(97 days)
For in vitro diagnostic use in the quantitative determination of carbon dioxide in human serum and plasma on ADVIA Chemistry Systems. Such measurements are used in the diagnosis and treatment of numerous potentially serious disorders associated with changes in body acid-base balance.
The ADVIA Chemistry Carbon Dioxide reagent is a solution containing buffer (pH 7.6 at 25°C), 12.5 mmol/L PEP, ≥ 400 U/L PEPC (microbial), ≥ 4100 U/L malate dehydrogenase (mammalian), 0.6 mmol/L NADH analog, activators, stabilizers, a surfactant, and a preservative.
The provided text describes a 510(k) summary for the ADVIA® Chemistry Carbon Dioxide Liquid (CO2 L) Reagent, demonstrating its substantial equivalence to a predicate device. The information focuses on the device's performance characteristics.
Here's an analysis of the acceptance criteria and study that proves the device meets them, based on the provided text:
Preamble: This document does not describe an AI/ML powered device, therefore some of the requested information (like MRMC study, AI assistance, training set details) is not applicable. The device is an in-vitro diagnostic reagent and its acceptance criteria are based on analytical performance compared to a predicate device, not on diagnostic accuracy involving human interpretation.
1. A table of acceptance criteria and the reported device performance
The document does not explicitly state pre-defined acceptance criteria in numerical terms for each performance characteristic. Instead, it states that "All of the evaluation studies gave acceptable results compared to the predicate device" and concludes with a "substantial equivalence" claim. However, we can infer the performance metrics reported.
| Performance Characteristic | Acceptance Criteria (Implied / Comparator) | Reported Device Performance (ADVIA CO2 L Assay) |
|---|---|---|
| Imprecision | Acceptable precision for clinical use | |
| - Within Run (Level 1) | (Not explicitly stated) | Mean: 16.1 mmol/L; SD: 0.19; CV: 1.2% |
| - Within Run (Level 2) | (Not explicitly stated) | Mean: 25.9 mmol/L; SD: 0.17; CV: 0.7% |
| - Within Run (Level 3) | (Not explicitly stated) | Mean: 34.6 mmol/L; SD: 0.33; CV: 1.0% |
| - Total (Level 1) | (Not explicitly stated) | Mean: 16.1 mmol/L; SD: 0.56; CV: 3.5% |
| - Total (Level 2) | (Not explicitly stated) | Mean: 25.9 mmol/L; SD: 0.92; CV: 3.5% |
| - Total (Level 3) | (Not explicitly stated) | Mean: 34.6 mmol/L; SD: 1.21; CV: 3.5% |
| Linearity/Reportable Range | Comparable to predicate device's range (2.9-50.0 mmol/L) and clinically appropriate | 10 to 40 mmol/L (mEq/L) |
| Limit of Detection (LoD) | Clinically appropriate (Implied 0.95) with predicate device | Correlation Coefficient: 0.995 |
| Method Comparison (Regression) | Slope close to 1, intercept close to 0 compared to predicate | Equation: This method = 1.07 (predicate) + 0.75 mmol/L |
| Analytical Specificity (Interference) | 10%, but the text states "with acceptable results" with hemoglobin, and "no significant interference" for unconjugated. This specific bilirubin result is slightly ambiguous in its acceptability within the " >10% variance from the control" significance criterion for icterus. | |
| - Hemolysis | (Not explicitly stated) | - Hemoglobin: Acceptable results from 0-500 mg/dL (in 22.6 mmol/L CO2 sample) |
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
- Imprecision: Assayed two control sera. Each sample was assayed 2 times per run, 2 runs per day, for at least 10 days. The 'N' column in the table refers to the number of replicates (40 for each level). This is a prospective lab-based study using control materials.
- Linearity/Assay Reportable Range: Established based on calculations from the limit of detection and linearity studies. No specific sample size of patient samples is given for the linearity study itself.
- Limit of Detection: 40 replicates of 0.9% saline and 40 replicates of a Low Sample, using one lot of reagent. Data obtained from a 10-day precision study. This is a prospective lab-based study.
- Method Comparison: Sixty-two (62) serum samples. The provenance (country of origin, retrospective/prospective) is not specified, but it's implied to be a prospective comparison of the new and predicate methods.
- Analytical Specificity: Interference studies used samples spiked with interferents (Intralipid, unconjugated bilirubin, hemoglobin). Specific sample sizes for each interference study are not given, but they refer to specific CO2 sample concentrations (e.g., 22.2 mmol/L CO2 sample). This is a controlled lab-based study using spiked samples.
Data Provenance (General): Not explicitly stated, but these are typical studies performed in a lab setting, likely in the US, during product development.
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 is not applicable to an in-vitro diagnostic assay. The "ground truth" for this type of device is established by the analytical reference methods or reference materials themselves, or by comparison to a legally marketed predicate device which serves as the comparative standard. There are no human experts "reading" or interpreting the results in a diagnostic imaging sense.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
This is not applicable. Adjudication methods like 2+1 or 3+1 are used in studies where human experts interpret data (e.g., radiology images) and a consensus is needed to establish ground truth or resolve discrepancies. For an in-vitro diagnostic, the measurements are quantitative chemical analyses, not subjective interpretations.
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 is not applicable. The device is a laboratory reagent, not an AI-powered diagnostic tool that assists human readers.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
The device itself is a "standalone" analytical system in that it chemically quantifies CO2. Its performance is entirely "algorithm only" in the sense of the chemical reactions and spectrophotometric measurements. There is no human interpretation of an image or signal that the device is intending to aid or replace.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The "ground truth" for this type of device is established by:
- Reference Methods/Materials: For imprecision, linearity, and LoD, the device measures against established control materials or standard concentrations.
- Predicate Device: For method comparison, the Genzyme Carbon Dioxide L3K® Assay (K042362) serves as the "ground truth" or reference for comparison, based on its established performance and market history. The goal is to show agreement with this predicate.
8. The sample size for the training set
This is not applicable. This is an IVD reagent developed through chemical and assay optimization, not a machine learning model that requires a "training set" in the context of AI.
9. How the ground truth for the training set was established
This is not applicable as there is no training set for an AI model. The development of the reagent involves standard chemical and analytical validation processes.
§ 862.1160 Bicarbonate/carbon dioxide test system.
(a)
Identification. A bicarbonate/carbon dioxide test system is a device intended to measure bicarbonate/carbon dioxide in plasma, serum, and whole blood. Bicarbonate/carbon dioxide measurements are used in the diagnosis and treatment of numerous potentially serious disorders associated with changes in body acid-base balance.(b)
Classification. Class II.