Search Results

SAFECARE Fentanyl Urine Test Cassette is a competitive binding, lateral flow immunochromatographic assay for qualitative detection of fentanyl in human urine at the cutoff concentrations of 1 ng/mL. The test provides only preliminary test results. A more specific alternative chemical method must be used in order to obtain a confirmed analytical result. GC/MS or LC/MS is the preferred confirmatory method. The test is intended for over-the-counter use and for in vitro diagnostic use only.

Device Description

The SAFECARE Fentanyl Urine Tests are immunoassays intended for the qualitative detection of fentanyl in human urine. Each SAFECARE Fentanyl Urine Test device consists of a Test Cassette, a Dropper and a package insert. Each Test Cassette is sealed with sachets of desiccant in an aluminum pouch.

AI/ML Overview

Here's a breakdown of the acceptance criteria and the study details for the SAFECARE Fentanyl Urine Test Cassette, based on the provided FDA 510(k) summary:

1. Table of Acceptance Criteria and Reported Device Performance

The document doesn't explicitly state "acceptance criteria" percentages (e.g., "must achieve X% sensitivity"). Instead, it presents performance data that implies the device met sufficient standards for clearance. The "performance" column below summarizes the key findings.

Category	Description of Performance (Implicit Acceptance Criteria)	Reported Device Performance
Precision	Demonstrate consistent and accurate results across different concentrations relative to the cutoff, ideally showing 100% agreement for concentrations significantly below or above the cutoff, and a mix around the cutoff.	Lot 1: 100% negative for -100%, -75%, -50% cutoff; 98% negative (1 positive) for -25% cutoff; 52% positive/48% negative for cutoff; 100% positive for +25%, +50%, +75%, +100% cutoff. Lot 2: 100% negative for -100%, -75%, -50% cutoff; 98% negative (1 positive) for -25% cutoff; 50% positive/50% negative for cutoff; 100% positive for +25%, +50%, +75%, +100% cutoff. Lot 3: 100% negative for -100%, -75%, -50% cutoff; 96% negative (2 positive) for -25% cutoff; 54% positive/46% negative for cutoff; 100% positive for +25%, +50%, +75%, +100% cutoff.
Stability	Device must remain functional and accurate for a specified shelf life under defined storage conditions.	Devices are stable for 24 months at 36-86°F based on accelerated stability studies.
Interference	No interference from a list of common physiological, pathological, or drug substances at specified concentrations when fentanyl is absent or at known concentrations (50% below/above cutoff).	No interference observed from a comprehensive list of compounds (e.g., Acetaminophen, Albumin, Glucose, Ibuprofen, Nicotine) at 100µg/mL or specified concentrations when fentanyl was at ±50% cutoff levels.
Specificity (Cross-Reactivity)	Minimal to no cross-reactivity with structurally similar compounds or other common opioids/drugs, especially for compounds listed as "no cross-reactivity." For fentanyl analogs, a defined cross-reactivity profile.	Detected various fentanyl analogs with varying cross-reactivity percentages (e.g., Acetyl fentanyl: 100%, Ocfentanil: 40%). Showed no cross-reactivity for a list of over 30 opioid compounds (e.g., Codeine, Morphine, Buprenorphine) tested at 100 µg/mL.
Effect of Urine Specific Gravity & pH	Device performance should not be significantly affected by variations in urine specific gravity (1.000-1.035) or pH (4-9) at critical fentanyl concentrations (50% below and 50% above cutoff).	All results were positive for samples at and above +50% Cut-Off and all negative for samples at and below -50% Cut-Off, regardless of specific gravity or pH within the tested ranges.
Method Comparison (Clinical Samples)	Demonstrated agreement with a confirmatory method (LC/MS) for both negative and positive clinical urine samples across a range of fentanyl concentrations, including those near the cutoff. A low number of discordant results is expected, especially near the cutoff.	Operator 1: - 39/40 (97.5%) agreement for Negative/Low Negative/Near Cutoff Negative. - 38/40 (95%) agreement for Near Cutoff Positive/High Positive. - Discordant (1 positive at 0.985 ng/mL LC/MS; 2 negative at 1.055 and 1.097 ng/mL LC/MS). Operator 2: - 40/40 (100%) agreement for Negative/Low Negative/Near Cutoff Negative. - 39/40 (97.5%) agreement for Near Cutoff Positive/High Positive. - Discordant (2 positive at 0.954 and 0.993 ng/mL LC/MS; 1 negative at 1.055 ng/mL LC/MS). Operator 3: - 40/40 (100%) agreement for Negative/Low Negative/Near Cutoff Negative. - 39/40 (97.5%) agreement for Near Cutoff Positive/High Positive. - Discordant (2 positive at 0.980 and 0.985 ng/mL LC/MS; 1 negative at 1.055 ng/mL LC/MS).
Lay-User Study	High percentage of correct results by lay users when following instructions, particularly for concentrations clearly below or above the cutoff. User instructions should be easily understood.	Correct result percentages: 100% for -100%, -75%, -50% cutoff; 95% for -25% cutoff; 100% for +25%, +50%, +75% cutoff. All lay users indicated instructions were easily followed. Flesch-Kincaid score indicated reading grade level < 7.

