Search Results

Magnesium measurements are used in the diagnosis and treatment of hypomagnesemia (abnormally low plasma levels of magnesium) and hypermagnesemia (abnormally high plasma levels of magnesium).

Device Description

The Magnesium reagent kit contains Reagent 1 and Reagent 2. Magnesium present in the sample is a cofactor in an enzymatic reaction with isocitrate dehydrogenase. The rate of increase in absorbance at 340 nm, due to the formation of NADPH, is directly proportional to the magnesium concentration.

AI/ML Overview

The provided document is a 510(k) summary for a medical device called "Magnesium" from Abbott Laboratories. It details the performance testing conducted to demonstrate its substantial equivalence to a predicate device. This document describes an in vitro diagnostic (IVD) device for measuring magnesium levels, not an AI/ML-based device. Therefore, many of the requested elements related to AI (e.g., ground truth establishment by experts, adjudication, MRMC studies, training set details) are not applicable to this submission.

However, I can extract information relevant to the acceptance criteria and performance of this IVD device.

Here's the breakdown of the acceptance criteria and study that proves the device meets them, based on the provided text:

Device: Abbott Laboratories Magnesium Assay (List No. 3P68)
Intended Use: Quantitation of magnesium in human serum, plasma, or urine on the ARCHITECT c8000 System for diagnosis and treatment of hypomagnesemia and hypermagnesemia.
Predicate Device: Roche Magnesium Gen.2 (K983416)

1. Acceptance Criteria and Reported Device Performance

The document describes various performance studies and their results. The implicit acceptance criteria are that the device's performance characteristics demonstrate substantial equivalence to the predicate device and are within acceptable ranges for clinical utility.

Study Type	Acceptance Criteria (Implicit)	Reported Device Performance
Limit of Blank (LoB)	LoB should be low, demonstrating minimal signal in the absence of analyte. (No explicit numerical criterion stated, but values are reported).	Urine application: LoB of 0.04 mg/dL
Limit of Detection (LoD)	LoD should be low enough to detect clinically relevant low levels. (No explicit numerical criterion stated, but values are reported).	Urine application: LoD of 0.09 mg/dL
Limit of Quantitation (LoQ)	LoQ should be low enough for accurate quantitation at clinically relevant low levels. (No explicit numerical criterion stated, but values are reported).	Urine application: LoQ of 0.75 mg/dL
Within-Laboratory Precision (Imprecision)	%CV should be clinically acceptable, demonstrating consistency of results over time. (No explicit numerical criterion stated, but values are reported as evidence of acceptable precision).	Urine (within-laboratory imprecision):- Bio-Rad Level 1: 1.3 %CV- Bio-Rad Level 2: 1.3 %CV- LoQ Urine Pool -Low Mg: 2.4 %CV- Human Urine Pool - Normal Mg: 1.8 %CV- Human Urine Pool Abnormal Mg: 1.8 %CV
Interference	Assay results should be impacted by no more than ±10% for specific interferent levels.	For magnesium samples targeted to 5 mg/dL: no more than ±10% interference for listed substances (Albumin ≤ 64.0 mg/dL, Ascorbic Acid ≤ 200 mg/dL, Bilirubin (Conjugated) ≤ 59.9 mg/dL, Calcium ≤ 26.0 mg/dL, Glucose ≤ 1220 mg/dL, Hemoglobin ≤ 1200 mg/dL, Phosphorous ≤ 307 mg/dL, Boric Acid ≤ 1000 mg/dL, 6N Hydrochloric Acid ≤ 3.0 mL/dL, Copper ≤ 21.6 µg/dL, Zinc ≤ 3504 µg/L, Iron ≤ 0.6 mg/dL).For magnesium samples targeted to 14 or 15 mg/dL: similar results for the same interferents with slightly different calcium and bilirubin levels.Acetic acid, nitric acid, and sodium fluoride did not meet the ±10% criterion and are noted as limitations.
Linearity	The assay should be linear across its analytical measuring interval. (No explicit R-squared or slope criterion, but stated that it was "demonstrated to be linear").	Urine application demonstrated linearity across 1.04 to 36.24 mg/dL, spanning the analytical measuring interval of 1.81 to 26.35 mg/dL.
Measuring Interval (Analytical Measuring Range)	Defined by LoQ and highest linear point.	1.81 to 26.35 mg/dL
Method Comparison (Correlation to Predicate)	Demonstrate acceptable correlation (slope and correlation coefficient) to the predicate device across the measuring interval.	Urine application showed acceptable correlation to predicate:- pH <2: slope 1.08, r-value 1.00- No acidification: slope 1.04, r-value 1.00- pH ~1: slope 1.05, r-value 1.00
Automated Dilution Protocol	Results impacted by not more than ±10% at high analyte concentrations.	Impacted by not more than ±10% for analyte concentration of 75 mg/dL.
Manual Dilution	Results impacted by not more than ±10% for specific analyte concentrations after manual dilution.	Impacted by not more than ±10% for analyte concentrations 15, 25, and 40 mg/dL with 1:2 manual dilution.

