Search Results
Found 1 results
510(k) Data Aggregation
(211 days)
ORTHO™ ImmunoCount Flow Cytometry System is intended to be used for lymphocyte immunophenotyping. Results may be reported as either percent positive cells, or as absolute counts of lymphocytes and lymphocyte subsets.
ORTHO ImmunoCount Flow Cytometry System consists of ORTHO CytoronAbsolute Laser Flow Cytometer, ImmunoCount II Software, ORTHO TRIO Monoclonal Antibodies for lymphocyte immunophenotyping, and Ortho-Count Calibration Kit for calibration and verification of system calibration.
Here's an analysis of the provided text, focusing on the acceptance criteria and study details for the ORTHO ImmunoCount Flow Cytometry System:
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria are implied by the comparison to a "predicate method" and the demonstration of linearity and reproducibility. The tables provided (TABLE A, B, C, D, E, F, G, H) present the performance data, which implicitly serves as the "reported device performance." Since explicit numerical acceptance criteria (e.g., "R-value > 0.95") are not stated, I will infer them from the data presented, particularly the R-values, slopes, and intercepts for comparability, and CVs for reproducibility. For comparability, "equivalent" means the values are close to the predicate. For reproducibility, "low CV" is the criterion. For linearity, an R-value close to 1.00 is the criterion.
| Acceptance Criteria Category | Specific Metric | Implied Acceptance Criterion (from Predicate/Reproducibility) | Reported Device Performance (ORTHO ImmunoCount) | Met/Not Met (Based on typical expectations for medical devices) |
|---|---|---|---|---|
| Comparability (Predicate Method) | ||||
| Percent Positive Cells (Normal Donors) | Mean % difference vs. Predicate (for various subsets like CD3+CD4+, CD16+CD3-, CD19+, Mean CD3+) | Mean % difference should be minimal, suggesting equivalence to the predicate. Range overlap. | See TABLE A. Differences in mean % are generally small (e.g., CD3+CD4+ ImmunoCount 45.5% vs. Predicate 45.2%; CD16+CD3- ImmunoCount 11.6% vs. Predicate 13.1%). Ranges largely overlap. | Met |
| Absolute Counts (Normal Donors) | Mean (cells/µL) difference vs. Predicate (for various subsets) | Mean (cells/µL) difference should be minimal, suggesting equivalence to the predicate. Range overlap. | See TABLE A. Differences in mean (cells/µL) are generally small (e.g., CD3+CD4+ ImmunoCount 855 vs. Predicate 900; CD16+CD3- ImmunoCount 221 vs. Predicate 266). Ranges largely overlap. | Met |
| Percent Positive Cells (HIV+ Donors) | Mean % difference vs. Predicate (for various subsets) | Mean % difference should be minimal, suggesting equivalence to the predicate. Range overlap. | See TABLE B. Differences in mean % are generally small (e.g., CD3+CD4+ ImmunoCount 17.7% vs. Predicate 17.6%; CD16+CD3- ImmunoCount 8.8% vs. Predicate 10.6%). Ranges largely overlap. | Met |
| Absolute Counts (HIV+ Donors) | Mean (cells/µL) difference vs. Predicate (for various subsets) | Mean (cells/µL) difference should be minimal, suggesting equivalence to the predicate. Range overlap. | See TABLE B. Differences in mean (cells/µL) are generally small (e.g., CD3+CD4+ ImmunoCount 260 vs. Predicate 313; CD16+CD3- ImmunoCount 93 vs. Predicate 144). Ranges largely overlap. | Met |
| Linear Regression (Normal Donors - Percent) | R-value between ImmunoCount and Predicate (for % positive cells) | R-value > 0.85 (typical for good correlation), Slope close to 1, Intercept close to 0. | See TABLE C. R-values range from 0.83 (CD3+ (4/8/3) vs CD3+ (3/4)) to 1.00 (CD4+ (4/8/3) vs CD4+ (3/4)). Slopes are generally close to 1 (0.87-1.01), intercepts close to 0 (0-12). | Met |
| Linear Regression (Normal + HIV+ Abs Counts) | R-value between ImmunoCount and Predicate (for absolute counts) | R-value > 0.85, Slope close to 1, Intercept close to 0. | See TABLE D. R-values range from 0.88 (CD3- (16/19/3) vs CD3+ (3/16+56)) to 0.96 (All lymphocyte subsets). Slopes are generally close to 1 (0.76-0.89), intercepts are slightly higher than 0 (3-124). | Met |
| Reproducibility | ||||
| Within-Laboratory Reproducibility | Coefficient of Variation (CV) for absolute lymphocyte counts (Low, Normal, High concentrations) | CV < 10% for most analytes, with slightly higher allowances for low count populations or those with inherent biological variability (e.g., CD16+, CD19+ subsets). The conclusion states "low CV values demonstrate strong within-laboratory reproducibility". | See TABLE E. CVs range from 4.3% to 16.3%. CD16+CD3- and CD19+ have higher CVs at low/normal counts (10.0-16.3%), but the report justifies this due to low counts in these populations, and notes it "do[es] not significantly affect the ImmunoSum (ISUM) CV" (which are generally low, 4.2-5.0%). | Met (with stated justifications) |
| Between-Laboratory Reproducibility | Coefficient of Variation (CV) for absolute lymphocyte counts (Low, Normal, High concentrations) across different labs | CV < 15% (or similar, depending on analyte/level), with slightly higher allowances for low count populations. The conclusion states "low CV values demonstrate acceptable between-laboratory reproducibility". | See TABLE F. CVs range from 0.8% to 13.1%. Similar to within-lab, CD16+CD3- and CD19+ show higher CVs (8.3-13.1%), also justified by low counts in these populations and negligible impact on ISUM CV (1.6-3.0%). | Met (with stated justifications) |
| Specimen Age Criteria | Mean Slope of absolute counts over 72 hours and associated Confidence Interval (CI) | Mean Slope close to 0 and CI indicating no significant change over time (e.g., within a predefined tolerance for change). "Comparable results" up to 72 hours. | See TABLE G. Mean slopes are very close to 0 (e.g., -1.5 to 0.8), and CIs are small (1.0-2.6). This indicates stability of results across 24, 48, and 72 hours. | Met |
| Linearity | R-value between observed counts and expected (diluted/concentrated) counts | R-value close to 1.00 across a wide range of cell counts. "Excellent linearity." | See TABLE H. R-values are 0.99 or 1.00 for all tested subsets across significant ranges (e.g., CD3+(4/8/3) 75 - 9829 cells/µL; ISUM Lymph Count 95 - 14694 cells/µL). | Met |
The study demonstrates that the ORTHO ImmunoCount Flow Cytometry System meets its acceptance criteria by showing substantial equivalence to the predicate method for both normal and HIV-positive donors in percent positive cells and absolute counts. Furthermore, it exhibits excellent within- and between-laboratory reproducibility, stability over a 72-hour period post-collection, and strong linearity across a broad range of cell counts.
2. Sample Size and Data Provenance
-
Comparability Study (Normal Donors):
- Test Set Size (Percent Positive): 200 normal whole blood specimens.
- Test Set Size (Absolute Counts): 151 normal whole blood specimens.
- Data Provenance: Retrospective, collected from apparently healthy donors (44.5% male, 55.5% female, age 21-62 years) in the United States. Three geographically distinct clinical sites were used for absolute counts, and four for percent positive cells.
-
Comparability Study (HIV-Antibody Positive Donors):
- Test Set Size (Percent Positive): 119 HIV-antibody positive whole blood specimens.
- Test Set Size (Absolute Counts): 89 HIV-antibody positive whole blood specimens.
- Data Provenance: Retrospective, collected from HIV-antibody positive donors (88.1% male, 11.9% female, age 3- years; likely meaning 3 years and above, or the data has a typo for the lower bound) in the United States. Three geographically distinct clinical sites were used for absolute counts, and four for percent positive cells.
-
Within-Laboratory Reproducibility: 5 normal whole blood specimens, tested in replicates of 10 at three independent laboratories.
-
Between-Laboratory Reproducibility: 5 normal whole blood specimens, compared between three independent laboratories, tested in replicates of 10.
-
Specimen Age Criteria: 64 whole blood specimens (34 normal, 30 HIV-antibody positive), tested at three external clinical sites.
-
Linearity: 4 normal whole blood specimens, tested at Ortho Diagnostic Systems.
3. Number of Experts and Qualifications (for Ground Truth)
The documentation does not explicitly state the number or qualifications of "experts" used to establish the ground truth for the test set. Instead, the ground truth is established by the "predicate method," which consists of:
- FACScan™ Flow Cytometer with Simulset Software (Becton Dickinson)
- Simultest™ Monoclonal Antibody Immunophenotyping Reagents (Becton Dickinson)
- Coulter® JT2 Automated Hematology Analyzer (Coulter Electronics)
- Coulter T Series Automated Hematology Analyzer (Coulter Electronics)
- Argos Hematology Analyzer (ABX France)
These are established, FDA-cleared (or equivalent at the time) medical devices used in clinical laboratories. The "experts" in this context would be the trained laboratory personnel operating these predicate devices according to their standard operating procedures, rather than individual "expert adjudicators" in the sense of image interpretation.
4. Adjudication Method for the Test Set
No explicit adjudication method (like 2+1 or 3+1 consensus) is described. The "ground truth" is derived directly from the measurements of the predicate devices. The implicit "adjudication" is inherent in the established and validated performance of these predicate devices and the standard laboratory protocols used for their operation. Differences between the ImmunoCount system and the predicate are then analyzed statistically (e.g., linear regression, mean differences).
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No MRMC comparative effectiveness study was done. This device is a flow cytometer system for quantitative measurements, not an AI-assisted diagnostic imaging device requiring human-in-the-loop performance evaluation. Therefore, the concept of "how much human readers improve with AI vs without AI assistance" is not applicable here.
6. Standalone Performance
Yes, a standalone performance evaluation was done in the sense that the ORTHO ImmunoCount Flow Cytometry System's measurements were directly compared against the established predicate methods, as well as evaluated for its own reproducibility, specimen age stability, and linearity. It provides a direct measurement, and its results are compared to the predicate device's direct measurements. There isn't a "human-in-the-loop" component for its primary intended use, making its performance inherently standalone in its measurement function.
7. Type of Ground Truth Used
The ground truth used is based on measurements from established, predicate medical devices and reagents (flow cytometers, immunophenotyping reagents, and hematology analyzers) which represent the "traditional method" for lymphocyte immunophenotyping and absolute count determination. This can be considered a form of reference standard measurement from existing, accepted clinical methodology.
8. Sample Size for the Training Set
The document does not specify a separate "training set" or its size. This type of submission (a 510(k) for an instrument and reagent system) typically focuses on verification and validation studies using clinical samples, rather than machine learning model training. The development (calibration, optimization) of the ImmunoCount system itself would have involved internal testing and calibration, but a distinct "training set" as understood in AI/ML is not mentioned or implied.
9. How the Ground Truth for the Training Set was Established
As no explicit training set is mentioned in the context of AI/ML, this question is not directly applicable. If there were internal development/calibration data used, the ground truth would likely have been established similar to the predicate method for the validation studies, by running samples on established instruments to set targets for the new system's development.
Ask a specific question about this device
Page 1 of 1