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

Precision Studies:
- Test Set Size: For each of 3 lots, 50 tests were performed at each of 9 concentrations (-100%, -75%, -50%, -25%, cutoff, +25%, +50%, +75%, +100%). This totals 3 lots * 9 concentrations * 50 tests/concentration = 1350 tests.
- Data Provenance: Samples were prepared by spiking fentanyl into negative samples. The document does not specify the origin of the "negative samples." It states the fentanyl concentration was confirmed by LC/MS.
Interference & Specificity Studies:
- Test Set Size: Not explicitly stated as a number of individual tests, but each interfering substance/cross-reactant was tested using "three batches of each device" with drug-free urine and target drug fentanyl urine (at ±50% cutoff).
- Data Provenance: Samples were prepared by spiking test substances into drug-free urine or urine containing fentanyl.
Effect of Urine Specific Gravity and Urine pH:
- Test Set Size: Not explicitly stated as an exact number of tests, but samples within specific gravity (1.000-1.035) and pH (4-9) ranges were spiked with fentanyl at 50% below and 50% above cutoff. These were then tested using "three lots of device."
- Data Provenance: Samples were prepared by altering the specific gravity/pH and spiking fentanyl into urine.
Method Comparison (Clinical Samples):
- Test Set Size: 80 unaltered clinical samples (40 negative and 40 positive).
- Data Provenance: "Unaltered clinical samples" - implies naturally occurring samples from a clinical population. The country of origin is not specified, but it's generally understood to be within the scope of where such tests would be used.
Lay-user Study:
- Test Set Size: 140 lay persons, each testing one blind-labeled sample. A total of 140 tests were performed. The samples themselves were 20 for each of 7 concentrations (-100%, -75%, -50%, -25%, +25%, +50%, +75% cutoff), so 7 * 20 = 140 unique prepared samples were used.
- Data Provenance: Samples were prepared by spiking fentanyl into "drug free-pooled urine specimens."

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

Method Comparison (Clinical Samples): The ground truth was established by LC/MS (Liquid Chromatography/Mass Spectrometry), which is a highly accurate and routinely used laboratory method for drug confirmation. This is a scientific instrument, not an "expert" in the human sense. The operators running the LC/MS would be trained laboratory personnel.
Precision, Interference, Specificity, Effect of Urine Specific Gravity & pH, Lay-User Study: For these studies, the "ground truth" was established by spiking known concentrations of fentanyl (or other substances) into urine samples and confirming by LC/MS. This process relies on the accuracy of the LC/MS method and the precision of the spiking process.

4. Adjudication Method for the Test Set

Method Comparison (Clinical Samples): The document states samples were "blind labeled" and compared directly to "LC/MS results." This implies a direct comparison without an explicit adjudication process involving multiple human readers beyond the initial device reading by each of the three operators. Discordant results were simply listed for each operator.
Lay-user Study: Samples were "blind-labeled and randomized." Each participant provided a result, which was then compared to the known LC/MS-confirmed concentration. There was no mention of an adjudication process among lay users or reviewers.

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

No MRMC comparative effectiveness study was done comparing human performance with and without AI assistance.
This device is a rapid diagnostic test (lateral flow immunoassay), not an AI-powered diagnostic imaging or interpretation system. Therefore, an MRMC study in the context of AI assistance is not applicable. The study involved three operators for the method comparison and 140 lay users for usability, but these were for direct device performance and ease of use, not AI-assisted interpretation.

6. Standalone Performance

Yes, a standalone (algorithm only without human-in-the-loop) performance study was done implicitly. This device is the standalone "algorithm" (the immunoassay itself). The precision, interference, and specificity studies evaluate the inherent performance of the device without a human interpreting a complex image or data set. The human involvement is limited to performing the test procedure and reading a clear positive/negative line, which is a direct output of the device's chemistry.

7. The Type of Ground Truth Used

LC/MS (Liquid Chromatography/Mass Spectrometry): This was the primary method used to establish the ground truth for all studies involving fentanyl concentrations (precision, method comparison, lay-user study, and confirming spiked concentrations for interference/specificity/pH/SG studies). LC/MS is a highly accurate and quantitative analytical chemistry technique considered the "gold standard" for confirming drug concentrations in toxicology.

8. The Sample Size for the Training Set

This document describes performance studies for a finished medical device (an immunoassay rapid test). Immunochromatographic assays do not have a "training set" in the machine learning sense. Their performance is inherent to their design, reagents, and manufacturing process. Therefore, there is no training set mentioned or applicable for this type of device.

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

As there is no "training set" for this type of device, this question is not applicable. The "ground truth" (LC/MS confirmation) is used for validation of the device's performance, not for training it.

Ask a Question

Ask a specific question about this device

Page 1 of 1