Search Results

The Atellica® CH Diazo Total Bilirubin (D TBil) assay is for in vitro diagnostic use in the quantitative determination of total bilirubin in adults and children (non-neonates) in human serum and plasma using the Atellica® CH Analyzer. Measurement of total bilirubin, an organic compound formed during the normal and abnormal destruction of red blood cells, is used in the diagnosis and treatment of liver, hemolytic hematological, and metabolic disorders, including hepatitis and gall bladder block.

Atellica CH Diazo Total Bilirubin is a photometric test using 2,4-dichloroaniline (DCA). Direct bilirubin in presence of diazotized 2,4-dichloroaniline forms a red colored azocompound in acidic solution. A specific mixture of detergents enables the determination of the total bilirubin.

The provided document describes the Siemens Atellica® CH Diazo Total Bilirubin (D_TBil) assay, an in vitro diagnostic device, and its performance characteristics to demonstrate substantial equivalence to a predicate device (Dimension TBI Flex reagent cartridge).

Here's an analysis of the acceptance criteria and the study proving the device meets them, based on the provided text:

1. Table of Acceptance Criteria and Reported Device Performance

The acceptance criteria for this in-vitro diagnostic device are generally defined by demonstrating performance within established statistical limits or comparison to a predicate device, as per CLSI (Clinical and Laboratory Standards Institute) guidelines. The "acceptance criteria" themselves are not always explicitly stated as pass/fail thresholds for each performance characteristic in a simple numerical format, but rather as meeting the objectives of the study design (e.g., correlation coefficient of ≥ 0.950).

2. Sample Size Used for the Test Set and Data Provenance

• Detection Capability: Not explicitly stated as a "test set" in the context of patient samples, but the determination was in accordance with CLSI Documents EP17-A2.
• Precision:
  • Repeatability/Within-Lab Precision: 80 measurements for each of the three serum levels (Serum 1, 2, 3), assayed in duplicate over 20 days.
  • Reproducibility: 225 measurements for each of the three serum levels (Sample ID 1, 2, 3), assayed with 5 replicates per run for 5 days using 3 instruments/sites and 3 reagent lots.
• Assay Comparison (Method Comparison): N = 103 patient samples (Serum)
• Specimen Equivalency: N = 57 patient samples for each plasma type (Lithium Heparin, Sodium Heparin, K2 EDTA) compared to serum.
• Interferences (HIL and Non-interfering Substances): The studies were performed with specific interferent concentrations and two bilirubin levels. The sample size refers to the number of spiked samples tested, but not explicitly the number of unique patient samples that may have been used to create these spiked materials.
• Data Provenance: The document does not specify the country of origin for the samples or whether the studies were retrospective or prospective. Given the nature of in-vitro diagnostic development, these are typically prospective studies using a mix of spiked and potentially remnant clinical samples.

3. Number of Experts Used to Establish Ground Truth and Qualifications

This information is not applicable and therefore not provided in the document. For an in-vitro diagnostic assay that quantitatively measures an analyte (total bilirubin), the "ground truth" is established by the reference method or comparison method (the predicate device in this case) and the inherent accuracy and traceability of the calibrators, rather than by human expert review of images or clinical cases. The ground truth for quantitative assays typically relies on metrological traceability to certified reference materials (NIST Standard Reference Material 916 in this case) and established laboratory protocols.

4. Adjudication Method for the Test Set

This is not applicable. Adjudication methods (like 2+1, 3+1) are common in studies involving human interpretation of medical images or clinical data, especially for AI applications where consensus among experts establishes the ground truth. For quantitative in-vitro diagnostics, the "ground truth" is determined by established analytical methods and reference standards, not by human adjudication of qualitative results.

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

This is not applicable. An MRMC study is designed to assess the performance of a diagnostic system that involves human readers (e.g., radiologists interpreting images) and typically compares the effectiveness of human readers with and without AI assistance. The Atellica CH Diazo Total Bilirubin assay is an automated in-vitro diagnostic test, meaning it does not involve human readers for interpretation beyond the initial sample collection and analysis setup. It is a standalone analytical device.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study

This is essentially what the entire submission describes. The performance characteristics presented (Detection Capability, Precision, Reproducibility, Assay Comparison, Specimen Equivalency, Interference) are all standalone performance studies of the medical device itself. There is no human-in-the-loop component for the interpretation or output of the assay. The device directly measures and reports total bilirubin levels.

7. Type of Ground Truth Used

The ground truth for this quantitative in-vitro diagnostic assay is established through:

• Metrological Traceability: The assay is traceable to the NIST Standard Reference Material 916. This provides a fundamental basis for accuracy.
• Comparison to a Predicate Device: The performance of the new device is compared against a legally marketed predicate device (Dimension TBI assay), which serves as the "truth" or reference in method comparison studies.
• Controlled Samples: Precision, reproducibility, and interference studies use manufactured control materials, internal validations, and spiked samples with known concentrations or interferent levels, whose 'truth' values are established through rigorous analytical methods and often validated against reference methods.

8. The Sample Size for the Training Set

This information is not provided in the document, and it's generally not applicable in the typical sense of "training set" for traditional in-vitro diagnostic devices. These devices are developed, validated, and optimized through a series of analytical studies using various types of samples (e.g., precision materials, linearity samples, spiked samples, patient samples). The "training" isn't a machine learning training phase, but rather the process of optimizing the reagent formulation, reaction kinetics, and instrument parameters. The data used for these optimizations would be proprietary and extensive, but not typically referred to as a "training set" in regulatory submissions for IVDs. The analytical performance studies (like those detailed above) serve as the validation and verification of the final, optimized product.

9. How the Ground Truth for the Training Set Was Established
As explained in point 8, the concept of a "training set" with a defined "ground truth" for a traditional IVD like this is generally not applicable in the same way as for AI/ML devices. The "ground truth" for the development and optimization of such an assay would be established through:

• Reference Methods: Using established, highly accurate reference methods to determine the true concentration of bilirubin in control materials and patient samples used during development.
• Certified Reference Materials: Calibrating and verifying the assay against NIST or other internationally recognized certified reference materials.
• Known Spiking: Creating samples with precisely known additions of bilirubin or interferents to test linearity, recovery, and interference effects.

In summary, the provided document details a comprehensive set of analytical studies to demonstrate the performance and substantial equivalence of the Atellica® CH Diazo Total Bilirubin assay as a standalone in-vitro diagnostic device. The evaluation follows a different paradigm than AI/ML algorithms, thus many questions related to expert review, adjudication, and MRMC studies are not relevant.

Ask a specific question about this device