Search Results

The STA® - Liatest® D-Di kit is an immuno-turbidimetric assay for the quantitative determination of D-dimer in venous plasma (3.2% sodium citrate) for use on STA-R®, STA Compact® and STA Satellite® analyzers by professional laboratory personnel. The STA® - Liates® D-Di is intended for use in conjunction with a clinical pretest probability (PTP) assessment model to exclude pulmonary embolism (PE) and deep venous thrombosis (DVT) in outpatients suspected of PE or DVT.

STA® - Liatest® D-Di kit contains: 6 x 5-ml vials of ready-for-use Tris buffer and 6 x 6-ml vials of a suspension of microlatex particles coated with two different mouse monoclonal anti-human D-dimer antibodies (8D2 and 2.1.16) stabilized with bovine albumin.

The test principle is based on the change in turbidity of a microparticle suspension that is measured by photometry. A suspension of latex microparticles, coated by covalent bonding with monoclonal antibodies specific for D-dimer is mixed with the test plasma for which the D-dimer level is to be assayed. An antigen-antibody reaction takes place, leading to an agglutination of the latex microparticles which causes an increase in turbidity of the reaction medium. This increase in turbidity is reflected by an increase in absorbance, the latter being measured photometrically. The increase in absorbance is a function of the D-dimer level present in the test sample.

This document describes the acceptance criteria and study proving the performance of the STA® - Liatest® D-Di kit, an immuno-turbidimetric assay for the quantitative determination of D-dimer.

1. Acceptance Criteria and Reported Device Performance

The acceptance criteria for the STA® - Liatest® D-Di kit were based on achieving specific Negative Predictive Value (NPV) and Sensitivity targets for the exclusion of Pulmonary Embolism (PE) and Deep Venous Thrombosis (DVT) in outpatients with low to moderate pretest probability (PTP).

Note: The document explicitly states that the results "meet both confirmatory hypotheses relating to NPV and Sensitivity." While the predicate device's PE and DVT results are listed in the "Similarities Chart," the detailed performance data for the new device (STA® - Liatest® D-Di) in the clinical performance section is specifically for DVT exclusion. The indications for use state "to exclude pulmonary embolism (PE) and deep venous thrombosis (DVT)", but the detailed study results presented focus on DVT.

2. Sample Size and Data Provenance

• Test Set Sample Size: 980 samples of patients with a low or moderate PTP were included in the primary efficacy analyses. This consisted of 79 suspects of DVT and PE, and 901 suspects of DVT only. Of the 980 samples, 85 were DVT positive and 895 were DVT negative.
• Data Provenance: The data was collected from a prospective, multi-center clinical study conducted at 16 sites across the United States, Europe, and Canada. The patients were consecutive, ambulatory outpatients presenting at emergency units or outpatient clinics suspected of having venous thromboembolism (VTE).

3. Number of Experts and Qualifications for Ground Truth

The document does not explicitly state the "number of experts" or their "qualifications" in the context of establishing ground truth in the traditional sense of human readers adjudicating medical images. Instead, the ground truth was established through a clinical protocol:

• Patients with positive D-dimer results were considered for an imaging procedure (e.g., ultrasound, CT angiography).
• Patients with negative D-dimer results were assigned to a three-month follow-up to confirm the absence of DVT/PE.
• The "Reference" column in the results tables (Tables 1, 2) is a combination of "imaging or 3-month follow-up."

Therefore, the establishment of ground truth was based on clinical diagnostic pathways and follow-up, rather than expert interpretation of a specific dataset for the purpose of algorithm validation in the way one might see for an imaging AI. The "experts" in this context would be the clinicians and radiologists involved in the standard of care diagnostic work-up.

4. Adjudication Method for the Test Set

Not applicable in the typical sense of expert adjudication of AI outputs. The ground truth was established by standard clinical practice: a positive D-dimer result led to imaging, and a negative D-dimer result led to a 3-month clinical follow-up for confirmation.

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

No MRMC study was done, as this is an in-vitro diagnostic (D-dimer assay), not an imaging AI diagnostic device intended to assist human readers. The study evaluates the standalone performance of the assay in a clinical pathway.

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

Yes, this study represents a standalone (algorithm only) performance evaluation. The STA® - Liatest® D-Di kit itself is the "algorithm" or diagnostic tool, and its performance (sensitivity and NPV) is reported based on its direct results in conjunction with a clinical pretest probability assessment model. Its output (D-dimer level) is then used to guide clinical decisions (imaging vs. observation).

7. Type of Ground Truth Used

The ground truth was established by clinical outcomes data and definitive diagnostic imaging (e.g., ultrasound for DVT, CT angiography for PE) or clinical follow-up confirming the absence of the condition. Specifically, the "Reference" for the test set was determined by "imaging or 3-month follow-up."

8. Sample Size for the Training Set

The document does not specify a separate "training set" or its sample size. This is common for traditional in-vitro diagnostic assays, where product development and analytical validation are often not described in terms of "training sets" like machine learning models. The study described is primarily a clinical validation study demonstrating performance on a test set.

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

As no specific "training set" (in the context of machine learning) is discussed for this in-vitro diagnostic device, this point is not applicable. The development and analytical validation of such assays typically involve laboratory studies (e.g., linearity, precision, interference) and internal developmental studies that establish the assay's performance characteristics, rather than a "ground truth" derived from a clinical dataset used for training, as would be the case for an AI/ML device.

Ask a specific question about this device