(58 days)
The DiaSorin Molecular Simplexa™ GBS Direct assay is a real-time polymerase chain reaction (PCR) assay intended for use on the LIAISON® MDX instrument for the in vitro qualitative detection of Group B Streptococcus (GBS) nucleic acid from 18 to 24 hour Lim broth enrichments of vaginal/rectal specimen swabs obtained from antepartum women. Assay results can be used as an aid in determining the colonization status of antepartum women, but are not intended to diagnose or monitor treatment of a GBS infection.
The Simplexa™ GBS Direct assay does not provide susceptibility results. Culture isolates are needed to perform susceptibility testing as recommended for penicillin-allergic women.
Simplexa™ GBS Positive Control Pack
The Simplexa™ GBS Positive Control Pack is intended to be used as a control with the Simplexa™ GBS Direct kit. This control is not intended for use with other assays or systems.
The Simplexa™ GBS Direct assay system is a real-time PCR system that enables the direct amplification and qualitative detection of Group B Strep bacterial DNA from vaginal swabs enriched in Lim Broth for eighteen to twenty-four (18 to 24) hours that have not undergone a nucleic acid extraction. The system consists of the Simplexa™ GBS Direct assay, the LIAISON® MDX (with LIAISON® MDX Studio Software), the Direct Amplification Disc (DAD) and associated accessories.
In the Simplexa™ GBS Direct assay, primers and fluorescent probes are used together to amplify Group B Streptococcus bacterial DNA and the Internal Control (DNA IC). The assay targets a conserved region of the cfb gene to identify Group B Streptococus in the specimen. The DNA IC is used to detect PCR failure and/or inhibition.
The provided document is a 510(k) Summary for a medical device called Simplexa™ GBS Direct. This document primarily focuses on the device's analytical and clinical performance for the qualitative detection of Group B Streptococcus (GBS) nucleic acid in Lim broth enrichments. It is not an AI/ML medical device in the typical sense that would involve human interpretation of images or other data assisted by an AI, but rather a PCR-based diagnostic assay.
Therefore, many of the requested items related to AI/ML device studies (e.g., number of experts, adjudication methods, MRMC studies, human reader improvement) are not applicable to this type of device. The study described focuses on the device's accuracy against a culture reference method and its analytical performance characteristics.
Here’s a breakdown of the information that can be extracted and which requested items are not applicable:
Acceptance Criteria and Study for Simplexa™ GBS Direct
The "acceptance criteria" for this device are implicitly tied to a demonstration of substantial equivalence to a predicate device, which includes showing acceptable clinical and analytical performance. The study described is the clinical performance evaluation.
1. A table of acceptance criteria and the reported device performance
The document does not explicitly state pre-defined "acceptance criteria" in terms of specific sensitivity and specificity thresholds that were set before the study. Instead, it presents the clinical performance results. The implicit acceptance is based on these results being deemed sufficient for substantial equivalence.
| Performance Metric | Acceptance Criteria (Implicit) | Reported Device Performance (Simplexa™ GBS Direct vs. Culture) |
|---|---|---|
| Clinical Performance: | ||
| Sensitivity | N/A (demonstrate acceptable clinical agreement) | 97.0% (97/100) with 95% CI: 91.5% to 99.0% |
| Specificity | N/A (demonstrate acceptable clinical agreement) | 96.1% (319/332) with 95% CI: 93.4% to 97.7% |
| Reproducibility (Overall Agreement with Expected Results): | ||
| GBS (FAM) Signal | N/A (demonstrate high agreement) | 99.6% (538/540) with 95% CI: 98.7% to 99.9% |
| Internal Control (Q670) Signal | N/A (demonstrate high agreement) | 100.0% (540/540) with 95% CI: 99.3% to 100.0% |
| Analytical Sensitivity (LoD): | N/A (establish lowest detectable concentration) | ATCC BAA-22 (serotype III): 80,000 CFU/mL |
| ATCC BAA-1138 (serotype Ia): 30,000 CFU/mL | ||
| Analytical Reactivity: | N/A (demonstrate detection of various GBS strains) | 100% detection for 18 additional GBS strains tested at 2 x LoD |
| Cross Reactivity (Analytical Specificity): | N/A (demonstrate no detection of non-GBS organisms) | 0% detection for 74 potential cross-reacting organisms |
| Interference: | N/A (demonstrate no interference from common substances) | No interference observed with 26 potentially interfering substances |
| Inhibition by other Microorganisms: | N/A (demonstrate reliable detection in presence of other organisms) | 100% detection for GBS BAA-22, |
| Some false negatives (88.9% or 77.8% detected) for GBS BAA-1138 with 7 specific organisms (e.g., Bacillus cereus, Candida parapsilosis, Clostridium perfringens) |
2. Sample size used for the test set and the data provenance
- Test Set Sample Size: 432 clinical samples were used for the prospective clinical performance study.
