(264 days)
AFIAS iFOB in conjunction with AFIAS 50 is a fluorescence immunoassay system for qualitative detection of fecal occult blood (FOB) in human fecal samples. AFIAS iFOB is an in vitro diagnostic test used by professional clinical laboratories and clinical reference laboratories for routine physical examination when gastrointestinal bleeding may be suspected. Intended users/operators for AFIAS iFOB is professional medical personnel.
AFIAS iFOB in conjunction with AFIAS-50 is a fluorescence immunoassay system for qualitative detection of fecal occult blood (FOB) in human fecal samples.
Components of AFIAS iFOB: AFIAS iFOB consist of a test cartridge, ID chip, sample collection tube contain the extraction buffer, package insert, applicator sticks, collection slide, mailing envelope, sample collection tissues, instruction for use and patient instructions.
The AFIAS iFOB test cartridge contains a test strip; with a nitrocellulose membrane of which, mouse monoclonal anti hemoglobin labeled with fluorescence and anti rabbit IgG labeled fluorescence have been immobilized at the glaze line, mouse monoclonal anti hemoglobin at the test line and rabbit IgG at the control line. Each test cartridge is individually sealed in an aluminum foil pouch containing a desiccant. Twenty-five sealed test cartridges are packed in a box which also contains an ID chip and 25 mailing envelopes which contain a collection slide, applicator sticks and sample collection tissues.
The ID chip contains a memory device that contains encoded calibration data/information for the batch (lot-to-lot) variation. With the ID chip inserted in the designated port, AFIAS-50 reads and utilizes the calibration data regarding the batch/lot under consideration and applies appropriate correction to the conversion formula while computing the test result.
The AFIAS iFOB extraction buffer tube with extraction buffer is sealed with plastic caps. The upper side is capped with a plastic cap without a sampling stick. The bottom side is capped with a plastic cap with a sampling stick. The extraction buffer contains bovine serum albumin (BSA) as a stabilizer, tween 20 as a surfactant and sodium azide in phosphate buffered saline (PBS) as a preservative. Each extraction buffer tube contains 1 mL extraction buffer. Twenty-five pre-filled extraction buffer tubes are packed in a test cartridge box.
Components of AFIAS-50: Power cable, Barcode reader, Thermal printer paper, Collection tube rack holder, Sample tips, Test cartridge magazine, Waste bin, System check cartridge set.
The provided text describes the performance characteristics of the AFIAS iFOB device, including precision, prozone effect, specificity, cross-reactivity, interference, stability studies, assay cut-off, and method comparison with a predicate device.
Here's the information extracted and organized according to your request:
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly define "acceptance criteria" in a separate section with specific numerical thresholds for each performance characteristic. Instead, it presents study results and then generally states whether the results "passed acceptance criteria" or demonstrated "acceptable overall percent agreement." For the purpose of this response, I will interpret the reported performance as meeting implicit acceptance criteria due to phrases like "passed acceptance criteria" or "acceptable overall percent agreement."
| Characteristic | Acceptance Criteria (Implicit) | Reported Device Performance |
|---|---|---|
| Precision | Passed acceptance criteria for repeatability and reproducibility. | Repeatability: Overall Percent Agreement: 98.0%, Positive Percent Agreement: 100.0% (92.7% – 100.0%), Negative Percent Agreement: 95.9% (86.3% – 98.9%). |
| Lot-to-Lot Reproducibility: Overall Percent Agreement: 98.3%, Positive Percent Agreement: 99.3% (96.3% – 99.9%), Negative Percent Agreement: 97.3% (93.2 – 98.9%). | ||
| Between-run Reproducibility: Overall Percent Agreement: 98.0%, Positive Percent Agreement: 99.3% (96.3% – 99.9%), Negative Percent Agreement: 96.6% (92.2% – 98.5%). | ||
| Between-Device Reproducibility: Overall Percent Agreement: 98.3%, Positive Percent Agreement: 98.7% (95.2% – 99.6%), Negative Percent Agreement: 98.0% (94.1% – 99.3%). | ||
| Between-site Reproducibility: Overall Percent Agreement: 98.3%, Positive Percent Agreement: 98.6% (95.2% – 99.6%), Negative Percent Agreement: 98.0% (94.2% – 99.3%). | ||
| Combined Reproducibility: Overall Percent Agreement: 98.2%, Positive Percent Agreement: 99.1% (98.0% – 99.6%), Negative Percent Agreement: 97.3% (95.8% – 98.3%). | ||
| Prozone (Hook Effect) | Not susceptible to prozone/hook effect up to a specified hemoglobin concentration. | Not susceptible to prozone/hook effect up to a hemoglobin concentration of 2000 ng/mL. |
| Specificity (Hb Variant) | Equally sensitive to 'Hemoglobin S' as human hemoglobin. | Equally sensitive to 'Hemoglobin S'. |
| Cross-Reactivity | No significant cross-reactivity with specified animal hemoglobin. | No significant cross-reactivity with bovine, chicken, fish, horse, goat, pig, rabbit, and sheep hemoglobin at specified concentrations. |
| Interference | No significant interference with specified biomolecules. | No significant interference with ascorbic acid, bilirubin, albumin, myoglobin, glucose, and triglyceride mixture at specified concentrations. |
| Test Kit Stability | Stable for an estimated period at a given temperature range. | Stable for an estimated period of 20 months at 4-30°C. |
| Stool Sample (Slide) Stability | Stable for a specified duration at room temperature. | Stable up to 30 days when stored at room temperature (25-30°C). |
| Stool Sample (Buffer Tube) Stability | Stable for a specified duration when refrigerated. | Stable up to 14 days when stored at 2 to 8°C. |
| Stool Sample (Cups) Stability | Stable for specified durations at various temperatures. | Stable for 30 days if stored at -20°C and for 2 days if stored at 37°C. |
| Controls Stability | Shelf-life and open-vial stability for controls at specified conditions. | i-CHROMA iFOB controls shelf-life stability was three months and open-vial stability was one month when stored at 2-8°C. |
| Humidity Effect | No humidity effect on test results up to a specified humidity level. | No humidity effect on AFIAS iFOB test results up to 75±5% of humidity. |
| Assay Cut-off | Validated cut-off value matching the specified threshold. | Cut-off determined to be 8.0 µg hemoglobin/g stool or 100 ng/mL (hemoglobin in fecal sample mixed with detection buffer). Performance at cut-off: For 100 ng/mL, 57.5% Positive (85.1%-100.0% CI). For values below (85, 90 ng/mL) and above (110, 130, 150 ng/mL) the cut-off, the percent positive/negative results align with expectations. |
| Method Comparison (vs. Predicate) | Acceptable overall percent agreement, positive percent agreement, and negative percent agreement with predicate device. | Combined Sites: Overall Percent Agreement: 98.5% (94.8%-99.4% CI), Positive Percent Agreement: 98.2% (94.8%-99.4% CI), Negative Percent Agreement: 98.6% (96.8%-99.4% CI). |
| Specimen Collection Verification | No statistical significance in analyses by two sampling methods, with acceptable agreement. | Acceptable overall percent agreement as well as positive percent agreement and negative percent agreement. No statistical significance in the analyses by the two sampling methods. |
2. Sample Sizes Used for the Test Set and Data Provenance
- Precision/Reproducibility:
- Sample Size: Seven fecal hemoglobin concentrations (4 µg Hb/g stool to 80 µg Hb/g stool or equivalent ng/mL). Fourteen replicates were performed for each sample and concentration level per study (repeatability, lot-to-lot, between-run, between-device, between-site, combined). Totaling 98 samples for repeatability and 294 for each reproducibility component, and 1274 for combined reproducibility results shown in the table.
- Data Provenance: Not explicitly stated, but implies laboratory-controlled spiked fecal samples. Conducted at "one intended use site" for repeatability, and "three intended use sites" for reproducibility. Locations of these sites (e.g., country of origin) are not specified. The study design (spiked samples) suggests a controlled laboratory setting rather than direct patient samples.
- Prozone (Hook Effect):
- Sample Size: Twelve hemoglobin concentrations (700 ng/mL to 2000 ng/mL). Twenty aliquots of each sample mixed with extraction buffer.
- Data Provenance: Spiked stool specimens with human blood (controlled laboratory setting).
- Specificity & Cross-Reactivity:
- Sample Size: Not explicitly stated for each test, but involved testing with specific animal hemoglobin concentrations and a 'Hemoglobin S' human variant.
- Data Provenance: Spiked test samples (controlled laboratory setting).
- Interference:
- Sample Size: Not explicitly stated for each test, but involved testing with specific biomolecule concentrations.
- Data Provenance: Spiked test samples (controlled laboratory setting).
- Stability Studies (Test Kit, Stool Sample on Slide, Stool Sample in Buffer Tube, Stool Sample in Cups):
- Sample Size: Nine hemoglobin concentrations (25 ng/mL to 500 ng/mL) for test kit, slide, and buffer tube stability. Twenty-one aliquots of each concentration for slide and buffer tube stability. Six hemoglobin concentrations (50 ng/mL to 150 ng/mL), with 384 specimen cups (64 cups x 6 concentrations x 4 storage temperatures) for stool sample in cups stability.
- Data Provenance: Spiked Hb-free stool samples with human blood (controlled laboratory setting).
- Stability of i-CHROMA iFOB Controls:
- Sample Size: Ten aliquots of each control level. Three lots of i-CHROMA iFOB Controls (negative and positive).
- Data Provenance: Not entirely clear if these are spiked or internal controls themselves, implying a controlled laboratory evaluation.
- Humidity Effect Stability Study:
- Sample Size: Seven hemoglobin concentrations (50 ng/mL to 1,000 ng/mL). 50 AFIAS iFOB test cartridges.
- Data Provenance: Spiked Hb-free stool samples with human blood (controlled laboratory setting).
- Assay Cut-off:
- Sample Size: Seven fecal hemoglobin concentrations (50 ng/mL to 150 ng/mL). Forty aliquots of each concentration.
- Data Provenance: Spiked stool samples with human blood (controlled laboratory setting).
- Method Comparison:
- Sample Size: 522 patient samples (165 positive, 357 negative).
- Data Provenance: Patient samples. Conducted at "one professional medical laboratory in the U.S. and two international professional medical laboratories." The document does not specify if these were retrospective or prospective, but the phrasing "patient samples" and "method comparison" often implies prospective or a mix.
- Specimen Collection Verification:
- Sample Size: Eight hemoglobin concentrations (25 ng/mL to 1,000 ng/mL). Twenty aliquots of each concentration.
- Data Provenance: Spiked Hb-free fecal samples with human blood (controlled laboratory setting).
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
- For Precision, Prozone, Specificity, Cross-Reactivity, Interference, Stability, Humidity Effect, Assay Cut-off, and Specimen Collection Verification studies: The "ground truth" was established by creating spiked fecal samples with known hemoglobin concentrations. This means the "expert" here is the precise control over the spiking process and analytical methods to determine the exact hemoglobin levels. Human experts were not involved in establishing this ground truth, rather the experimental design.
- For Method Comparison: The ground truth for the 522 patient samples was established by the predicate device, i-CHROMA iFOB test (K132167). This study compared the new device's results to the predicate device's results. No explicit mention of human experts defining the "true" positive or negative status of these patient samples beyond the predicate device's output.
4. Adjudication Method for the Test Set
- For Precision, Prozone, Specificity, Cross-Reactivity, Interference, Stability, Humidity Effect, Assay Cut-off, and Specimen Collection Verification studies: No adjudication method was mentioned as the ground truth was based on known, spiked concentrations or direct comparison to the predicate's established output for comparison.
- For Method Comparison: No specific adjudication method (like 2+1 or 3+1) is mentioned. The comparison was directly between the AFIAS iFOB and the predicate i-CHROMA iFOB device.
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, a Multi Reader Multi Case (MRMC) comparative effectiveness study was not done. This device is an in vitro diagnostic (IVD) fluorescence immunoassay system for qualitative detection of fecal occult blood (FOB), which automatically produces a result (positive or negative) without human interpretation of images or complex data where an "AI assistant" would typically improve a human reader's performance. The studies performed were analytical performance studies and a method comparison with a predicate device.
6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) was done
- Yes, the performance studies described are essentially standalone. The AFIAS iFOB system (device with its instrument AFIAS-50) is an automated system. The performance characteristics (precision, stability, cut-off, method comparison) evaluate the device's ability to accurately detect FOB based on its internal algorithms and fluorescence detection, without human interpretation of the final result. The professional medical personnel are "users/operators," not "interpreters" who then make a diagnostic decision on top of what the device outputs.
7. The Type of Ground Truth Used
- For most performance studies (Precision, Prozone, Specificity, Cross-Reactivity, Interference, Stability, Humidity Effect, Assay Cut-off, Specimen Collection Verification): The ground truth was spiked samples with known, carefully defined concentrations of human hemoglobin or other target analytes/interferents. This is a form of analytical ground truth.
- For Method Comparison: The ground truth for the patient samples was established by the predicate device (i-CHROMA iFOB test). This means the new device was compared against a previously cleared and accepted device's results.
8. The Sample Size for the Training Set
The document does not mention a "training set" or machine learning in the context of device development. This type of IVD device (fluorescence immunoassay) typically relies on pre-programmed calibration data and antigen-antibody reactions, not on machine learning algorithms that require explicit training sets. The "ID chip" contains encoded calibration data, which serves a similar function to calibration in traditional assays rather than training in machine learning.
9. How the Ground Truth for the Training Set Was Established
As no "training set" is mentioned or implied for machine learning, this question is not applicable. The device's operation is based on established immunochemistry principles and factory-calibrated parameters transmitted via an "ID chip" for each lot.
§ 864.6550 Occult blood test.
(a)
Identification. An occult blood test is a device used to detect occult blood in urine or feces. (Occult blood is blood present in such small quantities that it can be detected only by chemical tests of suspected material, or by microscopic or spectroscopic examination.)(b)
Classification. Class II (special controls). A control intended for use with an occult blood test is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 864.9.