Search Results
Found 1 results
510(k) Data Aggregation
(591 days)
RAMP NT-PROBNP ASSAY
The RAMP NT-proBNP Assay is a quantitative immunochromatographic test indicated for use as an in vitro diagnostic product used to measure N-terminal pro-brain natriuretic peptide (NT-proBNP) levels in EDTA whole blood. Measurement of NT-proBNP aids in the diagnosis and assessment of severity in individuals suspected of having heart failure and may aid in the risk stratification of patients with heart failure.
The RAMP NT-proBNP Assay is a quantitative immunochromatographic test indicated for use as an in vitro diagnostic product used with a RAMP reader to measure N-terminal pro-brain natriuretic peptide (NT-proBNP) levels in EDTA whole blood. Mixed EDTA whole blood is added to the sample well of the Test Cartridge which houses the immunochromatographic test strip. The red blood cells are retained in the sample pad, and the separated plasma migrates along the strip. Fluorescent-dyed latex particles coated with anti-NT-proBNP antibodies bind to NT-proBNP, if present in the sample. As the sample migrates along the strip. NT-proBNP bound particles are captured at the detection zone, and additional particles are captured at the internal standard zone.
The RAMP reader then measures the amount of fluorescence emitted by the complexes captured at the detection zone and at the internal standard zone. Using a ratio between the two fluorescence values, a quantitative reading is calculated.
Here's a breakdown of the acceptance criteria and study information for the RAMP® NT-proBNP Assay, based on the provided 510(k) summary:
1. Acceptance Criteria and Reported Device Performance
The acceptance criteria are not explicitly stated as distinct pass/fail thresholds against which the device performance is measured. Instead, the document presents detailed performance characteristics (precision, linearity, limits, analytical specificity, interference) and then compares the clinical performance (method comparison, sensitivity/specificity, ROC analysis) against a predicate device (Roche Elecsys proBNP Assay) and using established cut-offs.
However, based on the conclusions and the comparative data, we can infer some desired performance characteristics.
| Performance Metric | Acceptance Criteria (Inferred/Implicit) | Reported Device Performance |
|---|---|---|
| Precision (Control Material) | Within Run CV: Low concentrations 22,000 ng/L reported as > 22,000 ng/L). | |
| Analytical Specificity (Cross-Rxn) | Minimal or no cross-reactivity with specified physiological compounds (Ensuring accurate measurement of target analyte) | No cross-reactivity observed with listed compounds up to maximum levels tested (e.g., ANP28, BNP32, CNP22, Angiotensins). |
| Interference (Therapeutic Drugs) | Average difference 0.95; Slope near 1.0; Intercept near 0 (Demonstrating strong agreement with a legally marketed predicate device) | Slope 0.97 (95% CI: 0.95 to 1.00) |
| Intercept 19.39 ng/L (95% CI: 14.20 to 24.67) | ||
| Correlation coefficient (R) 0.98 (95% CI: 0.97 to 0.98) | ||
| Clinical Sensitivity (Age-Stratified) | High sensitivity for CHF patients (>85%) (Ensuring accurate identification of disease presence) | Age 75: Sensitivity 0.99 (95% CI: 0.92-1.0) |
| Clinical Specificity (No Comorbidity) | High specificity for non-CHF patients without comorbidities (>70-80%) (Ensuring accurate identification of disease absence in healthy individuals) | Age 75: Specificity 0.72 (95% CI: 0.53-0.87) |
| Clinical Specificity (With Comorbidity) | Clinically acceptable specificity, acknowledging the impact of comorbidities (Understanding device performance in complex patient populations) | Age 75: Specificity 0.48 (95% CI: 0.35-0.60) (Note: The wide CI for >75 suggests variability or smaller sample size, but the values are reported) |
| ROC Analysis AUC | AUC > 0.85 and comparable to predicate device (Demonstrating good discriminatory power) | AUC 0.87 (for both RAMP and Elecsys assays) |
| Correlation with NYHA Classification | Increasing NT-proBNP levels with increasing NYHA class (Supporting use in severity assessment) | Mean NT-proBNP (ng/L) for NYHA Class I: 1686; II: 2831; III: 5737; IV: 8308 (shows clear trend of increasing mean with disease severity). This supports the claim that the system provides NT-proBNP results that correlate with severity of heart failure. |
Study Details
2. Sample sizes for the test set and data provenance
- Method Comparison Test Set (RAMP vs. Elecsys):
- Total Patients Enrolled: 699
- Samples analyzed within reportable range (34 ng/L to 22,000 ng/L): 580
- HF Diagnosed: 274 (164 males, 110 females)
- Non-HF Reference Group: 306 (124 males, 182 females)
- Data Provenance: Not explicitly stated regarding country of origin, but implied to be from clinical sites where patients were enrolled. The use of "individual hospital criteria" suggests multiple sites. The study is prospective in nature, as samples were collected "for each of these subjects" and then analyzed by both devices.