2. Sample Size and Data Provenance

Test Set Sample Sizes:
- LoB, LoD, LoQ: 4 saline samples (zero-analyte), 2 low-analyte level samples at each of 5 target concentrations (0.390, 0.780, 1.285, 1.560, 3.130 mg/dL). Zero-analyte samples tested in 10 replicates. Low-analyte samples tested in 10 replicates.
- Within-Laboratory Precision: Control materials (Bio-Rad Level 1 & 2, LoQ Urine Pool -Low Mg, Human Urine Pool Normal Mg, Human Urine Pool Abnormal Mg) tested in 2 replicates, 2 times per day for 20 testing days.
- Interference (Endogenous/Preservatives): Control and test level samples tested in a minimum of 7 replicates.
- Interference (Cations - Copper/Zinc): Control sample (urine) and test level samples tested in 12 replicates.
- Linearity: 3 sets of linearity standards, each with 12 levels. Each level tested in a minimum of 4 replicates.
- Method Comparison: 118 patient urine specimens. 12 of these were normal urine specimens spiked with magnesium.
- Automated Dilution: 1 sample (75 mg/dL target) evaluated in 10 replicates.
- Manual Dilution: 3 urine pools (15, 25, 40 mg/dL target). Each pool evaluated after 1:2 manual dilution. Samples tested in a minimum of 12 replicates.
Data Provenance: The document does not explicitly state the country of origin for the patient samples or if they were retrospective or prospective. Given the submitter is an American company (Abbott Laboratories, Irving, TX) and the regulatory submission is to the U.S. FDA, it is highly likely the studies were conducted in the U.S. and the samples were generally from a U.S. population. The studies described are performance studies for an IVD, which typically use banked or collected samples, so they would be considered retrospective in terms of sample collection for the specific study; however, the study itself is prospective in its design and execution.

3. Number of Experts and Qualifications for Ground Truth

Not Applicable. This is an IVD device measuring analyte concentration using a chemical reaction. The "ground truth" for the test set is established by the analytical reference methods and gravimetric preparations, not by human expert interpretation (like radiologists for imaging).

4. Adjudication Method for Test Set

Not Applicable. As this is an IVD device measuring a chemical analyte, there is no expert adjudication process for the test set results. The ground truth for performance studies is based on traceable standards, established reference methods, or gravimetric preparations.

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

Not Applicable. MRMC studies are relevant for devices that involve human interpretation, particularly in diagnostic imaging or pathology. This is an automated IVD device.

6. Standalone (Algorithm Only) Performance

The performance data presented (LoB, LoD, LoQ, Precision, Interference, Linearity, Analytical Measuring Interval, Method Comparison, Dilution studies) are all standalone performance metrics of the Magnesium assay on the ARCHITECT c8000 System. There is no human-in-the-loop component for the measurement itself; the device provides a direct quantitative result.

7. Type of Ground Truth Used

The ground truth for these studies relies on:

Reference Standards/Materials: For LoB, LoD, LoQ, Linearity, and Interference studies, prepared samples with known concentrations (e.g., saline, diluted or spiked magnesium standards).
Established Methods/Predicate Device: For Method Comparison, the results are compared to a legally marketed predicate device (Roche Magnesium Gen.2), which serves as the "truth" for demonstrating substantial equivalence.
Control Materials: Commercially available quality control materials and in-house prepared human urine pools with expected magnesium levels are used for precision studies.

8. Sample Size for the Training Set

Not Applicable/Not Explicitly Stated. For an IVD like this, there isn't a "training set" in the sense of machine learning algorithms. The device's operational parameters (reagent formulation, reaction kinetics, detection methods) are developed through R&D and optimization, not typically by training on a distinct data set in the way an AI model is trained. The studies described are validation and verification studies to demonstrate performance, not a training process.

9. How Ground Truth for the Training Set Was Established

Not Applicable. As there's no "training set" in the typical AI sense, the establishment of ground truth for such a set is not relevant. The analytical parameters of the device are based on established chemical principles and optimized through research and development.

Ask a Question

Ask a specific question about this device

Page 1 of 1