(188 days)
The Visualine®Cup VI for Drugs of Abuse Test is used for qualitative testing for the presence of Cocaine and it's metabolites at or above 300 ng/ml, Cannabinoids and it's metabolites at or above 50 ng/ml, Morphine and it's metabolites at or above 300 ng/ml, Methamphetamine at or above 1,000 ng/ml, Amphetamine at or above 1,000 ng/ml, Phencyclidine at or above 25 ng/ml, and Benzodiazepines and it's metabolites at or above 300 ng/ml in human urine samples. This test provides only a preliminary screening result; a more specific alternative method should be used to confirm the test result.
The Visualine® CupVI for Drugs of Abuse Test is based on the principle of antigen-antibody complexation and is used for the analysis of Cocaine, Cannabinoids, Morphine, Methamphetamine or Amphetamines, Phencyclidine, and Benzodiazepines and their corresponding metabolites in human urine samples. The assay utilizes a competitive immunochromatographic technique involving a sample of test urine delivered through a wicking action as the test panel (holding the porous membrane) is dipped into the urine sample. The drug in the sample competes for the limited antibody sites on the colored microspheres. When an adequate amount of drug is present, it will fill the limited antibody binding sites. This will prevent attachment of the colored microspheres to the probe site on the membrane. Therefore, a positive urine sample will inhibit the formation of precipitin at the probe site.
A reference or control line with a secondary antibody reaction is added to the membrane strip to check for proper sample migration on the membrane. This control line should always be present. A negative urine sample will produce two colored lines and a positive urine sample will show only one, the control line.
Here's a breakdown of the acceptance criteria and study information for the Visualine® Cup VI for Drugs of Abuse Test, based on the provided text:
Acceptance Criteria and Device Performance
| Analyte | Acceptance Criteria (Cutoff Level) | Reported Device Performance (Positive Agreement) | Reported Device Performance (Negative Agreement) | Reported Device Performance (Overall Agreement) |
|---|---|---|---|---|
| Cocaine | 300 ng/ml | 97.7% (42/43) | >99% (381/381) | >99% (423/424) |
| Cannabinoids | 50 ng/ml | 96% (46/48) | >99% (376/376) | >99% (424/424) |
| Morphine | 300 ng/ml | >99% (48/48) | >99% (375/376) | >99% (424/424) |
| Methamphetamine | 1000 ng/ml | >99% (47/47) | >99% (377/377) | >99% (424/424) |
| Phencyclidine | 25 ng/ml | >99% (46/46) | >99% (378/378) | >99% (424/424) |
| Benzodiazepines | 300 ng/ml | >99% (55/55) | >99% (369/369) | >99% (424/424) |
| Amphetamines | 1000 ng/ml | 95% (40/42) | >99% (382/382) | >99% (424/424) |
Precision Acceptance Criteria:
- Within run and run to run: >99%
- Within day and day to day: >99%
- Within lot and lot to lot: >99%
Stability Acceptance Criteria:
- A urine specimen containing 0 ng/ml of the analyte of interest will render two distinct magenta lines (one test line and one control line).
- Samples containing +25% of cutoff analyte levels will show positive results >99% of the time, therefore yielding only the control line.
Study Information
-
Sample size used for the test set and the data provenance:
- Sample Size: 424 specimens for each drug category.
- Data Provenance: Presumptive positive samples were provided by Redwood Toxicology Laboratory (Santa Rosa, California). Negative samples consisted of UTAK Laboratories Drug Free Urine pool #5488 and in-house negative urines. The study was conducted at Redwood Toxicology Laboratory and Sun Biomedical Laboratories. This indicates a mix of external and internal data, likely retrospective as existing samples were used.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The document does not specify the number or qualifications of experts used to establish ground truth. It states that presumptive positive samples were "assayed on the Hitachi at Redwood." This implies a laboratory instrument provided the reference standard, rather than human experts interpreting the results for ground truth.
-
Adjudication method for the test set:
- Not applicable. The ground truth was established by laboratory instrument (Hitachi), not through human interpretation requiring adjudication.
-
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 MRMC comparative effectiveness study was done. This device is an in-vitro diagnostic test for direct visual interpretation (presence/absence of lines) and does not involve AI or human reader interpretation in the context of diagnostic imaging or similar fields where MRMC studies are typical. The comparison was between the new device and existing reference devices.
-
If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- The performance data presented is for the standalone performance of the Visualine® Cup VI test. It's a qualitative immunoassay that produces a visual result directly. There isn't an "algorithm" in the typical AI sense; the test itself is the "device" being evaluated for its direct output. The user simply interprets the visual lines.
-
The type of ground truth used:
- Laboratory Instrument Reference Standard: The ground truth for positive samples was established by assaying them on a Hitachi instrument. For negative samples, known drug-free urine pools (UTAK Laboratories Drug Free Urine pool #5488) and in-house negative urines were used. This is a laboratory-based, objective reference standard.
-
The sample size for the training set:
- The document does not explicitly mention a "training set" in the context of machine learning or AI. This device is a biochemical immunoassay, not an algorithmic device that requires a distinct training phase with a dataset. The development and optimization of the immunoassay itself would have involved internal validation and formulation studies, but these are not described as a "training set" in the common sense.
-
How the ground truth for the training set was established:
- As there's no explicit "training set" described for an AI/ML model, this question is not applicable. The development of the immunoassay (e.g., antibody selection, membrane formulation) would have been based on established chemical and biological principles and internal validation, rather than a "ground truth" derived from a specific dataset in the way an AI model is trained.
§ 862.3870 Cannabinoid test system.
(a)
Identification. A cannabinoid test system is a device intended to measure any of the cannabinoids, hallucinogenic compounds endogenous to marihuana, in serum, plasma, saliva, and urine. Cannabinoid compounds includedelta -9-tetrahydrocannabinol, cannabidiol, cannabinol, and cannabichromene. Measurements obtained by this device are used in the diagnosis and treatment of cannabinoid use or abuse and in monitoring levels of cannabinoids during clinical investigational use.(b)
Classification. Class II (special controls). A cannabinoid test system is not exempt if it is intended for any use other than employment or insurance testing or is intended for Federal drug testing programs. The device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 862.9, provided the test system is intended for employment and insurance testing and includes a statement in the labeling that the device is intended solely for use in employment and insurance testing, and does not include devices intended for Federal drug testing programs (e.g., programs run by the Substance Abuse and Mental Health Services Administration (SAMHSA), the Department of Transportation (DOT), and the U.S. military).