(96 days)
The ACE Direct Total Iron-Binding Capacity (TIBC) Reagent is intended for the quantitative determination of total iron-binding capacity in serum using the ACE Alera Clinical Chemistry System. Iron-binding capacity measurements are used in the diagnosis and treatment of anemia. This test is intended for use in clinical laboratories and physician office laboratories. For in vitro diagnostic use only.
The ACE Total Iron Reagent is intended for the quantitative determination of iron in serum using the ACE Alera Clinical Chemistry System. Iron (non-heme) measurements are used in the diagnosis and treatment of diseases such as iron deficiency anemia, hemochromatosis (a disease associated with widespread deposit in the tissues of two iron-containing pigments, hemosiderin and hemofuscin, and characterized by pigmentation of the skin), and chronic renal disease. This test is intended for use in clinical laboratories and physician office laboratories. For in vitro diagnostic use only.
The ACE LDH-L Reagent is intended for the quantitative determination of lactate dehydrogenase activity in serum using the ACE Alera Clinical Chemistry System. Lactate dehydrogenase measurements are used in the diagnosis and treatment of liver diseases such as acute viral hepatitis, cirrhosis, and metastatic carcinoma of the liver, cardiac diseases such as myocardial infarction and tumors of the lung or kidneys. This test is intended for use in clinical laboratories and physician office laboratories. For in vitro diagnostic use only.
In the ACE Direct Total Iron-Binding Capacity (TIBC) Reagent assay, Direct TIBC Color Reagent, an acidic buffer containing an iron-binding dye and ferric chloride, is added to the serum sample. The low pH of Direct TIBC Color Reagent releases iron from transferrin. The iron then forms a colored complex with the dye. The colored complex at the end of the first step represents both the serum iron and excess iron already present in Direct TIBC Color Reagent. Direct TIBC Buffer, a neutral buffer, is then added, shifting the pH and resulting in a large increase in the affinity of transferrin for iron. The serum transferrin rapidly binds the iron by abstracting it from the dye-iron complex. The observed decrease in absorbance of the colored dye-iron complex is directly proportional to the total iron-binding capacity of the serum sample. The absorbance is measured at 647 nm.
In the ACE Total Iron Reagent assay, transferrin-bound iron in serum is released at an acidic pH and reduced from ferric to ferrous ions. These ions react with ferrozine to form a violet colored complex, which is measured bichromatically at 554 nm/692 nm. The intensity of color produced is directly proportional to the serum iron concentration.
In the ACE LDH-L Reagent assay, lactate dehydrogenase catalyzes the conversion of L-lactate to pyruvate. Nicotinamide adenine dinucleotide (NAD+) acts as an acceptor for the hydrogen ions released from the L-lactate and is converted to reduced nicotinamide adenine dinucleotide (NADH). NADH absorbs strongly at 340 nm whereas NAD+ does not. Therefore, the rate of conversion of NAD+ to NADH can be determined by monitoring the increase in absorbance bichromatically at 340 nm/647 nm. This rate of conversion from NAD+ to NADH is directly proportional to the lactate dehydrogenase activity in the sample.
The provided document describes in vitro diagnostic (IVD) reagents (ACE Direct Total Iron-Binding Capacity (TIBC) Reagent, ACE Total Iron Reagent, and ACE LDH-L Reagent) for use on the ACE Alera Clinical Chemistry System. The acceptance criteria and performance data presented relate to the analytical performance of these reagents/systems, specifically their ability to accurately and precisely measure analytes in serum samples.
Crucially, this is not a study about an AI/ML powered medical device. Therefore, many of the typical acceptance criteria and study aspects requested in your prompt regarding AI/ML (e.g., ground truth established by experts, multi-reader multi-case studies, human-in-the-loop performance, training/test set sample sizes for AI, adjudication methods) are not applicable to this type of device and submission.
The "study" described here is a series of analytical performance tests (linearity, precision, method comparison, detection limits, interference) to demonstrate that the new device (ACE Alera system with these reagents) performs comparably to the predicate device (ACE Clinical Chemistry System with the same reagents) and meets established analytical performance specifications for clinical chemistry assays.
