(62 days)
Enzyme linked immunosorbent assay method for the semi-quantitative determination of specific IgG autoantibodies to Sm in human serum.
Uses:
The results of the anti-Sm assay can be used as an aid in the diagnosis of auto-immune diseases including Systemic Lupus Erythematosus (SLE).
For in vitro diagnostic use only.
Not Found
The provided text is related to a 510(k) premarket notification for an in vitro diagnostic device, specifically an ELISA method for detecting anti-Sm antibodies. This type of device is a laboratory test, not an imaging AI device, and therefore the concepts of "device performance," "ground truth," "multi-reader multi-case (MRMC) comparative effectiveness study," and "standalone (algorithm only) performance" as typically applied to imaging AI are not directly applicable in the same way.
Instead, the performance of such a device is typically assessed through analytical and clinical validation studies that measure parameters like sensitivity, specificity, accuracy, precision, linearity, and interference. The "acceptance criteria" would be defined thresholds for these parameters to demonstrate that the device performs equivalently or better than a legally marketed predicate device.
Given that context, here's an attempt to answer your request based on the spirit of the questions, inferring typical validation practices for this type of device, and acknowledging the limitations of the provided document which is a regulatory clearance letter, not a detailed study report.
1. Table of Acceptance Criteria and Reported Device Performance
The provided FDA clearance letter (K990625) does not explicitly detail the acceptance criteria or the reported performance data from the study submitted by Cogent Diagnostics Limited for their Autostat™ II Anti-Sm ELISA. For in vitro diagnostics like ELISA, typical acceptance criteria and performance metrics would include:
| Metric | Typical Acceptance Criteria (Hypothetical for ELISA) | Reported Device Performance (Not explicit in document) |
|---|---|---|
| Sensitivity | ≥ 90% (against a gold standard or predicate device for Systemic Lupus Erythematosus) | Not reported in provided document. |
| Specificity | ≥ 90% (against healthy controls or patients with other autoimmune diseases) | Not reported in provided document. |
| Accuracy | ≥ 90% overall agreement (against a predicate device or clinical diagnosis) | Not reported in provided document. |
| Precision | Coefficient of Variation (CV) ≤ 15% (intra-run, inter-run, lot-to-lot) | Not reported in provided document. |
| Agreement with Predicate Device | Substantial equivalence demonstrated through high correlation/concordance. | The FDA letter states "substantially equivalent". |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Test Set: Not explicitly stated in the provided FDA clearance letter. For an ELISA device, a clinical validation study typically involves hundreds to low thousands of serum samples from both patients with the target condition (SLE in this case) and appropriate control groups.
- Data Provenance: Not explicitly stated in the provided FDA clearance letter. Such studies often involve samples collected retrospectively or prospectively from multiple clinical sites, potentially across different countries, to ensure generalizability. The manufacturer, Cogent Diagnostics Limited, is based in the United Kingdom, suggesting data could originate from there or other international centers. This would be a clinical validation study.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
For an ELISA used as an aid in diagnosis, the "ground truth" for the test set is established through clinical diagnosis of Systemic Lupus Erythematosus (SLE) and related conditions.
- Number of Experts: Not explicitly stated. The diagnosis of SLE is typically a clinical diagnosis made by rheumatologists. These experts would use established diagnostic criteria (e.g., ACR or SLICC classification criteria) which integrate clinical findings, other laboratory tests, and potentially biopsy results. It's not a panel of experts reviewing the diagnostic device's output, but rather a consensus of clinical findings and existing diagnostic standards that defines the patient's disease status.
- Qualifications: Likely board-certified rheumatologists with extensive experience in diagnosing and managing autoimmune diseases, particularly SLE.
4. Adjudication Method for the Test Set
Not applicable in the way it's used for imaging AI. The "ground truth" for clinical samples is the patient's diagnosed clinical status based on established medical criteria, not an adjudication of multiple interpretations of a direct screening output. In clinical trials, patient diagnoses are often established by a primary clinician, sometimes confirmed by a second clinician or a diagnostic committee following specific criteria, but this is different from adjudicating interpretations of an AI output.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done
- No, an MRMC study was not done. MRMC studies are specific to imaging interpretation where multiple human readers interpret images (with and without AI assistance) and their performance is compared. This device is an in vitro diagnostic (ELISA), not an imaging device. The "reader" is the laboratory technologist running the assay, and the interpretation is based on a quantitative result compared to a cutoff, not a subjective interpretation of complex visual data.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
- Yes, in spirit. For an in vitro diagnostic (IVD) like the Autostat™ II Anti-Sm ELISA, the primary validation is standalone performance. The "algorithm" here is the entire assay procedure, including reagents, controls, testing protocol, and predetermined cut-off values for interpretation. The performance metrics (sensitivity, specificity, accuracy) are derived from how well the assay's biochemical result (antibody level) correlates with the clinical ground truth. Human involvement is in performing the lab test and interpreting the quantitative result against set thresholds, but the performance of the test itself is evaluated independently of a human's interpretive judgment of a complex visual pattern.
7. The Type of Ground Truth Used
- Expert Consensus / Clinical Diagnosis (using established criteria). The ground truth for evaluating an anti-Sm antibody test is the clinical status of the patient (e.g., definitively diagnosed with SLE, diagnosed with another autoimmune disease, healthy control). This diagnosis is established by clinical experts (rheumatologists) based on comprehensive clinical evaluation, physical examination, and a battery of laboratory tests (including other autoantibodies, inflammatory markers, etc.), often following recognized diagnostic criteria (e.g., ACR or SLICC). It is not pathology in the surgical sense, nor solely outcomes data, but rather a clinical determination.
8. The Sample Size for the Training Set
- Not applicable in the same way as AI. For an ELISA, there isn't a "training set" in the machine learning sense. The assay is developed based on biochemical principles (antigen-antibody binding) and optimized using smaller sets of prototype samples to establish reagent concentrations, reaction conditions, and preliminary cut-off values. These preliminary samples are often used for:
- Assay optimization and linearity studies
- Defining the dynamic range
- Establishing preliminary cut-off values (which are then validated in larger studies)
- Precision and reproducibility studies
- The exact number of samples used for these development and optimization phases is not typically disclosed in regulatory summaries and is distinct from the clinical validation 'test set'.
9. How the Ground Truth for the Training Set was Established
- As above, there isn't a "training set" in the AI sense. The development of the assay relies on known positive and negative samples, where "known" refers to samples from patients with confirmed SLE (positive) and confirmed healthy individuals or those with other non-SLE conditions (negative). The ground truth for these samples would be established through clinical diagnosis by medical experts (rheumatologists) using established diagnostic criteria for SLE.
§ 866.5100 Antinuclear antibody immunological test system.
(a)
Identification. An antinuclear antibody immunological test system is a device that consists of the reagents used to measure by immunochemical techniques the autoimmune antibodies in serum, other body fluids, and tissues that react with cellular nuclear constituents (molecules present in the nucleus of a cell, such as ribonucleic acid, deoxyribonucleic acid, or nuclear proteins). The measurements aid in the diagnosis of systemic lupus erythematosus (a multisystem autoimmune disease in which antibodies attack the victim's own tissues), hepatitis (a liver disease), rheumatoid arthritis, Sjögren's syndrome (arthritis with inflammation of the eye, eyelid, and salivary glands), and systemic sclerosis (chronic hardening and shrinking of many body tissues).(b)
Classification. Class II (performance standards).