- Clinical Sensitivity and Specificity Test Set:
- Total Subjects: 858
- Diagnosed with HF: 299 (using local hospital criteria)
- Non-HF with potentially confounding co-morbidity: 189
- Reference Individuals (healthy, no co-morbidities): 370 (includes an additional 159 subjects from an additional clinical site without concomitant testing in the Elecsys system).
- Data Provenance: Implied to be from clinical sites. The collection of "data collected from 858 subjects" suggests a mix of retrospective and prospective, however the descriptions of "presenting population" and "enrolled" subjects points more towards a prospective data collection. The additional 159 subjects were "added from an additional clinical site".
3. Number of experts used to establish the ground truth for the test set and their qualifications
The ground truth for Heart Failure (HF) diagnosis was primarily established "based on individual hospital criteria" and "using local hospital criteria." This implies that the diagnosis was made by treating physicians at the respective clinical sites. The document does not specify the number of experts or their specific qualifications (e.g., "radiologist with 10 years of experience").
4. Adjudication method for the test set
The document does not specify an explicit adjudication method (e.g., 2+1, 3+1, none) for the clinical diagnosis of heart failure. The ground truth ("local hospital criteria") suggests that hospital-based clinical judgment was used.
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, an MRMC comparative effectiveness study involving human readers and AI assistance was not done. This device is a quantitative immunochromatographic test for measuring NT-proBNP levels, not an AI-assisted diagnostic imaging device or a decision support system for human readers. The comparison is between the RAMP device and a predicate laboratory device (Roche Elecsys proBNP Assay) and the device's performance metrics (sensitivity, specificity) in diagnosing HF.
6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done
Yes, a standalone performance study was clearly done. The entire "Summary of Studies" section details the performance characteristics of the RAMP NT-proBNP Assay when used without human intervention in interpreting the raw data. The RAMP reader provides a quantitative result (ng/L) of NT-proBNP levels. The clinical efficacy (sensitivity, specificity) is calculated based on these quantitative results against a clinical diagnosis (ground truth), demonstrating the algorithm's standalone performance.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The primary ground truth used for heart failure diagnosis was clinical diagnosis based on individual hospital criteria. This would typically involve a combination of patient history, physical examination, imaging (e.g., echocardiogram), and other laboratory tests as judged by the treating clinicians. The document does not mention pathology or long-term outcomes data being used as the primary ground truth for the diagnostic studies.
8. The sample size for the training set
The document does not explicitly describe a separate "training set" for the RAMP NT-proBNP Assay algorithm in the context of machine learning. For chemical assays like this, the "training" (or development and optimization) of the assay itself is done during product development using various known concentrations and interfering substances. The performance characteristics studies (precision, linearity, LoD, LoQ, analytical specificity, interference) could be considered part of the internal validation/optimization process which might indirectly involve a "training set" of samples, but these are presented as characterization studies rather than specific training data for a machine learning model. The clinical evaluation samples (699 for method comparison, 858 for sensitivity/specificity) serve as the test sets for demonstrating performance.
9. How the ground truth for the training set was established
As there is no explicitly defined "training set" in the machine learning sense, the method for establishing its ground truth is not applicable/not provided. The ground truth for the performance characterization studies (e.g., linearity, LoD) would be established by preparing samples with known, precise concentrations of NT-proBNP (spiked blood or control material where the true concentration is known).
Ask a specific question about this device
Page 1 of 1