(54 days)
The Piccolo® Lactate Test System (presently contained on the MetLac 12 Panel Reagent Disc) used with the Piccolo xpress™ Chemistry Analyzer is intended to be used for the in vitro quantitative determination of lactate concentration in heparinized whole blood or heparinized plasma in a clinical laboratory setting or point-of-care location.
Lactate measurements are used in the diagnosis and treatment of lactate acidosis, monitoring tissue hypoxia, and diagnosis of hyperlactatemia.
The Piccolo MetLac 12 Panel Reagent Disc (which contains the Piccolo Lactate Test System) is designed for lithium heparinized whole blood and lithium heparinized plasma. The disc meters the required quantity of sample and diluent, mixes the sample with diluent, and delivers the mixture to the reaction cuvettes along the disc perimeter. The diluted sample mixes with the reagent beads, initiating the chemical reactions that are then monitored by the analyzer.
Here's a breakdown of the acceptance criteria and study information for the Piccolo® LAC Test System, based on the provided text:
Acceptance Criteria and Device Performance
The acceptance criteria are generally inferred from the "Summary of Safety and Effectiveness" document, which aims to demonstrate substantial equivalence to a predicate device. For each performance characteristic, the "reported device performance" is the result demonstrated by the Piccolo Lactate Test System.
| Acceptance Criteria | Reported Device Performance (Piccolo Lactate Test System) | Study/Method Used |
|---|---|---|
| Linearity | Slope: 1.00 | Linearity Study |
| Intercept: 0.00 | ||
| Correlation Coefficient: 0.999 | ||
| Sensitivity | LOB (Limit of Blank): 0.02 mmol/L | CLSI EP17-A study |
| LOD (Limit of Detection): 0.07 mmol/L | ||
| LOQ (Limit of Quantitation): 0.11 mmol/L | ||
| Precision | Within-Run & Total Precision (Control & Plasma Pool Testing) | Precision Studies |
| Control Level 1: | Mean: 1.62 mmol/L, SD: 0.03 mmol/L, %CV: 1.8 (Within-Run) | |
| Mean: 1.62 mmol/L, SD: 0.04 mmol/L, %CV: 2.2 (Total) | ||
| Control Level 2: | Mean: 3.63 mmol/L, SD: 0.05 mmol/L, %CV: 1.5 (Within-Run) | |
| Mean: 3.63 mmol/L, SD: 0.08 mmol/L, %CV: 2.3 (Total) | ||
| Control Level 3: | Mean: 6.99 mmol/L, SD: 0.18 mmol/L, %CV: 2.6 (Within-Run) | |
| Mean: 6.99 mmol/L, SD: 0.36 mmol/L, %CV: 5.2 (Total) | ||
| Plasma Pool 1: | Mean: 0.86 mmol/L, SD: 0.02 mmol/L, %CV: 1.9 (Within-Run) | |
| Mean: 0.86 mmol/L, SD: 0.02 mmol/L, %CV: 1.9 (Total) | ||
| Plasma Pool 2: | Mean: 6.22 mmol/L, SD: 0.20 mmol/L, %CV: N/A (Within-Run) | |
| Precision (Fresh Whole Blood) - Internal | %CV ranges from 1.7% to 3.3% across 6 samples. | Inter-assay Precision Study |
| Precision (Fresh Whole Blood) - External | Site 1 Combined %CV: 2.0-3.5% | Whole Blood Precision Study at Three External Sites |
| Site 2 Combined %CV: 2.5-4.2% | ||
| Site 3 Combined %CV: 3.2-3.8% | ||
| Method Comparison with Predicate Device (i-STAT) | Slope (Linear Regression): 1.02 (95% CI: 1.01 to 1.04) | Method Comparison Study |
| Intercept (Linear Regression): 0.13 (95% CI: 0.07 to 0.19) | ||
| Correlation Coefficient, R: 0.996 | ||
| Slope (Deming Regression): 1.03 (95% CI: 0.99 to 1.06) | ||
| Intercept (Deming Regression): 0.06 (95% CI: -0.01 to 0.14) | ||
| Interference (Endogenous Substances) | Hemolysis: No interference up to 500 mg/dL | Interference Studies |
| Icterus: No interference up to 15 mg/dL | ||
| Lipemia: No interference up to 3,000 mg/dL | ||
| Interference (Exogenous Substances) | Dopamine: No interference at 0.52 mg/dL | Interference Studies |
| L-dopa: No interference at 0.5 mg/dL | ||
| Reference Interval | 0.53 - 2.10 mmol/L (95% of values) | Reference Interval Determination |
Study Details
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
- Linearity Study: Seven pools (including a saline blank and lowest non-zero calibrator and its dilutions). Each pool assayed 60 times.
