(85 days)
The BACT/ALERT® MP Reagent System consists of the BACT/ALERT® MP culture bottle with a removable closure used in conjunction with the BACT/ALERT® MP Antimicrobial Supplement and the BACT/ALERT® MP Nutrient Supplement. The BACT/ALERT® MP Reagent System is designed for use with BACT/ALERT® 3D Mycobacteria Detection Systems for recovery and detection of mycobacteria from sterile body specimens other than blood, and from digested/decontaminated clinical specimens.
BACT/ALERT® MP Reagent System consists of three reagents: BACT/ALERT® MP Culture Bottle, the BACT/ALERT® MP Antimicrobial Supplement and the BACT/ALERT® MP Nutrient Supplement.
The provided text describes the performance characteristics of the BACT/ALERT® MP Reagent System, comparing it to a predicate device (BACT/ALERT® MP Culture Bottle) for the recovery and detection of mycobacteria.
Here's a breakdown of the requested information:
1. Table of Acceptance Criteria and Reported Device Performance:
The document doesn't explicitly state "acceptance criteria" in a table format with pass/fail thresholds. However, it presents the performance of the new device and the predicate device, implying that the new device aims to be equivalent or superior, particularly in Time To Detection. The study demonstrates that the new device has increased sensitivity and specificity and a faster time to detection for certain mycobacteria species.
For the purpose of this response, I will interpret the performance metrics presented as the "reported device performance" and infer that the device meets implied acceptance criteria by being equivalent or better than the predicate.
| Performance Metric | BACT/ALERT® MP Reagent System (REF 419744) (Test Device) | BACT/ALERT® MP Culture Bottle (REF 259797) (Predicate Device) | Implied Acceptance Criterion (e.g., must be equivalent or better than predicate) |
|---|---|---|---|
| Analytical Sensitivity: LoD | |||
| M. avium | ≤ 15 CFU/bottle (100% detection) | Not explicitly stated for predicate in LoD section | Low LoD (e.g., ≤ the predicate device's LoD) |
| M. kansasii | ≤ 4 CFU/bottle (100% detection) | Not explicitly stated for predicate in LoD section | Low LoD (e.g., ≤ the predicate device's LoD) |
| M. tuberculosis | ≤ 14 CFU/bottle (100% detection) | Not explicitly stated for predicate in LoD section | Low LoD (e.g., ≤ the predicate device's LoD) |
| Analytical Sensitivity: Growth Performance | 100% detection at high (10^6^ CFU/bottle) and low (10^2^ CFU/bottle) inoculum levels | Not explicitly stated for predicate | 100% detection for tested organisms at varying inoculum concentrations |
| Within-Laboratory Precision - Recovery Rate | 100% recovery for all test events and concentrations | Not explicitly stated for predicate | 100% recovery rate |
| Reproducibility - Overall Recovery Rate (Pleural Fluid) | 99.6% | Not explicitly stated for predicate | High recovery rate (e.g., ≥ 95%) |
| Reproducibility - Overall Recovery Rate (Simulated Sputum) | 99.9% | Not explicitly stated for predicate | High recovery rate (e.g., ≥ 95%) |
| Clinical Performance - Sensitivity | 86.58% (95% CI: 81.50%, 90.70%) | 81.39% (95% CI: 75.76%, 86.19%) | Equivalent or superior to predicate; the test device showed a 5.19% increase in sensitivity. |
| Clinical Performance - Specificity | 96.75% (95% CI: 95.49%, 97.74%) | 93.79% (95% CI: 92.16%, 95.18%) | Equivalent or superior to predicate; the test device showed a 2.96% increase in specificity. |
| Clinical Performance - Overall Recovery Rate (Positive Mycobacteria Samples) | 94.34% | 88.68% | Equivalent or superior to predicate; the test device showed a 5.66% increase. |
| False Negative Rate (Clinical Samples) | 0.13% (2/1,488) | Not explicitly stated for predicate | Low false negative rate (e.g., ≤ predicate's or a clinically acceptable threshold). |
| False Positive Rate (Clinical Samples) | 2.28% (34/1,488) | Not explicitly stated for predicate | Low false positive rate (e.g., ≤ predicate's or a clinically acceptable threshold). |
| Time to Detection (TTD) - M. tuberculosis complex species | Mean: 11.6 days | Mean: 14.4 days | Faster than predicate (demonstrated -2.7 days mean difference). |
| Time to Detection (TTD) - Non-tuberculous Mycobacteria Rapid Growers | Mean: 7.3 days | Mean: 7.4 days | Faster than or equivalent to predicate (demonstrated -0.1 days mean difference, considered faster). |
| Time to Detection (TTD) - Non-tuberculous Mycobacteria Slow Growers | Mean: 11.5 days | Mean: 11.8 days | Faster than or equivalent to predicate (demonstrated -0.3 days mean difference, considered faster). |
2. Sample size used for the test set and data provenance:
-
Clinical Performance Studies (for Sensitivity, Specificity, Recovery Rates):
- Sample Size: 1,488 specimens obtained from 1,162 patients.
