The QMS® Tobramycin assay is intended for the quantitative determination of tobramycin in human serum or plasma on automated clinical chemistry analyzers.

The results obtained are used in the diagnosis and treatment of tobramycin overdose and in monitoring levels of tobramycin to help ensure appropriate therapy.

The QMS® Tobramycin assay system is a homogeneous assay utilizing particle agglutination technology and is based on the competitive binding principle. The assay consists of reagents R1: anti- tobramycin monoclonal antibody and R2: tobramycin -coated microparticles. A six-level set of QMS® Tobramycin Calibrators (A through F) is used to calibrate the assay.

Here's a breakdown of the acceptance criteria and study information for the Seradyn QMS® Tobramycin assay, based on the provided 510(k) summary:

Acceptance Criteria and Device Performance

Study Details

1. Sample size used for the test set and the data provenance:

   • Accuracy: The exact number of spiked samples is not explicitly stated beyond "across the range of the assay," but each sample was analyzed in triplicate. The samples used were human serum negative for the drug. Data provenance is not specified (e.g., country of origin), but it is a retrospective laboratory study using controlled samples.
   • Linearity: Similar to accuracy, a tobramycin in human serum pool was diluted, and samples were analyzed in triplicate. Data provenance is not specified.
   • Sensitivity (LOQ): Not specified.
   • Method Comparison: N = 67 patient samples. Data provenance is not specified, but it's a retrospective comparison using patient samples.
   • Precision: 80 measurements per control level (Low, Mid, High) over a period of time (e.g., 20 days with 4 replicates per day, which is a common NCCLS EP5-A2 setup, though not explicitly stated here). Data provenance not specified.
   • Interferences: 3 replicates per interfering substance. Data provenance not specified.
   • Cross-reactivity: Each cross-reactant drug was tested at a specific concentration (e.g., 30 µg/mL for 5-Fluorocytosine, 200 µg/mL for Amikacin).
   • Anticoagulants: Not explicitly stated but implies testing with different types of plasma samples and serum.

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

   • Not applicable. This is a quantitative immunoassay for a drug, and the "ground truth" for the test set samples is typically established by precisely preparing known concentrations of tobramycin (for accuracy, linearity, sensitivity) or by using a a well-established comparative method (for method comparison). There are no human expert readers involved in establishing the ground truth for these types of in vitro diagnostic tests.

3. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

   • Not applicable. As described above, this is laboratory testing of a quantitative assay, not an image interpretation or diagnostic performance study requiring expert adjudication. Recovered concentrations are compared to theoretical concentrations or to a predicate device's results.

4. 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 not an AI-assisted diagnostic device, nor does it involve human readers interpreting results in the way an MRMC study evaluates. It's an automated clinical chemistry analyzer.

5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

   • Yes, the performance studies described (Accuracy, Linearity, Sensitivity, Precision, Method Comparison, Interferences, Stability) are all standalone performance evaluations of the assay system itself. This device is entirely automated; there is no "human-in-the-loop" once the sample is loaded and the assay initiated. The results are quantitative measurements produced directly by the instrument.

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

   • For Accuracy and Linearity: The ground truth was established by known, theoretical concentrations of tobramycin prepared by spiking or diluting human serum.
   • For Method Comparison: The "ground truth" was essentially the results obtained from the legally marketed predicate device (Abbott TDx/TDxFLx Tobramycin assay). The study aimed to show correlation and agreement with an already established method.
   • For Precision: The ground truth was the mean concentration of the control materials run multiple times.
   • For Interferences/Cross-reactivity: The ground truth involved known concentrations of tobramycin with the addition of known concentrations of interfering substances or cross-reactants.

7. The sample size for the training set:

   • Not applicable. This is a traditional immunoassay, not a machine learning or AI-based device that requires a "training set" in the computational sense. The reagents and assay parameters are developed and optimized through standard biochemical and analytical chemistry processes, not through training on a dataset.

8. How the ground truth for the training set was established:

   • Not applicable, as there is no "training set" in the context of this device. Development of the assay's reagents and methodologies would involve empirical testing and optimization to achieve desired analytical performance characteristics.