Here's a breakdown of the relevant information from the document in the format you requested, with an explanation of why certain AI/ML-centric points are not applicable:
Device: ACE Direct Total Iron-Binding Capacity (TIBC) Reagent, ACE Total Iron Reagent, ACE LDH-L Reagent (for use on ACE Alera Clinical Chemistry System)
1. Table of acceptance criteria and reported device performance:
Since the document does not explicitly present "acceptance criteria" alongside "reported performance" in a single table, I will infer the acceptance criteria from the context of method comparison, linearity, and precision studies, which are standard for IVD device validation, often aiming for performance comparable to predicate devices or within clinically acceptable limits. The reported performance is directly extracted from the tables provided.
Interference:
The acceptance criterion for interference studies in IVD assays is typically that the interferent, up to a specified concentration, does not cause a "significant interference" (e.g., a bias exceeding a defined clinical or analytical threshold). The document lists the concentrations at which no significant interference was observed.
| Interferent | No Significant Interference at or below (Acceptance Criteria Implicit) | Reported Device Performance (Concentration where no significant interference was observed) |
|---|---|---|
| TIBC | ||
| Icterus | Assumes standard limits for non-interference | 59 mg/dL |
| Hemolysis | Assumes standard limits for non-interference | 188 mg/dL* |
| Lipemia | Assumes standard limits for non-interference | 1000 mg/dL |
| Ascorbic Acid | Assumes standard limits for non-interference | 3 mg/dL |
| Iron | ||
| Icterus | Assumes standard limits for non-interference | 59 mg/dL |
| Hemolysis | Assumes standard limits for non-interference | 125 mg/dL* |
| Lipemia | Assumes standard limits for non-interference | 125 mg/dL |
| Ascorbic Acid | Assumes standard limits for non-interference | 6 mg/dL |
| LDH-L | ||
| Icterus | Assumes standard limits for non-interference | 50 mg/dL |
| Hemolysis | Assumes standard limits for non-interference | 0.99), ideally with the confidence intervals for slope encompassing 1 and for intercept encompassing 0. This indicates analytical equivalence between the two systems. |
| Reagent | n (samples) | Range (of samples tested) | Reported Slope | Reported Intercept | Reported Correlation Coefficient | CI Slope | CI Intercept |
|---|---|---|---|---|---|---|---|
| TIBC | 50 | 59 to 676 µg/dL | 0.987 | 3.6 | 0.9960 | 0.962 to 1.013 | -7.2 to 14.4 |
| Iron | 48 | 13 to 549 µg/dL | 0.993 | 0.9 | 0.9995 | 0.984 to 1.003 | -0.6 to 2.3 |
| LDH-L | 58 | 20 to 799 U/L | 0.997 | -3.6 | 0.9991 | 0.985 to 1.008 | -6.1 to -1.1 |
Precision (POL - Point of Care/Physician Office Lab):
Similar to in-house precision, specific %CV or SD limits would be the acceptance criteria. The data shows results from 3 POLs compared to in-house.
| Reagent | Lab | Sample Level | Mean | Within-Run SD, %CV | Total SD, %CV |
|---|---|---|---|---|---|
| Direct TIBC | In-House | 1 | 330 | 5.1, 1.5% | 5.8, 1.8% |
| POL 1 | 1 | 284 | 8.3, 2.9% | 9.6, 3.4% | |
| POL 2 | 1 | 259 | 5.6, 2.2% | 8.5, 3.3% | |
| POL 3 | 1 | 276 | 9.1, 3.3% | 16.7, 6.0% | |
| In-House | 2 | 450 | 4.9, 1.1% | 6.8, 1.5% | |
