(200 days)
The Simplexa C. auris Direct is a real-time polymerase chain reaction (RT-PCR) assay intended for use on the LIAISON MDX instrument for the direct in vitro qualitative detection of Candida auris DNA from a composite swab of bilateral axilla/groin from patients suspected of C. auris colonization.
The test is intended to aid in the prevention and control of C. auris infection in healthcare settings by detecting C. auris from colonized patients.
Positive results indicate that the patient is colonized with C. auris. A positive result cannot rule out co-colonization with other pathogens. A negative result does not preclude C. auris colonization or infection and should not be used as the sole basis for treatment or other patient management decisions. Results are meant to be used in conjunction with other clinical, epidemiologic, and laboratory information available to the clinician evaluating the patient. The test is not intended to diagnose or monitor treatment for C. auris infection. Concomitant cultures are necessary to recover organisms for epidemiological typing or for antimicrobial susceptibility testing.
The Simplexa C. auris RT-PCR system is intended for the amplification and qualitative detection of nucleic acid from Candida auris in composite bilateral axilla/groin swab specimens and consists of the following:
- The Simplexa C. auris Direct is the RT-PCR assay kit that contains all the reagents for the amplification reaction, including the primers and fluorescent probes for the detection of nucleic acid from Candida auris. The primers and fluorescent probes amplifies the C. auris DNA and Internal Control DNA. In addition, the kit comes with a barcode card, which contains assay specific parameters and lot information.
- The Simplexa C. auris Positive Control Pack is the separately packaged external positive quality control kit for use with the Simplexa C. auris Direct assay.
- The Simplexa C. auris Sample Prep Kit is the enzymatic buffer solution to receive the sample solution (bilateral axilla/groin swab in Amies transport media) from the patient.
The Simplexa C. auris RT-PCR system is for use with the LIAISON MDX instrument (with LIAISON MDX Studio Software), the RT-PCR thermocycler that amplifies the nucleic acid from biological specimens and uses real-time fluorescence detection to identify targets, and the Direct Amplification Disc (DAD), which is the accessory containing the input sample wells for use on the LIAISON MDX. The instrument and accessory were previously cleared under K102314 and K120413. The instrument is controlled by an external laptop running the software. The DAD consumable is compartmentalized into eight (8) separate wedges and can process up to eight (8) separate specimens or controls on each disc. Each wedge contains sample and reagent input wells, microfluidic channels and laser activated valves to control the fluid flow as well as a reaction/detection chamber.
The provided document describes the analytical and clinical performance of the Simplexa C. auris Direct RT-PCR system. However, it does not explicitly state pre-defined acceptance criteria in a table format that the device needed to meet. Instead, it presents the results of various studies (precision, analytical specificity, limit of detection, inclusivity, clinical performance) and then often concludes whether these results are "acceptable."
For example, for precision, it states: "For the multisite study, the test device showed ≥ 98.9% agreement of the qualitative result and ≤ 8.2% CV for each of the variance components, which is acceptable." This implies that ≥ 98.9% agreement and ≤ 8.2% CV were the internal acceptance criteria for precision.
Similarly, for the clinical performance, the reported Positive Percent Agreement (PPA) and Negative Percent Agreement (NPA) values are presented as the device's performance, but explicit pre-defined minimum thresholds for PPA and NPA as acceptance criteria are not given. They are implied by the fact that the De Novo request was granted.
Given this, I will infer the acceptance criteria from the reported "acceptable" results where possible and present the device performance based on the clinical study results.
Here's the information requested based on the provided text:
1. Table of Acceptance Criteria (Inferred) and Reported Device Performance
| Performance Characteristic | Inferred Acceptance Criteria (Based on "Acceptable" Results) | Reported Device Performance |
|---|---|---|
| Analytical Performance | ||
| Multisite Precision (% Agreement) | ≥ 98.9% (qualitative result) | 98.9% (Clade I South Asian (LP)) and 100.0% (other variants and controls) |
| Multisite Precision (% CV) | ≤ 8.2% (for each variance component) | ≤ 8.2% (observed max) |
| Lot-to-Lot Precision (% Agreement) | 100% (expected results) | 100% |
| Lot-to-Lot Precision (% CV) | ≤ 7.2% (for all panel members) | ≤ 7.2% (observed max combined) |
| Cross-reactivity & Microbial Interference | No observed cross-reactivity or interference | Not observed with any of the 34 organisms tested (wet testing) and none predicted by in silico analysis for 13 organisms |
| Interfering Substances | Expected detection rates, with documented interferences where applicable | Some interferences noted (anti-breathable deodorant cream @ 10% v/v, Benzalkonium chloride @ 0.07% v/v resulted in invalid results; detection restored at lower concentrations). Other 34 substances showed 100% detection. |
| Specimen Stability (5x LoD) | 100% positivity | 100% |
| Specimen Stability (2x LoD) | ≥ 95% positivity | 93-100% (Clade I 2xLoD fresh result was 93%, others were ≥97%) |
| Specimen Stability (0.5x LoD) | 10-90% positivity (expected variability) | 20-100% (varies by condition and clade) |
| Specimen Stability (Negative Samples) | 0% positivity | 0% |
| Limit of Detection (LoD) | ≥ 95% detection rate for the lowest concentration (confirmatory LoD) | Clade I: 127 CFU/mL (98% detection) Clade IV: 260 CFU/mL (98% detection) |
| Inclusivity (% Detection) | 100% detection of tested strains (wet testing) | 100% (all 9 strains from 6 clades) |
| Inclusivity (% Homology predicted) | ≥ 90% (oligo identity) with full coverage and predicted inclusivity | 98% (721/736 sequences) with two new Clade VI sequences showing 100% homology. |
| Carry-Over/Cross Contamination (% Detection of Negatives) | 0% | 0% (56/56 negative samples) |
| Clinical Performance | ||
| Positive Percent Agreement (PPA) - Prospective Cohort | Implied "acceptable" given clearance | 94.1% (32/34) (95% CI: 80.9% - 98.4%) |
| Negative Percent Agreement (NPA) - Prospective Cohort | Implied "acceptable" given clearance | 98.8% (1874/1896) (95% CI: 98.2% - 99.2%) |
| PPA - Combined Cohort (Prospective, Enriched, Retrospective) | Implied "acceptable" given clearance | 94.8% (55/58) (95% CI: 85.9% - 98.2%) |
| NPA - Combined Cohort (Prospective, Enriched, Retrospective) | Implied "acceptable" given clearance | 98.7% (1937/1962) (95% CI: 98.1% - 99.1%) |
2. Sample Size Used for the Test Set and Data Provenance
Test Set (Clinical Study):
- Total Evaluable Clinical Specimens (Combined Cohort): 2,020 specimens.
