Search Results

The ADVIA Centaur® NT-proBNPII (PBNPII) assay is for in vitro diagnostic use in the quantitative determination of N-terminal pro-brain natriuretic peptide (NT-proBNP) in human serum and plasma (EDTA and lithium heparin) using the ADVIA Centaur® XP system.

In the Emergency Department (ED) and Outpatient (OP) populations, measurements of NT-proBNP are used as an aid in the diagnosis of heart failure (HF) in patients with clinical suspicion of new onset or worsening HF.

The ADVIA Centaur® NT-proBNPII (PBNPII) assay kit includes the Primary Reagent ReadyPack and the Calibrator. The Primary Reagent ReadyPack contains Lite Reagent, Solid Phase Reagent, and Ancillary Well Reagent. The Calibrator includes Low and High Calibrators which are lyophilized.

The provided document discusses the ADVIA Centaur® NT-proBNPII (PBNPII) assay, an in vitro diagnostic device for measuring N-terminal pro-brain natriuretic peptide (NT-proBNP) to aid in the diagnosis of heart failure in Emergency Department (ED) and Outpatient (OP) populations. The submission aims to demonstrate substantial equivalence to the predicate device, the Roche Elecsys proBNP II assay (K072437).

Here's an analysis of the acceptance criteria and the studies that support them:

1. Table of Acceptance Criteria and Reported Device Performance

The document doesn't explicitly state "acceptance criteria" as a single, consolidated table with pass/fail values for all performance characteristics. Instead, it presents various performance studies with their results. Based on the "Comparison of Technological Characteristics with the Predicate Device" and the "Performance Characteristics" sections, we can infer some criteria and compare the device's performance.

Note: The "Acceptable Criteria" values for precision were primarily taken directly from comments in the tables (indicated by "<" followed by a number), which appear to be the applicant's internal criteria rather than directly from the predicate. The predicate's precision data is shown for comparison purposes in the "Comparison of Technological Characteristics."

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

• Precision and Reproducibility:
  • Precision: Assayed in duplicate in 2 runs per day for 20 days. Number of samples not explicitly stated but implied to be multiple levels of serum and QCs (6 serum, 3 QC samples listed).
  • Reproducibility: N=90 samples assayed in duplicate in 2 runs per day for 5 days at 3 sites.
• Linearity: Not specified.
• Detection Limit (LoB, LoD, LoQ): Not specified.
• Interference (HIL & Other Substances): Not specified.
• Cross-Reactivity: Not specified.
• Specimen Equivalency: N=50 samples for each tube type comparison to serum.
• Expected Values (Reference Study Group): 723 apparently healthy subjects (362 females, 361 males).
• Clinical Performance (ED Population): 3128 subjects with signs and symptoms of acute HF who presented to the ED. 1148 adjudicated as acute HF, 1980 as without HF.
• Clinical Performance (OP Population): 1033 OP subjects with signs and symptoms of new onset HF. 185 adjudicated as new onset HF, 848 as without HF.

Data Provenance: The document does not explicitly state the country of origin for the clinical study data or if it was retrospective or prospective. However, the term "prospectively enrolled" is used for both the ED and OP clinical performance studies, indicating these were prospective studies.

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

• Clinical Performance (ED Population): An "independent central adjudication panel of expert clinicians (Cardiologists)" determined the diagnosis and severity of HF. The exact number of experts is not stated, nor are their specific qualifications (e.g., number of years of experience).
• Clinical Performance (OP Population): An "independent central adjudication panel of expert clinicians (Cardiologists)" determined the diagnosis of new onset HF. Similar to the ED population, the exact number and specific qualifications are not detailed.

4. Adjudication Method (for the test set)

The adjudication method used by the "independent central adjudication panel of expert clinicians (Cardiologists)" is not explicitly described (e.g., 2+1, 3+1, none). It only states they "determined" the diagnosis.

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

No MRMC comparative effectiveness study was done. This device is an in vitro diagnostic assay (laboratory test), not an imaging or interpretive device that typically involves human readers in the same way an MRMC study would apply. Therefore, there is no effect size reported for human readers improving with or without AI assistance.

6. Standalone Performance Study

Yes, a standalone (algorithm only) performance study was performed for the assay. The entire "Performance Characteristics" section (9.1 through 9.12) describes the analytical and clinical performance of the ADVIA Centaur PBNPII assay in isolation, without human interpretation of the assay results in the context of the study design. The clinical utility is assessed by how accurately the assay's quantitative results correlate with the adjudicated clinical diagnosis.

7. Type of Ground Truth Used

• Clinical Performance (ED and OP Populations): Clinical diagnosis adjudicated by an "independent central adjudication panel of expert clinicians (Cardiologists)." This falls under expert consensus clinical diagnosis.
• Analytical Performance (Precision, Linearity, Detection Limits, Interference, Cross-Reactivity, Specimen Equivalency): These are quantitative measurements against established reference methods, standards, or known concentrations, serving as the ground truth for analytical validity.

8. Sample Size for the Training Set

The document describes an in vitro diagnostic assay for measuring biomarkers. These types of assays typically do not have a "training set" in the same sense as machine learning algorithms, which require labeled data for model development. Instead, they involve method development and validation experiments. The document describes studies for:

• Establishing expected values (reference study group, N=723).
• Establishing disease study groups for statistical analysis of correlation (ED population N=1148, OP population N=185).
• Various analytical validation studies such as precision, linearity, and detection limits.

Therefore, the concept of a "training set" for the assay, in the context of algorithm development, is not directly applicable or discussed. The studies mentioned above contribute to the overall validation and characterization of the assay's performance.

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

As mentioned above, the concept of a "training set" for an assay like this is not directly applicable. If we consider the data used to characterize the assay's performance as analogous to "training data" (i.e., the data used to define the assay's characteristics and clinical cut-offs), then:

• Clinical Cut-offs: The clinical cut-offs for the ED and OP populations were established based on the measured NT-proBNP values in patient populations (ED and OP disease study groups) where the diagnosis of heart failure was determined by an "independent central adjudication panel of expert clinicians (Cardiologists)." This expert consensus clinical diagnosis serves as the ground truth for establishing the clinical utility of the various cut-off values.
• Reference Intervals: The reference interval was established using 723 apparently healthy subjects. The "healthy" status would have been determined through standard medical screening and criteria, serving as the ground truth for defining the normal range.
• Analytical Studies: Ground truth for analytical studies (e.g., linearity, detection limit) is established through reference materials, spiked samples with known concentrations, and comparison to established methods or gold standard instrumentation.

Ask a specific question about this device