(151 days)
In vitro test for the quantitative determination of vancomycin in serum and plasma on Roche/Hitachi cobas c systems.
A vancomycin test system is a device intended to measure vancomycin, an antibiotic drug, in serum and plasma. Measurements obtained by this device are used in the diagnosis and treatment of vancomycin overdose and in monitoring the level of vancomycin to ensure appropriate therapy.
The ONLINE TDM Vancomycin Gen.3 is a two reagent assay for the in vitro quantitative determination of vancomycin in human serum or plasma on automated clinical chemistry analyzers. It is a homogeneous microparticle agglutination immunoassay based on the kinetic interaction of microparticles in solution (KIMS). A competitive reaction takes place between the drug conjugate and vancomycin in the serum sample for binding to the vancomycin antibody on the microparticles. The resulting kinetic interaction of microparticles is indirectly proportional to the amount of drug present in the sample.
The provided text details the performance evaluation of the "ONLINE TDM Vancomycin Gen.3" device. Here's a breakdown of the requested information:
1. A table of acceptance criteria and the reported device performance
The document doesn't explicitly state "acceptance criteria" as separate rows but presents performance study results, implying these results are considered acceptable for substantial equivalence. The table below compiles the reported performance data.
| Performance Metric | Acceptance Criteria (Implied by reported performance) | Reported Device Performance (ONLINE TDM Vancomycin Gen.3) |
|---|---|---|
| Detection Limit | ||
| Limit of Blank (LoB) | $\le$ 1.0 µg/mL | MP Lot: 0.6 µg/mL |
| P2 Lot: 0.7 µg/mL | ||
| P3 Lot: 0.5 µg/mL | ||
| Limit of Detection (LoD) | $\le$ 1.5 µg/mL | MP Lot: 1.4 µg/mL |
| P2 Lot: 1.3 µg/mL | ||
| P3 Lot: 1.1 µg/mL | ||
| Limit of Quantitation (LoQ) | $\le$ 4.0 µg/mL | MP Lot: 2.0 µg/mL |
| P2 Lot: 3.3 µg/mL | ||
| P3 Lot: 3.1 µg/mL | ||
| Precision (Repeatability) | Not explicitly stated, but consistent low CVs are expected | TDM Control 1: Mean 7.45 µg/mL, SD 0.4 µg/mL, CV 5.2% |
| TDM Control 2: Mean 21.5 µg/mL, SD 0.5 µg/mL, CV 2.3% | ||
| TDM Control 3: Mean 36.2 µg/mL, SD 0.9 µg/mL, CV 2.4% | ||
| Human Serum 1: Mean 4.82 µg/mL, SD 0.4 µg/mL, CV 8.2% | ||
| Human Serum 2: Mean 7.95 µg/mL, SD 0.4 µg/mL, CV 5.2% | ||
| Human Serum 3: Mean 32.1 µg/mL, SD 0.8 µg/mL, CV 2.5% | ||
| Human Serum 4: Mean 40.0 µg/mL, SD 1.0 µg/mL, CV 2.5% | ||
| Human Serum 5: Mean 71.4 µg/mL, SD 2.0 µg/mL, CV 2.8% | ||
| Precision (Intermediate) | Not explicitly stated, but consistent low CVs are expected | TDM Control 1: Mean 7.45 µg/mL, SD 0.5 µg/mL, CV 6.2% |
| TDM Control 2: Mean 21.5 µg/mL, SD 0.8 µg/mL, CV 3.7% | ||
| TDM Control 3: Mean 35.5 µg/mL, SD 1.1 µg/mL, CV 3.2% | ||
| Human Serum 1: Mean 4.93 µg/mL, SD 0.5 µg/mL, CV 10.5% | ||
| Human Serum 2: Mean 7.95 µg/mL, SD 0.5 µg/mL, CV 5.9% | ||
| Human Serum 3: Mean 32.1 µg/mL, SD 1.1 µg/mL, CV 3.4% | ||
| Human Serum 4: Mean 39.5 µg/mL, SD 1.1 µg/mL, CV 2.9% | ||
| Human Serum 5: Mean 71.4 µg/mL, SD 2.2 µg/mL, CV 3.1% | ||
| Linearity | Pearson correlation coefficient (R) close to 1, slope close to 1, intercept close to 0 | Serum: y=1.000x-0.000, R=0.9985 |
| Plasma: y=1.000x-0.000, R=0.9976 | ||
| Measuring Range | 4.0 to 80.0 µg/mL | 4.0 to 80.0 µg/mL (Claimed, consistent with linearity results) |
| Matrix Comparison | Strong correlation (r value close to 1) between plasma and serum measurements | Serum vs. Li-heparin: y = 1.01x -0.3, r = 0.996 |
| Serum vs. K2-EDTA: y = 0.99x -0.0, r = 0.996 | ||
| Serum vs. K3-EDTA: y = 1.00x - 0.3, r = 0.995 | ||
| Endogenous Interference | No significant interference up to stated levels | Hemolysis: No interference up to H index of 1000 (1000 mg/dL hemoglobin) |
| Lipemia: No interference up to L index of 1000 (1000 mg/dL triglycerides) | ||
| Icterus (Unconjugated Bilirubin): No interference up to I index of 60 (60 mg/dL or 1026 umol/L) | ||
| Icterus (Conjugated Bilirubin): No interference up to I index of 60 (60 mg/dL or 1026 umol/L) | ||
| Drug Interference | No interference up to specified concentrations | No interference observed for a list of common drugs at specified concentrations (e.g., Acetylsalicylic acid 1000 mg/L, Acetaminophen 200 mg/L, Heparin 5000 U/L, etc.) |
| Method Comparison to Predicate | Strong correlation with predicate device, Passing Bablok regression results with slope close to 1 and intercept close to 0 | y = 0.993x + 0.641, r = 0.994 |
2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
- Detection Limit (LoB, LoD):
- LoB: 60 measurements (5-fold determinations per run, 6 runs over 3 days, across one or two instruments) per lot.