Ask a specific question about this device

The test provides qualitative and semi-quantitative determination of rheumatoid factors (RF) in human serum. The test provides results for one of the seven criteria used in the diagnosis of rheumatoid arthritis as suggested by the American College of Rheumatology

ColorSlide® RF is a rapid latex particle agglutination assay for the determination of circulating Rheumatoid Factors (RF) in human serum.

The provided text describes the Seradyn ColorSlide® RF device and its performance in comparison to predicate devices, SeraTest™ RF and Behring's Rapitex RF.

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly state pre-defined acceptance criteria in terms of numerical thresholds for sensitivity, specificity, and efficiency. Instead, it states that "An acceptable correlation was obtained between the two tests" (ColorSlide® RF vs. SeraTest™ RF and ColorSlide® RF vs. RapiTex RF). The reported performance values are:

Comparison to SeraTest™ RF (Qualitative Assay)

Comparison to RapiTex RF (Qualitative Assay)

Semi-Quantitative Agreement

Lot-to-Lot Reproducibility: All titers fell within one doubling dilution, indicating satisfactory performance with little or no lot-to-lot variability.

Sensitivity based on international standard: Approximately 20 IU/mL, traceable to a WHO reference serum.

Passive Interference: No adverse interference from hemoglobin (at 1000 mg/dL), bilirubin (at 40 mg/dL), or lipid (at 20 g/L).

Stability Data: Performance within specifications for approximately 400 days for one ColorSlide® RF lot, with ongoing studies for at least two years.

2. Sample Size Used for the Test Set and Data Provenance

• Comparison to SeraTest™ RF:
  • Sample Size: 103 patient samples.
  • Data Provenance: Testing was conducted at one site. The country of origin is not specified but is implicitly the USA, given the company's address and the FDA 510(k) submission. It's a retrospective comparison study, as patient samples were tested by both methods.
• Comparison to Behring's RapiTex RF:
  • Sample Size: 309 patient samples.
  • Data Provenance: Testing was conducted at three sites. Country of origin implicitly USA. Retrospective comparison study.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications

The studies compare the ColorSlide® RF device to two legally marketed predicate devices (SeraTest™ RF and RapiTex RF). The ground truth for sensitivity and specificity is established by the results of these predicate devices. The document does not mention the involvement of "experts" in establishing a separate ground truth or their qualifications. The readings of the agglutination reactions for both the ColorSlide® RF and the predicate devices are "read visually," implying trained laboratory personnel but no specifically defined "experts" for ground truth establishment.

4. Adjudication Method for the Test Set

No explicit adjudication method is described. The comparison is made directly between the results of the ColorSlide® RF and the predicate device. For the semi-quantitative assay, agreement is categorized. There is no mention of a process like 2+1 or 3+1 for resolving discrepancies.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No MRMC comparative effectiveness study was done. The document does not describe human readers using AI assistance or measuring its effect size. The device is a diagnostic assay, not an AI software.

6. Standalone (Algorithm Only) Performance

The device is a laboratory assay (latex agglutination test), not an algorithm or AI software. Therefore, the concept of "standalone performance" in the context of an algorithm does not apply. Its performance is inherent in the chemical and physical reaction, read visually by laboratory personnel.

7. Type of Ground Truth Used

The ground truth for the performance assessment (sensitivity, specificity, predictive values) is established by the results of the predicate devices (Seradyn's SeraTest™ RF and Behring's Rapitex RF). For the semi-quantitative assays, the ground truth for agreement is also based on the titers obtained from the predicate devices. The document does not mention pathology, outcomes data, or expert consensus as primary ground truth references for the clinical performance assessment.

8. Sample Size for the Training Set

This information is not applicable. The ColorSlide® RF is a latex agglutination assay, not an AI or machine learning algorithm that requires a "training set." Its development would involve chemical and manufacturing processes, and its characteristics are assessed against established methods, not by training.

9. How the Ground Truth for the Training Set was Established

As mentioned above, there is no "training set" in the context of an AI/ML algorithm. Therefore, this question is not applicable to the ColorSlide® RF device. The assay development would have involved validation against known positive and negative samples, but these are part of R&D and not typically referred to as a "training set" in this context.

Ask a specific question about this device

Ask a specific question about this device