- Data Provenance: Prospective clinical study conducted at 3 different sites. The document does not specify the country of origin, but generally, FDA submissions involve studies conducted in or for the US market, potentially with multi-national sites.
-
Analytical Performance Studies (LoD, Growth Performance, Precision, Reproducibility):
- Sample Size: Varies by study (e.g., LoD specified as lowest inoculum level with >95% recovery, Growth Performance at high and low inoculum levels, Precision on a minimum of 3 lots over 12 days, Reproducibility with 5 organisms at low and high concentrations in two matrices). Specific total sample numbers for these analytical studies are not given, but generally involve laboratory-controlled experiments.
- Data Provenance: In-house seeded studies and tests at clinical trial sites (for reproducibility). The specific country of origin is not mentioned for these internal studies.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
This information is not provided in the document. The ground truth for clinical samples appears to be established by laboratory results using both the test and predicate devices, and primary solid medium culture ("a sample has a positive patient infected status when one of the BACT/ALERT® MP (REF 419744 or 259797) subcultures is positive, or the primary solid medium is positive"). This implies a laboratory-based gold standard rather than expert clinical consensus or adjudication on imaging, for example.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set:
The document does not describe an adjudication method involving multiple human readers or a specific consensus process for reviewing diagnostic results in the test set. The ground truth seems to be determined by a combination of positive results from subcultures of both the test and predicate devices, and primary solid medium culture. This is a laboratory-based ground truth, not one requiring human expert 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 device is a microbial growth monitor and not an AI-powered diagnostic imaging tool or a system that requires human "readers" in the context of an MRMC study. Therefore, an MRMC comparative effectiveness study, as typically understood for AI-assisted image interpretation, was not performed and is not applicable here. The comparison is between two laboratory systems (the new device vs. the predicate).
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
The device (BACT/ALERT® MP Reagent System) is an automated system for detecting microbial growth. Its performance is inherently standalone in the sense that the instrument provides the detection signal without continuous human intervention for interpretation in the same way an AI algorithm for image analysis would. The "algorithm" here is the system's ability to detect CO2 production. The study evaluates the performance of this system directly.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
The ground truth for the clinical performance studies (sensitivity, specificity, recovery) in the document is defined as follows:
- Positive Patient Infected Status: "A sample has a positive patient infected status when one of the BACT/ALERT® MP (REF 419744 or 259797) subcultures is positive, or the primary solid medium is positive."
- Negative Patient Infected Status: "A sample has a negative patient infected status when both BACT/ALERT® MP (REF 419744 and 259797) subcultures are negative and the primary solid media is negative (i.e. there is no growth of mycobacteria)."
This is a laboratory-based and culture-confirmed ground truth, relying on the isolation and identification of mycobacteria from subcultures and primary solid media.
8. The sample size for the training set:
The document does not mention a training set because this is a medical device for in-vitro diagnostics based on chemical reactions and optical detection, not typically a machine learning or AI algorithm that requires a separate training set. The "formulation changes" and "analytical performance studies" represent the development and validation of the reagent system itself.
9. How the ground truth for the training set was established:
As no training set is mentioned or applicable in the context of this device's type, there is no information on how ground truth for a training set was established.
§ 866.2560 Microbial growth monitor.
(a)
Identification. A microbial growth monitor is a device intended for medical purposes that measures the concentration of bacteria suspended in a liquid medium by measuring changes in light scattering properties, optical density, electrical impedance, or by making direct bacterial counts. The device aids in the diagnosis of disease caused by pathogenic microorganisms.(b)
Classification. Class I. With the exception of automated blood culturing system devices that are used in testing for bacteria, fungi, and other microorganisms in blood and other normally sterile body fluids, this device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter.