(340 days)
The GDS® HemoSite® Test System is a device for in vitro diagnostic use only. The GDS® HemoSite® Test Card is intended for the quantitative determination of hemoglobin in whole blood. It is intended for use with the Stat-Site® Meter.
The HemoSite® Test Cards are dry-chemistry Test Cards that contain chemicals which react and produce color when hemoglobin is present. The Stat-Site® Meter is a reflectance colorimeter that reads color intensity. A Test Module is provided with each box of Test Cards. The Test-Module contains all current calibration information for the specific lot of reagent. The Test Module and Test Cards are coded to each other, and the Stat-Site® Meter will only operate if they are properly matched. HemoSite® Test Cards are dry reagent cards based on the azidemethemoglobin method. When a drop of whole blood is applied to the top of the Test Card, hemolysis occurs, with the release of hemoglobin. Sodium nitrite converts the hemoglobin to methemoglobin. Sodium azide converts methemoglobin to azidemethemoglobin to produce a brown color that is measured at 580 nm using the Stat-Site® Meter. The Stat-Site® Meter measures the color intensity and calculates the hemoglobin concentration.
Here's an analysis of the provided text regarding the HemoSite® Test System, addressing your specific questions:
1. Table of Acceptance Criteria and Reported Device Performance
The submission does not explicitly state pre-defined "acceptance criteria" in terms of specific thresholds for the regression analysis or precision. Instead, it presents the results of a comparison study against a predicate device and concludes that the results "compare well." Based on the provided data, we can infer the performance metrics reported.
| Acceptance Criteria (Inferred) | Reported Device Performance |
|---|---|
| Comparison to Predicate Device (Coulter® JT) | |
| - Strong Linear Correlation Coefficient (r) for Venous Blood | Y = 0.90x + 1.3, r = 0.8878 |
| - Strong Linear Correlation Coefficient (r) for Capillary Blood | Y = 1.0x - 0.18, r = 0.8358 |
| Precision (Whole Blood Hemoglobin) | |
| - Low %CV for Within-Run Precision (all hemoglobin levels) | Below 4 %CV |
| - Low %CV for Total Precision at High Hemoglobin Levels | 3.0-3.5 %CV |
| - Low %CV for Total Precision at Normal Hemoglobin Levels | 5.7-5.9 %CV |
| - Low %CV for Total Precision at Low Hemoglobin Levels | 5.9-7.2 %CV |
| Qualitative Conclusion: Device compares well to predicate. | The data demonstrates that blood hemoglobin results from the Stat-Site®/HemoSite® Test System... compare well to those of laboratory instruments such as the Coulter® JT when determining blood hemoglobin concentrations. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for the Test Set: A total of 212 venous and capillary blood samples were used.
- Data Provenance: The samples were obtained internally and at three different physician's office sites. This suggests a blend of controlled laboratory and real-world point-of-care settings. The document does not specify the country of origin, but given the submission is to the FDA, it is highly probable the data is from the United States. The study appears to be prospective in nature, as it describes obtaining samples to compare the new device against a reference method.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
This type of information is not applicable to this submission. The ground truth for this device (hemoglobin concentration) is established by a quantitative measurement from a predicate laboratory instrument, not by expert interpretation.
4. Adjudication Method for the Test Set
This type of information is not applicable to this submission. Adjudication methods are typically used when subjective interpretations (e.g., image readings) require consensus among experts. Here, the comparison is direct quantitative measurement against a reference method.
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
This type of study is not applicable to the HemoSite® Test System. This device is a standalone diagnostic test for hemoglobin, not an AI-assisted interpretation system for human readers.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Was Done
Yes, a standalone performance evaluation was done. The HemoSite® Test System itself is the "algorithm only" in this context. Its quantitative measurements were compared against the established reference method (Coulter® JT) without human interpretation as part of the measurement process. The device's performance (regression statistics and precision) represents its standalone capability.
7. The Type of Ground Truth Used
The ground truth used was measurements from a predicate laboratory instrument: the Coulter® JT hematology analyzer. This is a form of reference standard/comparative method rather than expert consensus, pathology, or outcomes data.
8. The Sample Size for the Training Set
The document does not explicitly mention a "training set" or its sample size. This is a 1998 510(k) submission for a diagnostic device that performs a chemical reaction and colorimetry. While the device likely has internal calibration and validation processes (which could be considered analogous to "training" in a very broad sense), the submission focuses on the performance of the final device against a predicate, not on the development of a complex algorithm that requires a distinct, external training set in the modern AI sense. The "Test Module" provided with each box of Test Cards contains "all current calibration information for the specific lot of reagent," implying internal calibration data, but the size or nature of this data is not detailed as a "training set."
9. How the Ground Truth for the Training Set Was Established
Since a "training set" in the modern sense of AI/machine learning isn't explicitly described or applicable, the method for establishing its ground truth is also not applicable. The device's function relies on established chemical principles and colorimetric measurement, with calibration data likely derived from known hemoglobin standards.
§ 864.7500 Whole blood hemoglobin assays.
(a)
Identification. A whole blood hemoglobin assay is a device consisting or reagents, calibrators, controls, or photometric or spectrophotometric instrumentation used to measure the hemoglobin content of whole blood for the detection of anemia. This generic device category does not include automated hemoglobin systems.(b)
Classification. Class II (special controls). An acid hematin intended for use with whole blood hemoglobin assays is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 864.9.