- Data Provenance:
- Country of Origin: Not explicitly stated, but the applicant (DiaSorin Molecular LLC) is based in Cypress, California, USA, suggesting the data is likely from the United States.
- Retrospective or Prospective: The clinical study was prospective. Samples were collected from pregnant women at 35-37 weeks of gestation, and Lim broth enrichments were made and tested fresh.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)
This is not applicable as this is not an AI/ML device involving human interpretation of data where expert consensus is required for ground truth. The ground truth for the clinical performance study was established by a "GBS culture reference method." This refers to standard microbiological laboratory procedures for culturing and identifying GBS, not to human expert interpretation of an AI's output.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
This is not applicable for the same reason as above. Discrepancies between the device and the culture reference method (ground truth) were investigated by testing with an "alternate FDA cleared NAAT" (Nucleic Acid Amplification Test). For example:
- 3/3 samples where Simplexa™ GBS Direct was negative and Culture was positive, were found negative by alternate NAAT.
- 11/13 samples where Simplexa™ GBS Direct was positive and Culture was negative, were found positive by alternate NAAT.
This indicates a form of discrepant analysis, not human reader adjudication.
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 not applicable as this is not an AI/ML device where human readers interact with or are assisted by AI. The device is a standalone diagnostic assay.
6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done
Yes, the primary clinical performance study (Table 1) represents the standalone performance of the Simplexa™ GBS Direct assay against a culture reference method. The results provided (Sensitivity and Specificity) represent the algorithm's diagnostic accuracy.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
The ground truth for the clinical performance study was established by a "GBS culture reference method." This is a laboratory-based microbiological culture and identification process, considered the gold standard for GBS detection in this context.
8. The sample size for the training set
The document describes a 510(k) submission for a diagnostic assay, not an AI/ML model that undergoes a "training" phase with a large dataset. Therefore, the concept of a "training set" for an AI model is not directly applicable here. The device's PCR primers and probes are designed based on biological knowledge (targeting the cfb gene) rather than being learned from data. Extensive analytical studies (LoD, reactivity, cross-reactivity, interference, inhibition) are done during development and validation, but this isn't analogous to training an AI model.
9. How the ground truth for the training set was established
As the concept of a "training set" in the context of AI/ML is not applicable to this diagnostic assay, this question is also not applicable. The underlying assays are developed based on known biological targets and validated through a series of analytical and clinical studies.
§ 866.3740
Streptococcus spp. serological reagents.(a)
Identification. Streptococcus spp. serological reagents are devices that consist of antigens and antisera (excluding streptococcal exoenzyme reagents made from enzymes secreted by streptococci) used in serological tests to identifyStreptococcus spp. from cultured isolates derived from clinical specimens. The identification aids in the diagnosis of diseases caused by bacteria belonging to the genusStreptococcus and provides epidemiological information on these diseases. Pathogenic streptococci are associated with infections, such as sore throat, impetigo (an infection characterized by small pustules on the skin), urinary tract infections, rheumatic fever, and kidney 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.