Manual vs. Automated (IgG)	Manual and MAGO Plus methods should show strong correlation.	Y = 1.1058 + 1.1976 X (r = 0.9883, n=172). 6-point (automated vs. manual): (automated) = 1.0696 (manual) + 4.0821; 95% CI slope [1.0174 to 1.1218], intercept [1.3042 to 6.8600]. (Strong linear correlation demonstrated.)
Manual vs. Automated (IgM)	Manual and MAGO Plus methods should show strong correlation.	Y = 1.3732 + 1.1510 X (r = 0.9917, n=162). 6-point (automated vs. manual): (automated) = 1.0169 (manual) + 2.2121; r=0.9772; 95% CI slope [0.9824 to 1.0514], intercept [1.1343 to 3.2899]. (Strong linear correlation demonstrated.)
Precision (CV%)	Coefficients of Variation (CV%) for intra-assay and interassay should be within acceptable limits (typically <15-20% for immunoassays).	Manual IgG: Interassay CV% range: 4.42-13.93%. MAGO Plus IgG: Interassay CV% range: 10.17-29.69%. Manual IgM: Interassay CV% range: 4.84-20.03%. MAGO Plus IgM: Interassay CV% range: 6.79-17.62%. (Generally acceptable, with some higher CVs for very low and high range samples, particularly for MAGO Plus IgG. Smallest values often have higher CVs inherently.)
Limit of Detection (Analytical Sensitivity)	Low detection limit for clinical relevance.	Determined as 0.4 GPL or MPL U/ml.

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

Calibration Equivalence:
- IgG: 172 samples
- IgM: 193 samples
- Provenance: Retrospective sera (not specified by country).
Relative Sensitivity and Specificity:
- IgG: 172 frozen retrospective sera
- IgM: 194 frozen retrospective sera
- Provenance: Retrospective sera (not specified by country).
Clinical Sensitivity and Specificity:
- Total: 354 frozen retrospective, clinically characterized sera
  - 214 normal sera
  - 57 sera from patients with Antiphospholipid Syndrome (APS)
  - 33 sera from patients with Systemic Lupus Erythematosus (SLE)
  - 35 sera from patients with other autoimmune diseases (Sjögren's, scleroderma, polymyositis/dermatomyositis, rheumatoid arthritis)
  - 15 sera from RPR-positive patients
- Provenance: Retrospective sera (not specified by country).
Cross Reactivity: 36 samples (positive for SSA/SSB, Scl-70, Jo-1, dsDNA, and RF antibodies). Provenance: Not specified.
Linearity: "Several highly positive samples" were serially diluted. (Number not specified). Provenance: Not specified.
Correlation of Manual and MAGO Plus Results:
- IgG: 172 serum samples
- IgM: 162 serum samples
- Provenance: Not specified.
Precision: 6 serum samples of varying reactivity tested. Provenance: Not specified.
Expected Values (Normal Population): 148 healthy blood donors from South Florida. Provenance: Prospective, South Florida, USA.
Expected Values (Clinical Population): 57 sera from patients with lupus anticoagulant antibodies. Provenance: Not specified.

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

The studies described are in vitro diagnostic (IVD) performance studies, not typically requiring human readers or expert radiologists to establish ground truth.

For Relative Sensitivity and Specificity, a "commercially available ELISA kit for detecting IgG and/or IgM cardiolipin antibodies" and a "referee EIA method" were used as the comparator/ground truth. No human expert adjudication is mentioned as part of this.
For Clinical Sensitivity and Specificity, samples were "clinically characterized sera" from specific patient groups (Normals, APS, SLE, Other Autoimmune Diseases, RPR Positive). The characterization of these patient groups (e.g., diagnosis of APS or SLE) would have been established by clinical experts (e.g., rheumatologists) based on diagnostic criteria, but the document does not specify how many or the qualifications of these clinicians, as it's assumed to be part of the standard diagnostic process for collecting such samples. The ground truth here is the patient's clinical diagnosis.

4. Adjudication Method for the Test Set

No explicit adjudication method (like 2+1 or 3+1) is mentioned or appears applicable, as the ground truth is established either by:

Comparison to another legally marketed (predicate/referee) assay.
Pre-existing clinical diagnosis of patient samples.

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. This device is an in-vitro diagnostic (ELISA) kit. It is designed to directly measure antibody levels in human serum, not to assist human readers in interpreting medical images or other complex data. Therefore, an MRMC study and the concept of "human readers improving with AI assistance" are not applicable to this device.

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

Yes, in essence. The primary performance evaluations (Sensitivity, Specificity, Linearity, Precision, Cross-reactivity) were performed on the device itself, either manually or using the automated MAGO® Plus system. The results are quantitative measurements from the assay, not interpretations by an algorithm that would then be compared to human interpretation. The "algorithm" in this context is the assay's biochemical reaction and detection system.

7. The Type of Ground Truth Used

Relative Sensitivity and Specificity: Comparison to a "commercially available ELISA kit" and a "referee EIA method."
Clinical Sensitivity and Specificity: "Clinically characterized sera" (i.e., outcomes/clinical diagnosis data for patient groups like APS, SLE, Normals).
Calibration Equivalence, Linearity, Precision, Manual vs. Automated Correlation: The ground truth is the expected output based on known standards or a well-established reference measurement method (e.g., higher concentration samples, manual method as a reference for automated).
Cross Reactivity: Known positive samples for other autoantibodies.
Expected Values: Clinically healthy or diagnosed patient populations.

8. The Sample Size for the Training Set

The document does not explicitly describe a "training set" in the context of machine learning or AI models. This is a traditional IVD device, not an AI/ML product. The samples used for validation were used to characterize the device's performance directly.

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

Not applicable, as there is no explicitly defined "training set" for an AI/ML model. The ground truth for the various performance characteristic studies was established as described in point 7.

Ask a Question

Ask a specific question about this device

Page 1 of 1