- Precision (Control and Plasma Pool Testing): 80 replicates for each Control Level (N=80), 40 replicates for each Plasma Pool (N=40). Data provenance not specified directly but implied to be internal (Abaxis) based on the context.
- Precision (Fresh Whole Blood - Internal): 20 replicates for each of the 6 samples (N=20 for each). Data provenance is internal (Abaxis).
- Precision (Fresh Whole Blood - External): For each of the three external sites, 4 whole blood samples were tested. Each sample was tested by 2 operators (10 replicates each), totaling 20 replicates (N=20 combined) per sample per site. Total samples across 3 sites: 12. Total tests: over 240. Data provenance is implied to be external study sites but the country is not specified.
- Method Comparison with Predicate Device: 126 heparinized whole blood samples from 126 subjects. Data provenance is an external site, country not specified.
- Interference Studies: For endogenous substances, multiple test pools (at least 4) and a control pool were prepared for each of the two lactate levels. For exogenous substances, human plasma pools contained lactate at 0.70 and 2.60 mmol/L. "Two levels of lactate were used in all testing." Data provenance not specified.
- Reference Interval Determination: 130 heparinized whole blood samples from apparently healthy (self-reported) individuals. Data provenance not specified.
All studies appear to be prospective in nature, as they involve testing samples specifically for the purpose of validating the device. The country of origin of the data is not explicitly stated, but the company (Abaxis, Inc.) is based in Union City, CA, USA.
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)
- No "experts" in the traditional sense (e.g., radiologists, pathologists) were used to establish ground truth for the test set in this context. The studies involve analytical performance.
- For the Method Comparison Study, the "ground truth" or reference was the predicate device, the i-STAT Lactate test (Abbott). This device is an already legally marketed and established lactate testing system.
- For other studies (linearity, precision, sensitivity, interference), the "ground truth" is defined by the known concentrations of calibrators, controls, and spiked samples, or by established laboratory reference methods and protocols (e.g., CLSI guidelines).
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
- None of the described studies utilize an adjudication method involving multiple human readers, as these are in vitro diagnostic device performance studies, not image interpretation or clinical decision-making studies. The performance is assessed against quantitative analytical targets or a reference 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
- No MRMC comparative effectiveness study was done. This device is an in vitro diagnostic instrument, not an AI-assisted diagnostic tool that helps human readers. Therefore, the concept of "human readers improve with AI vs without AI assistance" does not apply.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Yes, a standalone performance assessment was done. The entire submission describes the analytical performance of the Piccolo Lactate Test System, which is an automated instrument, functioning as a "standalone" device to measure lactate concentrations. The "human-in-the-loop" component primarily relates to sample collection, loading, and interpreting the final quantitative result.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)
- Known concentrations / Reference Method:
- Linearity, Sensitivity, Precision, Interference: Ground truth was established by using calibrators, controls, and spiked samples with known concentrations of lactate or interferents, according to industry standards (e.g., CLSI guidelines).
- Method Comparison: The predicate device, i-STAT Lactate test (Abbott), served as the reference method for comparison.
- Reference Interval Determination: The interval was determined by testing samples from a population of "apparently healthy" individuals, and the ground truth for their health status was based on "self-reported" information.
8. The sample size for the training set
- The document does not explicitly describe a "training set" in the context of machine learning or AI models. This device is an in vitro diagnostic device based on enzymatic colorimetric assay methodology.
- However, if by "training set" we consider the data used to initially develop and optimize the assay and its performance characteristics (e.g., establishing reagent concentrations, reaction kinetics, calibration algorithms), this information is not detailed in the 510(k) summary. The document focuses on the validation of the final device.
9. How the ground truth for the training set was established
- As noted above, a "training set" as understood in AI/ML is not applicable here. The development and optimization of such a diagnostic assay would typically involve extensive laboratory work to establish optimal reagent concentrations, reaction conditions, and calibration curves. The ground truth for this initial development would come from known standard solutions, reference methods, and clinical samples whose lactate concentrations are reliably determined by established laboratory techniques.
§ 862.1450 Lactic acid test system.
(a)
Identification. A lactic acid test system is a device intended to measure lactic acid in whole blood and plasma. Lactic acid measurements that evaluate the acid-base status are used in the diagnosis and treatment of lactic acidosis (abnormally high acidity of the blood).(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 § 862.9.