(203 days)
The NeuMoDx™ GBS Assay as implemented on the NeuMoDx™ 288 Molecular System is a qualitative in vitro diagnostic test designed to detect Group B Streptococcus (GBS) DNA from 18-24 hour Lim broth enrichments of vaginal/rectal swabs from pregnant women. The test incorporates automated DNA extraction to isolate the target nucleic acid from the specimen and real-time polymerase chain reaction (PCR) to detect an 88 bp region of the pcsB gene sequence in the Streptococcus agalactiae chromosome. Results from the NeuMoDx™ GBS Assay can be used as an aid in determining colonization status in antepartum women.
The NeuMoDx™ GBS Assay does not provide susceptibility results. Cultured isolates are needed for performing susceptibility testing as recommended for penicillin-allergic women.
The NeuMoDx™ GBS Assay (Subject Device) as implemented on the NeuMoDx™ 288 Molecular System is an automated, qualitative, in vitro diagnostic test for the detection of group B Streptococcus (GBS), also known as Streptococcus agalactiae from vaginal/rectal swabs collected from pregnant women at 35 - 37 weeks of gestation and enriched in a commercially available Lim broth medium. An aliquot of an overnight Lim broth culture added to the NeuMoDx™ GBS Assay is used for the testing. All further specimen handling is automated.
The GBS Assay test strip, in combination with required NeuMoDx buffers, extraction reagents, wash and release solutions, as well as the microfluidic cartridge (non-active,) and the fully automated NeuMoDx™ 288 Molecular System (a real time nucleic acid amplification system), utilizes real-time polymerase chain reaction (PCR) for the amplification of GBS DNA and fluorogenic target-specific TaqMan® probes for the detection of the amplified GBS DNA. General use components and the System are packaged and provided separately by NeuMoDx.
After the test is processed, a determination of the presence/absence of GBS DNA in the specimen is automatically made based on the amplification status of GBS and the Sample Process Control using preestablished decision criteria. The test results will be reported as Negative, Positive, Indeterminate or Unresolved based on the amplification status of the target and sample processing control. Results are reported based on the decision algorithm noted in Table 1.
The provided text describes the NeuMoDx™ GBS Assay, an in vitro diagnostic test for Group B Streptococcus (GBS), and its performance data to establish substantial equivalence to a predicate device. This is a molecular diagnostic test, not an AI/imaging device, so many of the requested elements pertaining to AI and imaging studies (e.g., number of experts, adjudication methods, MRMC studies, effect size of human readers improving with AI) are not applicable. However, I will detail the available information relevant to acceptance criteria and performance as presented in the document.
Here's a breakdown of the acceptance criteria and study proving the device meets them, based on the provided FDA 510(k) summary:
Acceptance Criteria and Reported Device Performance
Note: For molecular diagnostic devices, acceptance criteria are typically defined by performance characteristics such as sensitivity, specificity, Limit of Detection (LoD), inclusivity, and exclusivity. The document doesn't explicitly state "acceptance criteria" for clinical performance in a table, but it presents the results of a comparative study against a reference method, which implicitly serves as the standard for acceptance. For analytical performance, criteria are demonstrated through detailed studies (e.g., 100% detection rate at LoD).
Clinical Performance:
| Performance Metric | Acceptance Criteria (Implied/Expected) | Reported Device Performance (NeuMoDx™ GBS Assay) |
|---|---|---|
| Sensitivity | High detection rate of true positives | 96.9% (253/261) |
| 95% CI (94.1 - 98.4) | ||
| Specificity | High detection rate of true negatives | 96.0% (895/932) |
| 95% CI (94.6-97.1) |
Analytical Performance:
| Performance Metric | Acceptance Criteria (Implied/Expected) | Reported Device Performance (NeuMoDx™ GBS Assay) |
|---|---|---|
| Limit of Detection (LoD) | Concentration at which 95% or 100% of replicates are detected. | 500 CFU/mL (100% detection at 500 CFU/mL) |
| Precision/Reproducibility | Consistent results across instruments, operators, and sites, with low variability. | Demonstrated with high percent positive/negative rates across runs, operators, and sites (e.g., 97.7% for Low Positive, 98.7% for Low Negative in Inter-Lab Reproducibility). |
| Analytical Reactivity (Inclusivity) | Detection of all major GBS serotypes. | Detected all 12 tested GBS strains/serotypes at 100% detection (concentrations 400-1500 CFU/mL). |
| Analytical Specificity (Exclusivity) | No cross-reactivity with common urogenital/digestive organisms or related species. | None of 136 screened organisms demonstrated cross-reactivity. |
| Interference | No adverse effect on detection in presence of exogenous/endogenous substances. | No adverse effect on GBS detection observed in the presence of 20 exogenous and 6 endogenous interfering substances. |
| Carry-Over/Cross-Contamination | No contamination in negative samples adjacent to/following high positive samples. | No contamination observed. |
Study Details:
-
A table of acceptance criteria and the reported device performance: Done in the table above.
-
Sample size used for the test set and the data provenance:
- Clinical Performance Test Set: 1193 specimens with complete, valid, and compliant results were included in the study.
- Data Provenance: The study was a "prospective clinical method comparison study" conducted at "three (3) geographically diverse laboratory locations." The data were collected from pregnant women in the US (implied by FDA submission to US authority). Specimens were collected for "routine standard of care screening purposes."
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- This is a molecular diagnostic test. The "ground truth" for clinical performance was established by "conventional culture methods recommended by the Center for Disease Control (CDC) to identify GBS from subcultures of enriched Lim broth." This involves laboratory procedures and standard microbiological protocols rather than expert human interpretation of images. Therefore, the concept of "experts" as in radiologists for imaging is not directly applicable here.
- The document implies that "qualified laboratory personnel" were involved in processing samples and establishing the ground truth via culture.
-
Adjudication method (e.g., 2+1, 3+1, none) for the test set:
- Not applicable as this is a molecular diagnostic test where the ground truth is established by a predefined laboratory culture method, not human adjudication of ambiguous cases. Discordant results between the device and the reference culture method would be analyzed, but not "adjudicated" in the sense of multiple human readers coming to a consensus.
-
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 is a standalone diagnostic device (an automated in vitro diagnostic test) and not an AI-assisted imaging device. There are no "human readers" in the context of interpreting the primary test output or "AI assistance" provided to human readers.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- Yes, the performance characteristics (sensitivity, specificity, LoD, inclusivity, exclusivity, etc.) described for the NeuMoDx™ GBS Assay are "standalone" performance, meaning they relate directly to the output of the automated system. The device automatically determines "Positive, Negative, Indeterminate or Unresolved" results based on pre-established decision criteria (Table 1).
-
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- Clinical Performance Ground Truth: "Conventional culture methods recommended by the Center for Disease Control (CDC) to identify GBS from subcultures of enriched Lim broth" (specifically the 2010 CDC guidelines). This is a laboratory-based reference method, considered the "gold standard" for GBS colonization status.
- Analytical Performance Ground Truth: Established through prepared spiked samples with known concentrations (CFU/mL) of GBS and other organisms.
-
The sample size for the training set:
- The document describes a 510(k) submission for substantial equivalence, not a de novo premarket submission that typically details a separate training set for a machine learning algorithm.
- For a molecular diagnostic test like this, "training" often refers to the internal development and optimization of the assay parameters using characterized samples, rather than a distinct "training set" in the machine learning sense. The document does not specify a formal "training set" size.
-
How the ground truth for the training set was established:
- As noted above, a formal machine learning "training set" is not explicitly defined. Assay development and optimization would have relied on internally validated methods and known concentrations of GBS and non-GBS organisms to establish parameters such as Ct thresholds and call criteria (Table 1). This process implicitly establishes the "ground truth" for the test's internal logic.
§ 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.