- LoD: 30 samples/measurements (5 serum samples, 6 runs over 3 days, 1-fold or 2-fold determination per run, across one or two instruments) per lot.
- LoQ: 15 serum samples, tested in two aliquots over 6 runs for 4 days.
- Precision: Not explicitly stated, but "two runs per day for $\geq$ 21 days" were performed. The number of samples for controls (3) and human serum (5) are listed, with replicates per run not specified but implied to be sufficient for precision calculations.
- Linearity: Sixteen levels (dilution series from a human serum sample pool and diluent) were prepared. The process was repeated for plasma samples.
- Matrix Comparison: 67 full tubes and 9 half-filled tubes of serum and plasma from a single donor. (K2-EDTA plasma had 10 half-filled tubes).
- Interferences (H, L, I Indices): Two human serum sample pools spiked with Vancomycin. An 11-step dilution series prepared for each interferent, with 3 aliquots per level tested.
- Interferences (Drugs): Two human serum sample pools spiked with Vancomycin. Tested with 3 replicates.
- Method Comparison to Predicate: 125 single native human serum samples from patients taking Vancomycin. 8 additional native Vancomycin samples were spiked, and 1 sample diluted to cover the range.
The data provenance regarding the country of origin is not specified. All studies appear to be prospective experimental studies conducted in a controlled lab setting, rather than retrospective patient data analysis.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)
This device is an in-vitro diagnostic (IVD) assay for quantitative determination of vancomycin concentration. The "ground truth" for these types of devices is based on established analytical reference methods or assigned values.
- For LoQ, the expected value was determined with Vancomycin LCMS/MS, which is a highly accurate reference method, not expert consensus.
- For Method Comparison to Predicate, the predicate device (ONLINE TDM Vancomycin, K060586) served as the reference standard for comparison, not human experts.
- No human experts were involved in establishing the ground truth for any of these analytical performance studies. These are laboratory-based measurements.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
Not applicable. This is an IVD device for quantitative measurement, not an imaging or qualitative diagnostic device requiring expert adjudication. The "truth" is determined by analytical reference methods or comparison to a predicate device.
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 is an IVD device for quantitative measurement, not an AI-assisted diagnostic tool for human readers.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
This entire submission describes the standalone performance of the ONLINE TDM Vancomycin Gen.3 assay. The device itself is an automated chemical analyzer system (Roche/Hitachi cobas c systems) that performs the assay, not an AI algorithm. The performance metrics listed (detection limit, precision, linearity, interference, method comparison) are all standalone analytical performance characteristics of the device.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The ground truth or reference methods used for evaluation include:
- LCMS/MS (Liquid Chromatography-Mass Spectrometry/Mass Spectrometry): Used for determining expected values for the Limit of Quantitation (LoQ).
- Predicate Device (ONLINE TDM Vancomycin, K060586): Used as the comparative reference for the method comparison study.
- Standard Analytical Protocols: Studies like precision, linearity, and interference follow established CLSI guidelines, where the "truth" is defined by the experimental setup (e.g., known concentrations of analytes, spiked interferents).
8. The sample size for the training set
Not applicable. This device is an IVD assay, not a machine learning or AI-based system that requires a "training set" in the conventional sense. The development of such assays involves reagent formulation, optimization, and extensive analytical verification and validation, but not typically "training data" for an algorithm.
9. How the ground truth for the training set was established
Not applicable, as there is no "training set" for an AI algorithm here.
§ 862.3950 Vancomycin test system.
(a)
Identification. A vancomycin test system is a device intended to measure vancomycin, an antibiotic drug, in serum. Measurements obtained by this device are used in the diagnosis and treatment of vancomycin overdose and in monitoring the level of vancomycin to ensure appropriate therapy.(b)
Classification. Class II.