(137 days)
The Diazyme LDL-Cholesterol Assay is intended for the in vitro quantitative determination of Low Density Lipoprotein Cholesterol in human serum or plasma. The reagents can assist in the diagnosis and treatment of patients at risk of developing coronary heart disease. Elevated LDL cholesterol is the primary target of cholesterol-lowering therapy.
The assay is based on a modified polyvinyl sulfonic acid (PVS) and polyethylene-glycol methyl ether (PEGME) coupled classic precipitation method with the improvements in using optimized quantities of PVS/PEGME and selected detergents. LDL, VLDL. and chylomicron (CM) react with PVS and PEGME and the reaction results in inaccessibility of LDL, VLDL and CM by cholesterol oxidase (CHOD) and cholesterol esterase (CHER), whereas HDL reacts with the enzymes. Addition of R2 containing a specific detergent releases LDL from the PVS/PEGME complex. The released LDL reacts with the enzymes to produce H2O2 which is quantified by the Trinder reaction.
Here's an analysis of the acceptance criteria and study details for the Diazyme LDL-Cholesterol Reagent, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
The submission does not explicitly state "acceptance criteria" as distinct, quantifiable thresholds that the device must meet on its own for approval. Instead, the performance of the Diazyme LDL-Cholesterol Reagent is presented and compared primarily to a legally marketed predicate device (Genzyme N-Geneous LDL Cholesterol Reagent, K971573) to demonstrate substantial equivalence. The implication is that performance comparable to the predicate (and within generally accepted clinical laboratory standards for such assays) constitutes the acceptance.
Therefore, the table below reflects the reported performance of the Diazyme device and, for comparison, the predicate device as presented in the submission. The "acceptance criteria" are inferred as comparable performance to the predicate and generally good laboratory practice for quantitative assays.
| Performance Metric | Diazyme LDL-Cholesterol Reagent Reported Performance | Inferred Acceptance Criteria (based on predicate & good lab practice) |
|---|---|---|
| Reportable Range | Serum: 2.04 - 250 mg/dL | Comparable to predicate (6.6-992 mg/dL) and clinically relevant. Diazyme's range is narrower on the high end but acceptable. |
| Linearity | 1.64 - 830 mg/dL (for serum samples) | Sufficiently wide to cover clinically relevant range. |
| Precision (Within Run) | 0.7% - 1.0% ($C_v$%) at different LDL levels ($95, 146, 210 mg/dL$) | Comparable to predicate (0.62% - 0.73%) and within expected analytic variability for clinical assays. |
| Precision (Total) | 1.4% - 1.6% ($C_v$%) at different LDL levels ($95, 146, 210 mg/dL$) | Comparable to predicate (1.73% - 2.27%) and within expected analytic variability for clinical assays. |
| Accuracy (Correlation Coefficient vs. Predicate) | 0.996 (for serum samples) | High correlation (e.g., > 0.95) with a legally marketed device. |
| Accuracy (Slope/Intercept vs. Predicate) | y = 1.0883x + 0.6078 mg/dL | Slope close to 1 and intercept close to 0 when compared to a legally marketed device. (Predicate: y = 0.95x + 3.02mg/dL) |
| Interference | Less than 10% interference with: | Minimal interference from common endogenous substances at specified concentrations. |
| - Triglycerides (1000 mg/dL) | ||
| - Ascorbic acid (10 mmol/L) | ||
| - Bilirubin (40 mg/dL) | ||
| - Bilirubin Conjugated (40 mg/dL) | ||
| - Hemoglobin (1000 mg/dL) |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size: The document repeatedly mentions "samples tested with Diazyme LDL-Cholesterol Reagent showed good correlation with Genzyme N-geneous LDL Cholesterol Reagent (K971573) with correlation coefficients of 0.996 for serum samples" and refers to "clinical patient samples." However, the exact number of samples used for the method comparison (test set) is not explicitly stated in the provided text.
- Data Provenance: The text does not specify the country of origin of the data. It is implied that the data is retrospective, as it involves "samples tested" and a comparison was made using these samples. The term "clinical patient samples" also suggests that these were real-world samples collected from patients.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
This information is not provided in the document. For an in vitro diagnostic reagent like this, the "ground truth" for the test set is established by testing the same samples using a well-established, often reference, method or a legally marketed predicate device. The expertise lies in the laboratory performing the tests and ensuring proper execution, not typically in a panel of experts reviewing the results in the same way as, for example, image-based diagnostics.
In this case, the predicate device (Genzyme N-geneous LDL Cholesterol Reagent) serves as the comparator or "reference" for demonstrating substantial equivalence. The ground truth for the predicate device itself is stated to be traceable to the CDC HDL reference method, which is a highly standardized and rigorous method.
4. Adjudication Method for the Test Set
This is not applicable in the context of this type of quantitative diagnostic assay. Adjudication methods (like 2+1 or 3+1) are typically used in studies involving subjective interpretation, such as image analysis where multiple readers might disagree, and a consensus needs to be formed for the ground truth. Here, the "truth" is a quantitative value obtained from a chemical reaction. The comparison is statistical, between two quantitative results.
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, a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was not done. This type of study is relevant for AI-powered diagnostic devices, especially in imaging, where human readers interpret results with and without AI assistance. The Diazyme LDL-Cholesterol Reagent is a laboratory reagent for quantitative chemical analysis, not an AI-assisted diagnostic tool that aids human readers.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
The entire performance evaluation described for the Diazyme LDL-Cholesterol Reagent is a standalone (algorithm/reagent only) performance. The device is a reagent system that produces a quantitative result. There is no "human-in-the-loop" in the interpretation of the final quantitative value (though humans operate the instruments and handle the samples). The accuracy and precision data presented are reflective of the reagent's performance on its own.
7. The Type of Ground Truth Used
The ground truth used for comparison (and thus, implicitly, for evaluating the Diazyme device) is derived from the results obtained from a legally marketed predicate device (Genzyme N-geneous LDL Cholesterol Reagent, K971573). The submission states: "Genzyme N-geneous LDL Cholesterol Reagent (K971573) was selected for comparing serum samples with to the results generated by Diazyme LDL-Cholesterol Reagent." The predicate device itself traces its calibrator to the NIST SRM 1915b and its method to the CDC HDL reference method, indicating a robust underlying ground truth reference.
8. The Sample Size for the Training Set
This information is not applicable and therefore not provided in the document. The Diazyme LDL-Cholesterol Reagent is a chemical reagent assay, not an AI/machine learning model that requires a "training set." Its operating parameters (e.g., reagent concentrations, reaction times) would be optimized during its development process, but this is a different concept from training an algorithm.
9. How the Ground Truth for the Training Set Was Established
This information is not applicable as there is no "training set" in the context of this chemical reagent device.
§ 862.1475 Lipoprotein test system.
(a)
Identification. A lipoprotein test system is a device intended to measure lipoprotein in serum and plasma. Lipoprotein measurements are used in the diagnosis and treatment of lipid disorders (such as diabetes mellitus), atherosclerosis, and various liver and renal diseases.(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.