Search Results

In vitro test for the quantitative determination of protein in urine and cerebrospinal fluid on Roche automated clinical chemistry analyzers.

Measurements obtained by this device are used in the diagnosis and treatment of a variety of diseases involving the liver, kidney or bone marrow as well as metabolic or nutritional disorders.

Protein measurements in urine are used in the diagnosis and treatment of disease conditions such as renal or heart diseases, or thyroid disorders, which are characterized by proteinuria or albuminuria.

CSF protein measurements are used in diagnosis and treatment of conditions such as meningitis, brain tumors and infections of the central nervous systems.

The Roche/Hitachi Urinary/CSF Protein reagent is an in vitro test for the quantitative determination of protein in urine and cerebrospinal fluid on Roche automated clinical chemistry analyzers.

The modified device includes both the original endpoint assay and the additional rate assay. The new rate assay was developed to provide absorbance limits that will flag high protein samples with high absorbance, thus eliminating the need for prescreening samples for high protein levels. The endpoint assay still requires sample prescreening or inspection of the Reaction Monitor display after completion of the reaction to ensure that high samples are detected and appropriately diluted for rerun. The attached labeling provides a more complete description of this potential high sample / prozone effect.

Here's an analysis of the provided 510(k) summary regarding the Roche/Hitachi Urinary/CSF Protein test, focusing on acceptance criteria and supporting study details:

1. Table of Acceptance Criteria and Reported Device Performance

The provided document is a 510(k) Special Modification submission, which focuses on comparing a modified device (with an added "rate" application) to a predicate device (the original "endpoint" application). Therefore, the "acceptance criteria" discussed are largely implicit in demonstrating substantial equivalence to the predicate, with performance characteristics being compared directly.

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

• Precision:
  • Urine:
    • Within-run: n=21 (Modified device) compared to n=120 (Predicate device).
    • Between-run: n=10 (Modified device).
  • CSF:
    • Within-run: n=20 (Modified device) compared to n=120/119 (Predicate device).
    • Between-run: n=10 (Modified device).
• Method Comparison:
  • Urine samples: n=60 (Modified device vs. Endpoint application). Concentrations between 1.7 and 3286.5 mg/dL.
  • CSF samples: n=50 (Modified device vs. Endpoint application). Concentrations between 5.8 and 110.2 mg/dL.
• Data Provenance: Not explicitly stated (e.g., country of origin, specific institution, retrospective/prospective). However, the study involves comparisons against an existing, cleared device, implying lab-based performance verification. It is implied to be retrospective analysis of performance characteristics, likely from laboratory testing.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts

This type of submission (modification to an in vitro diagnostic test kit) does not typically involve expert review for "ground truth" in the way, for example, an imaging AI device would. The "ground truth" for the performance studies (precision, method comparison) is based on the quantitative results generated by the laboratory instruments themselves, often using reference methods or the predicate device as a comparator. Therefore, this section is not applicable to this type of device.

4. Adjudication Method for the Test Set

As this is a quantitative in vitro diagnostic device, there is no "adjudication method" in the human-centric sense typically applied to image-based AI or complex diagnostic interpretations. The results are numerical values, and their accuracy is assessed against accepted laboratory standards or comparative methods. Therefore, this section is not applicable.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and Effect Size of Human Readers' Improvement with AI vs. Without AI Assistance

This is an in vitro diagnostic device for laboratory use, not an AI device for interpretation by human readers. Therefore, an MRMC study is not applicable.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done

The device itself is a reagent and an application on an automated clinical chemistry analyzer. The performance characteristics reported (precision, method comparison) represent the standalone performance of this "rate application" on the analyzer. The results are quantitative outputs from the instrument. So, in essence, the performance studies reflect the standalone performance of the modified application.

7. The Type of Ground Truth Used (Expert Consensus, Pathology, Outcomes Data, etc.)

The "ground truth" for the performance studies relied on:

• Reference Methods/Predicate Device: For method comparison, the modified rate application was compared against the original endpoint application (which was previously cleared based on comparison with the DuPont ACA method, standardized against NBS Reference Material SRM 927a using the biuret method).
• Known Concentrations: For precision studies, control materials and human samples with referenceable concentrations (though not explicitly stated whether these were clinically confirmed pathology or outcome data) were used. The controls themselves serve as a form of "ground truth" for expected values.

8. The Sample Size for the Training Set

This is a chemical reagent and an application for an automated analyzer, not a machine learning or AI model that requires a "training set" in the typical sense. The original development of the rate application would have involved optimization and testing, but it's not described as a formal "training set" like in AI. Therefore, this section is not applicable for an AI training set, but the development likely involved numerous samples for internal optimization.

9. How the Ground Truth for the Training Set Was Established

Since there is no "training set" in the AI sense, this question is not applicable. The "ground truth" for the predicate device's original clearance (K913615, which this modification refers back to) was established by standardizing against the National Bureau of Standards Reference Material SRM 927a using the biuret method for protein quantitation. The current submission demonstrates the equivalence of the modified rate application to this established methodology.

Ask a specific question about this device