- Prospective Cohort: 1,930 evaluable specimens.
- Retrospective/Enriched Cohort: 90 evaluable specimens (11 retrospective pre-selected C. auris positive + 202 enriched specimens, of which 90 were evaluable).
- Data Provenance:
- Geographic Locations: Six study sites across four geographically diverse locations within the United States and one in Italy.
- Type of Data:
- Prospective: Prospectively collected specimens (axilla/groin swabs) from patients suspected of C. auris colonization. Tested either fresh or frozen. Collected from April to July 2023.
- Retrospective/Enriched: Leftover, de-identified composite bilateral axilla/groin swabs. Included 11 pre-selected C. auris positive retrospective specimens and 202 enriched specimens (identified as positive by a laboratory-verified RT-PCR test, then blinded and tested).
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document states that the ground truth for the clinical study was established by a reference method consisting of "standard of care (SOC) culture followed by matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) for identification." These are laboratory methods, not human expert interpretation of images or other subjective data.
For discordant analysis, "Bi-directional sequencing (BDS) assays were performed when the candidate assay results differed from the comparator method." Again, this is a laboratory method.
The document does not specify human experts or their qualifications for establishing the ground truth for the test set.
4. Adjudication Method for the Test Set
The primary ground truth was established by the reference method (SOC culture + MALDI-TOF MS). For discordant results between the candidate assay and the reference method, Bi-directional Sequencing (BDS) was performed. However, "The results from discordant analysis were not used to alter the original performance but are provided in footnotes to the performance tables."
Therefore, there wasn't a human-based adjudication method in the traditional sense (e.g., 2+1 or 3+1 radiologists making a consensus decision). The reference method and supplemental BDS purely relied on objective laboratory techniques.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was Done
No, this document describes a diagnostic device (RT-PCR assay) for detecting nucleic acids of Candida auris. It is not an AI-assisted diagnostic imaging device for which an MRMC study comparing human readers with and without AI assistance would be relevant. The study focuses solely on the direct performance of the molecular diagnostic test against a laboratory reference method.
6. If a Standalone (i.e. Algorithm only without human-in-the-loop performance) was Done
Yes, the entire clinical evaluation (Sections VI.C) and analytical performance studies (Sections VI.A) describe the standalone performance of the Simplexa C. auris Direct RT-PCR system. This device is an automated molecular diagnostic test and does not involve human interpretation in a loop, except for the user performing the test steps according to the instructions. The results (Ct values, positive/negative calls) are generated directly by the instrument platform (LIAISON MDX).
7. The Type of Ground Truth Used
The primary ground truth used for the clinical study was Standard of Care (SOC) culture followed by Matrix-Assisted Laser Desorption/Ionization-Time of Flight Mass Spectrometry (MALDI-TOF MS) for identification. For discordant results, Bi-directional Sequencing (BDS) was used as a supplemental ground truth, though these results did not alter the original performance metrics. This is a form of laboratory reference standard.
8. The Sample Size for the Training Set
The document does not explicitly mention a "training set" in the context of an AI/ML algorithm that would undergo a distinct training phase. This document describes a molecular diagnostic assay, not a machine learning model. The development of such assays involves establishing parameters (like fluorescence and Ct thresholds) using initial data, which could be considered an internal "calibration" or "development" process rather than a "training set" in the AI sense.
It states: "The fluorescence and Ct thresholds for C. auris and Internal Control were established using 717 sample runs of No Template Control (NTC), Limit of Detection (LoD), Microbial Inhibition, Interference and Limiting Dilution samples. The established thresholds were then confirmed using an independent data set comprising 2,924 sample runs..." These 717 runs could be considered the data used for establishing (or "training") the assay's interpretive parameters.
9. How the Ground Truth for the Training Set Was Established
Given that this is a molecular diagnostic assay and not an AI/ML system, the concept of "ground truth for a training set" as it pertains to labeled examples for model learning is not directly applicable.
Instead, the "establishment" of the assay's operating parameters (like Ct thresholds) was based on experimental data where the expected outcome (presence/absence of C. auris, inhibition, etc.) was known by design:
- No Template Control (NTC): Expected negative.
- Limit of Detection (LoD): Contrived samples with known concentrations of C. auris.
- Microbial Inhibition/Interference: Samples with C. auris at known concentrations (e.g., 3x LoD) spiked with other substances/organisms.
- Limiting Dilution: Samples serially diluted to determine the lowest detectable concentration.
These experiments provide the "ground truth" for setting the assay's interpretive parameters and analytical performance characteristics. The known concentrations, presence/absence of target nucleic acids, and presence/absence of interfering substances served as the factual basis for defining how the device should interpret its signals (e.g., Ct value thresholds for positivity).
N/A