| POL 1 | 2 | 464 | 6.3, 1.4% | 6.6, 1.4% | |
| POL 2 | 2 | 444 | 4.2, 1.0% | 5.4, 1.2% | |
| POL 3 | 2 | 453 | 3.2, 0.7% | 15.5, 3.4% | |
| In-House | 3 | 530 | 9.4, 1.8% | 10.8, 2.0% | |
| POL 1 | 3 | 544 | 8.2, 1.5% | 8.3, 1.5% | |
| POL 2 | 3 | 520 | 5.0, 1.0% | 9.0, 1.7% | |
| POL 3 | 3 | 533 | 12.6, 2.4% | 20.2, 3.8% | |
| Total Iron | In-House | 1 | 119 | 1.8, 1.5% | 2.5, 2.1% |
| POL 1 | 1 | 119 | 2.7, 2.3% | 3.2, 2.7% | |
| POL 2 | 1 | 122 | 3.1, 2.6% | 3.1, 2.6% | |
| POL 3 | 1 | 116 | 3.2, 2.8% | 3.4, 3.0% | |
| In-House | 2 | 222 | 3.8, 1.7% | 5.1, 2.3% | |
| POL 1 | 2 | 229 | 2.0, 0.9% | 2.5, 1.1% | |
| POL 2 | 2 | 235 | 2.3, 1.0% | 2.4, 1.0% | |
| POL 3 | 2 | 229 | 3.4, 1.5% | 3.9, 1.7% | |
| In-House | 3 | 412 | 5.2, 1.3% | 5.7, 1.4% | |
| POL 1 | 3 | 424 | 4.0, 0.9% | 4.6, 1.1% | |
| POL 2 | 3 | 435 | 2.4, 0.5% | 5.3, 1.2% | |
| POL 3 | 3 | 428 | 11.1, 2.6% | 11.1, 2.6% | |
| LDH-L | In-House | 1 | 118 | 2.9, 2.4% | 5.7, 4.8% |
| POL 1 | 1 | 116 | 1.7, 1.5% | 4.9, 4.3% | |
| POL 2 | 1 | 118 | 3.0, 2.5% | 5.1, 4.3% | |
| POL 3 | 1 | 124 | 3.4, 2.7% | 4.7, 3.8% | |
| In-House | 2 | 433 | 4.7, 1.1% | 6.5, 1.5% | |
| POL 1 | 2 | 437 | 2.9, 0.7% | 5.8, 1.3% | |
| POL 2 | 2 | 449 | 3.7, 0.8% | 5.2, 1.2% | |
| POL 3 | 2 | 446 | 5.8, 1.3% | 6.6, 1.5% | |
| In-House | 3 | 699 | 5.3, 0.8% | 8.5, 1.2% | |
| POL 1 | 3 | 698 | 8.6, 1.2% | 11.5, 1.6% | |
| POL 2 | 3 | 726 | 5.4, 0.8% | 10.0, 1.4% | |
| POL 3 | 3 | 716 | 14.3, 2.0% | 16.9, 2.4% |
Method Comparison (POLs vs. In-House (ACE Alera (x) vs. POL ACE Alera (y))):
Similar to the in-house method comparison, the acceptance criteria are for slopes to be near 1, intercepts near 0, and high correlation coefficients (e.g., >0.99), indicating consistent performance across different lab environments.
| Reagent | Lab Comparison | n (samples) | Range | Reported Regression | Reported Correlation | CI Slope | CI Intercept |
|---|---|---|---|---|---|---|---|
| TIBC | In-House vs. POL 1 | 50 | 59 to 676 | y = 0.994x + 12.4 | 0.9934 | 0.961 to 1.027 | -1.7 to 26.5 |
| In-House vs. POL 2 | 50 | 59 to 676 | y = 0.973x + 0.1 | 0.9954 | 0.946 to 1.001 | -11.4 to 11.6 | |
| In-House vs. POL 3 | 50 | 59 to 676 | y = 1.005x + 9.0 | 0.9898 | 0.963 to 1.047 | -8.7 to 26.6 | |
| Iron | In-House vs. POL 1 | 48 | 13 to 549 | y = 0.976x + 1.0 | 0.9986 | 0.960 to 0.991 | -1.4 to 3.3 |
| In-House vs. POL 2 | 48 | 13 to 549 | y = 0.976x + 2.3 | 0.9981 | 0.959 to 0.994 | -0.4 to 5.0 | |
| In-House vs. POL 3 | 48 | 13 to 549 | y = 0.951x + 0.8 | 0.9966 | 0.927 to 0.974 | -2.7 to 4.4 | |
| LDH-L | In-House vs. POL 1 | 51 | 74 to 799 | y = 0.992x + 3.5 | 0.9986 | 0.977 to 1.008 | -0.1 to 7.1 |
| In-House vs. POL 2 | 51 | 74 to 799 | y = 1.027x + 3.4 | 0.9989 | 1.013 to 1.041 | 0.2 to 6.7 | |
| In-House vs. POL 3 | 51 | 74 to 799 | y = 1.010x + 2.5 | 0.9984 | 0.994 to 1.026 | -1.3 to 6.2 |
2. Sample sizes used for the test set and the data provenance:
-
Sample Sizes for analytical performance studies (Test Set):
- Method Comparison:
- TIBC: 50 samples
- Iron: 48 samples
- LDH-L: 58 (in-house comparison) / 51 (POL comparison) samples
- Linearity: The number of samples/levels for linearity is not explicitly stated as 'n', but standard practice involves multiple levels (typically 5-7) prepared from diluted/spiked samples.
