(258 days)
The T2Bacteria Panel run on the T2Dx Instrument is a qualitative T2 magnetic resonance (T2MR) test for the direct detection of bacterial species in K2EDTA human whole blood specimens with suspected bacteremia. The T2Bacteria Panel identifies five species of bacteria: Enterococcus faecium, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Staphylococcus aureus.
The T2Bacteria Panel is indicated as an aid in the diagnosis of bacteremia and results should be used in conjunction with other clinical and laboratory data. Concomitant blood cultures are necessary to recover organisms for susceptibility testing or further identification, and for organisms not detected by the T2Bacteria Panel.
Results from the T2Bacteria Panel are not intended to be used as the sole basis for diagnosis, treatment management decisions in patients with suspected bacteremia.
The T2Bacteria Panel detects and identifies five bacterial target species directly from whole blood specimens and independent of blood culture using nucleic acid amplification and proprietary T2MR® detection technology. The assay is performed on the proprietary T2Dx platform.
The whole blood specimen, drawn into a blood collection tube containing K₂EDTA is used for the test. The blood collection tube containing a minimum of 3 mL of blood is loaded directly onto the T2Dx instrument as part of the assembled Cartridge, a single use self-contained unit that contains all of the reagents and disposables required to run a single test.
Fully automated on the T2Dx, the blood specimen is mixed with the Lysis Reagent to lyse the red blood cells and the bacterial cells are concentrated by centrifugation. The Internal Control is added to the concentrated bacterial cells, which then undergo a bead beating step to lyse the bacteria cells. The supernatant containing the DNA from the lysed bacterial cells and the Internal Control are amplified with the target and Internal Control specific primers. The generated amplicon is then aliquoted into individual tubes containing targetspecific conjugated particles for Enterococcus faecium, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Staphylococcus aureus, and the Internal Control. These individual tubes are read in the MR reader and a signal is generated.
Up to seven specimens can be loaded onto the T2Dx Instrument in parallel. When running a single specimen, the first result is reported in 3.5 hours from the time the specimen is loaded onto the instrument. The results are interpreted using the T2Dx applications software as valid or invalid, and if valid, target specific results are reported as Positive or Target not Detected. For one target, the Escherichia coli channel, results are reported as Positive, Indeterminate, or Target not Detected. An Indeterminate result is a valid result, but the presence or absence of Escherichia coli cannot be definitively assessed, and the indeterminate status applies only to the Escherichia coli channel. Results are displayed on the T2Dx touchscreen and can be printed. Raw T2MR data are not available to the end user.
Here's a breakdown of the acceptance criteria and study information for the T2Bacteria Panel, based on the provided text:
1. Acceptance Criteria and Reported Device Performance
The document does not explicitly state "acceptance criteria" for the clinical study in a dedicated table format. However, it presents the performance metrics (PPA and NPA) for each target species. We can infer the implicit acceptance criteria from these reported values, typically striving for high sensitivity (PPA) and specificity (NPA). For the purpose of this response, I will list the reported performance from the combined prospective and contrived arms as the "reported device performance."
| Target Species | Metric | Reported Device Performance (Value) | Reported Device Performance (95% CI) |
|---|---|---|---|
| E. faecium | PPA | 100% (40/40) (Contrived) / 100% (1/1) (Prospective) | 91.2 - 100% (Contrived) / 20.7 - 100% (Prospective) |
| NPA | 100% (300/300) (Contrived) / 99.4% (1417/1426) (Prospective) | 98.7 - 100% (Contrived) / 98.8 - 99.7% (Prospective) | |
| S. aureus | PPA | 92.3% (36/39) (Contrived) / 81.3% (13/16) (Prospective) | 79.7 - 97.3% (Contrived) / 57.0 - 93.4% (Prospective) |
| NPA | 100% (300/300) (Contrived) / 98.0% (1383/1411) (Prospective) | 98.7 - 100% (Contrived) / 97.1 - 98.6% (Prospective) | |
| K. pneumoniae | PPA | 100% (40/40) (Contrived) / 100% (6/6) (Prospective) | 91.2 - 100% (Contrived) / 61.0 - 100% (Prospective) |
| NPA | 99.3% (298/300) (Contrived) / 98.5% (1399/1421) (Prospective) | 97.6 - 99.8% (Contrived) / 97.7 - 99.0% (Prospective) | |
| P. aeruginosa | PPA | 97.4% (38/39) (Contrived) / 100% (5/5) (Prospective) | 86.8 - 99.5% (Contrived) / 56.6 - 100% (Prospective) |
| NPA | 97.7% (293/300) (Contrived) / 97.7% (1389/1422) (Prospective) | 95.3 - 98.9% (Contrived) / 96.8 - 98.3% (Prospective) | |
| E. coli | PPA | 90.9% (20/22) (Contrived) / 90.9% (10/11) (Prospective) | 72.2 - 97.5% (Contrived) / 62.3 - 98.4% (Prospective) |
| NPA | 97.3% (292/300) (Contrived) / 95.0% (1345/1416) (Prospective) | 94.8 - 98.6% (Contrived) / 93.7 - 96.0% (Prospective) |
Note: PPA (Sensitivity) was calculated against samples with titer levels at or above the limit of detection (LoD) in the Contrived Arm and blood culture positives in the Prospective Arm. NPA (Specificity) was calculated from all samples (including below LoD and unspiked negative samples) as the total number of negative channels divided by total number of non-spiked channels in the Contrived Arm and blood culture negatives in the Prospective Arm.