- Precision: Standard runs (e.g., 2 runs per day for 20 days for total precision, with replicates per run for within-run precision) would involve a substantial number of measurements (e.g., 20 days x 2 runs/day x 2 replicates/run = 80 measurements per level). The POL precision data shows n=20, likely referring to 20 days of testing.
- Interference: The number of samples used for interference studies is not explicitly stated.
- Method Comparison:
-
Data Provenance: "In-House" and "POL" (Physician Office Laboratories). The specific country of origin is not explicitly stated, but given the company's location (New Jersey, USA) and FDA 510(k) submission, it's highly likely to be United States. The studies are prospective analytical validation studies, meaning the data was collected specifically to demonstrate the performance of the device.
3. Number of experts used to establish the ground truth for the test set and qualifications of those experts:
Not applicable. This is an in vitro diagnostic (IVD) chemistry analyzer and reagent system. "Ground truth" for IVD analytical performance is established by reference methods, certified reference materials, or highly accurate comparative methods, not by human expert consensus or radiologists. The performance is assessed against quantitative values, not qualitative interpretations requiring expert review.
4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:
Not applicable. Adjudication methods like 2+1 or 3+1 are used in studies involving human interpretation (e.g., imaging studies where radiologists disagree). For analytical performance of a chemistry analyzer, the "ground truth" is typically the quantitative value obtained from a reference method or the predicate device, and differences are assessed statistically (e.g., bias, correlation).
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. MRMC studies are specific to evaluating the impact of a device on human readers' performance, typically in diagnostic imaging with AI assistance. This device is an automated chemistry analyzer, not an AI-assisted diagnostic imaging tool. There are no human "readers" in the context of this device's operation.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
Yes, in essence. The performance data provided (linearity, precision, detection limits, interference, method comparison) represents the "standalone" analytical performance of the automated chemistry system (ACE Alera with the new reagents) in measuring the target analytes in patient samples. There isn't an "algorithm only" in the AI sense, but the chemical reactions and photometric measurements are entirely automated by the device. The data shown is the raw analytical output.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
The "ground truth" for these analytical studies is primarily:
- Highly characterized samples: For linearity, samples with known, precise concentrations (often prepared by dilution of high-concentration materials or spiking low-concentration materials).
- Comparative method/Predicate device: For method comparison, the results generated by the predicate device (ACE Clinical Chemistry System) are treated as the reference or comparative method against which the new ACE Alera system's results are compared. This is a common and accepted "ground truth" for chemical analyzers seeking substantial equivalence.
- Reference materials/controls: For precision and detection limits, control materials with established target values are used.
8. The sample size for the training set:
Not applicable. This is a traditional IVD device (chemical reagents and analyzer), not an AI/ML device that requires a "training set" in the context of machine learning model development. The reagents perform chemical reactions, and the analyzer reads photometric changes; it does not "learn" from data.
9. How the ground truth for the training set was established:
Not applicable, as there is no training set in the AI/ML sense for this device.
§ 862.1440 Lactate dehydrogenase test system.
(a)
Identification. A lactate dehydrogenase test system is a device intended to measure the activity of the enzyme lactate dehydrogenase in serum. Lactate dehydrogenase measurements are used in the diagnosis and treatment of liver diseases such as acute viral hepatitis, cirrhosis, and metastatic carcinoma of the liver, cardiac diseases such as myocardial infarction, and tumors of the lung or kidneys.(b)
Classification. Class II (special controls). The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to § 862.9.