2. Sample Size Used for the Test Set and Data Provenance
- Prospective Arm: 1,427 subjects tested. The study was conducted at eleven sites within the US. The data is prospective.
- Contrived Arm: 250 contrived specimens (50 strains of each of the five bacterial species) and an additional 100 blood samples not spiked with T2Bacteria Panel members. These were evaluated at three sites. The data is contrived (laboratory-prepared).
3. Number of Experts Used to Establish the Ground Truth for the Test Set and the Qualifications of Those Experts
The document does not explicitly state the number of experts or their qualifications used to establish the ground truth for the test set.
For the Prospective Arm, the ground truth was "the reference method of blood culture," with species identification performed on all positive bacteria cultures using methods like Gram stain, bioMerieux Vitek® 2, bioMerieux or Bruker MALDI TOF, and PCR. This implies standard laboratory professionals and potentially clinicians reviewed and confirmed results, but specific details on expert involvement for ground truth establishment are not provided.
For the Contrived Arm, the ground truth was established by spiking known concentrations of bacterial species into healthy donor whole blood. This is a controlled, laboratory-defined ground truth, not reliant on expert interpretation for "truth" but rather on the setup of the experiment.
4. Adjudication Method for the Test Set
The document does not explicitly state an adjudication method (such as 2+1, 3+1, etc.) for establishing ground truth from the blood cultures in the clinical study. It refers to "species identification was performed on all positive bacteria cultures and methods included Gram stain, bioMerieux Vitek® 2, bioMerieux or Bruker MALDI TOF, and PCR." Discordant analysis for T2(+)/BC(-) cases was performed, identifying "strong evidence of infection" based on other blood cultures or sequencing, or "other evidence of infection" from non-blood cultures, but this is a reconciliation process rather than an initial ground truth 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
Not applicable. This device is a diagnostic test (T2Bacteria Panel) that directly detects bacterial species using T2 magnetic resonance technology. It is not an AI-assisted diagnostic tool for human readers; it provides a direct result. Therefore, no MRMC comparative effectiveness study involving human readers improving with or without AI assistance was performed.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
Yes, this was a standalone performance study. The T2Bacteria Panel is an automated diagnostic instrument. "Results are interpreted using the T2Dx applications software as valid or invalid, and if valid, target specific results are reported as Positive or Target not Detected." The clinical performance evaluation directly compares the device's results to blood culture results. There isn't a human-in-the-loop component for interpreting the T2Bacteria Panel's output described in the document.
7. The Type of Ground Truth Used (Expert Consensus, Pathology, Outcomes Data, Etc.)
The primary ground truth used was blood culture results with species identification. For the contrived arm, the ground truth was the known spiked bacterial species and concentrations.
For the discordant analysis of T2(+)/BC(-) results, additional forms of evidence were used to assess the likelihood of a true positive, which included:
- Other blood cultures positive for the same organism within ±14 days.
- Sequencing of concurrently drawn blood samples.
- Other non-blood matrices cultured positive for the same organism within ±14 days.
8. The Sample Size for the Training Set
The document does not explicitly mention a separate "training set" for the T2Bacteria Panel. This is typically a characteristic of machine learning models. For a molecular diagnostic assay like this, development usually involves analytical studies to optimize reagents and parameters (LoD, specificity, etc.), rather than a distinct "training set" in the machine learning sense. The performance data presented (LoD, Reproducibility, Inclusivity, Exclusivity, Competitive Inhibition, Interfering Substances) represent analytical validation studies.
9. How the Ground Truth for the Training Set Was Established
As no explicit "training set" is mentioned in the machine learning context, this question is not applicable. The ground truth for analytical validation studies (which could be considered analogous to development/optimization) was established through controlled laboratory experiments, such as spiking known organisms at specific concentrations into whole blood.
§ 866.3960 Nucleic acid-based device for the amplification, detection, and identification of microbial pathogens directly from whole blood specimens.
(a)
Identification. A nucleic acid-based device for the amplification, detection, and identification of microbial pathogens directly from whole blood specimens is a qualitative in vitro device intended for the amplification, detection, and identification of microbial-associated nucleic acid sequences from patients with suspected bloodstream infections. This device is intended to aid in the diagnosis of bloodstream infection when used in conjunction with clinical signs and symptoms and other laboratory findings.(b)
Classification. Class II (special controls). The special controls for this device are:(1) Premarket notification submissions must include detailed device description documentation, including the device components, ancillary reagents required but not provided, and a detailed explanation of the methodology, including primer/probe sequence, design, and rationale for sequence selection.
(2) Premarket notification submissions must include detailed documentation from the following analytical and clinical performance studies: Analytical sensitivity (limit of detection), reactivity, inclusivity, precision, reproducibility, interference, cross reactivity, carryover, and cross contamination.
(3) Premarket notification submissions must include detailed documentation from a clinical study. The study, performed on a study population consistent with the intended use population, must compare the device performance to results obtained from well-accepted reference methods.
(4) Premarket notification submissions must include detailed documentation for device software, including, but not limited to, software applications and hardware-based devices that incorporate software.
(5) The device labeling must include limitations regarding the need for culture confirmation of negative specimens, as appropriate.
(6) A detailed explanation of the interpretation of results and acceptance criteria must be included in the device's 21 CFR 809.10(b)(9) compliant labeling.
(7) Premarket notification submissions must include details on an end user device training program that will be offered while marketing the device, as appropriate.
(8) As part of the risk management activities performed as part of your 21 CFR 820.30 design controls, you must document an appropriate end user device training program that will be offered as part of your efforts to mitigate the risk of failure to correctly operate the instrument.