Search Results

CR3 Keyless Split Sample Cup Amphetamine-Cocaine is a rapid test for the qualitative detection of d-Amphetamine (major metabolite of Amphetamine) and Benzoylecgonine (major metabolite of Cocaine) in human urine at a cutoff concentration of 1000ng/mL and 300ng/mL, respectively.

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 is the preferred confirmatory method. Clinical consideration and professional judgment should be exercised with any drug of abuse test result, particularly when the preliminary result is positive.

For in vitro diagnostic use only. It is intended for over-the-counter and for prescription use.

The CR3 Keyless Split Sample Cup Amphetamine-Cocaine test uses immunochromatographic assays for amphetamine and cocaine. The test is a lateral flow, one step system for the qualitative detection of d-Amphetamine (major metabolite of Amphetamine) and Benzoylecgonine (major metabolite of Cocaine) in human urine.

Here's an analysis of the acceptance criteria and the study that proves the device meets them, based on the provided text:

Acceptance Criteria and Device Performance for CR3 Keyless Split Sample Cup Amphetamine-Cocaine

1. Table of Acceptance Criteria and Reported Device Performance

The document does not explicitly present a formal table of "acceptance criteria" with specific thresholds for sensitivity, specificity, accuracy, etc., that the device must meet. Instead, the performance characteristics section describes the results of various studies, implying that these results are considered acceptable for demonstrating substantial equivalence.

Here's a summary of the reported device performance for Amphetamine (AMP) and Cocaine (COC) testing, based on the precision, cut-off verification, and comparison studies:

2. Sample Size and Data Provenance

• Sample size for the test set:

  • Precision Studies: For each of the 9 concentration levels (for both AMP and COC), 50 tests were performed per concentration (2 runs/day for 25 days). This results in a total of 9 concentration levels * 50 tests/level = 450 tests per drug (AMP/COC).
  • Cut-off Verification: 125 amphetamine samples and 125 cocaine samples (equally distributed across 5 concentrations: -50%, -25%, at cut-off, +25%, +50%). Each tested using 3 lots by 3 operators.
  • Interference & Specificity (Cross-reactivity): Substances were added to drug-free urine and urine containing drugs at ±25% cut-off. These were tested using 3 batches by 3 operators. The exact number of samples for each interfering substance or cross-reactant isn't specified, but implies multiple samples per substance.
  • Effect of Urinary Density and pH: 12 urine samples (density) and an aliquot of negative urine pool adjusted to 6 pH ranges. Each spiked with drugs at ±25% cut-off. Tested using 3 batches by 3 operators.
  • Comparison Studies: 80 clinical samples (40 negative and 40 positive) for each drug (AMP and COC). Therefore, 80 * 2 = 160 clinical samples in total.
  • Lay-user Study: 260 lay persons. 20 for drug-free samples, 120 for d-Amphetamine samples, 120 for benzoylecgonine samples. Each individual tested one sample.

• Data Provenance:

  • Clinical Samples (Comparison Studies): "unaltered clinical samples." The geographic origin (country) is not explicitly stated, but the submission is from Guangzhou, P.R. China.
  • Other Studies (Precision, Cut-off, Interference, Specificity, Density/pH, Lay-user): Samples were likely prepared in-house (e.g., drug-free pooled urine spiked with drugs). The "clinical samples" mentioned only in the Comparison Study section suggest the others used prepared samples.
  • Retrospective or Prospective: Not explicitly stated. The description of preparing and testing samples suggests a prospective setup for most analytical performance studies and the lay-user study. The "unaltered clinical samples" for the comparison study could be retrospective or prospectively collected for the study.

3. Number of Experts and Qualifications for Ground Truth

• For Analytical and Comparison Studies:

  • Comparison Study: The ground truth for the 80 clinical samples (40 negative, 40 positive for each drug) was established using GC/MS (Gas Chromatography/Mass Spectrometry). This is a highly accurate and widely accepted reference method for drug testing, often considered the "gold standard." These were then compared to the device's results as interpreted by "three laboratory assistants."
  • Cut-off, Precision, Lay-user Studies: The ground truth for these studies was established by spiking known concentrations of d-Amphetamine or benzoylecgonine into drug-free urine, with concentrations confirmed by GC/MS. The "person who prepared them" (cut-off study) was masked from the testing, ensuring blinding.

• Number of Experts:

  • For the analytical performance studies (Precision, Cut-off, Interference, Specificity, Density/pH), results were read by "three different operators." Their qualifications are not specified beyond being "operators" or "laboratory assistants."
  • For the comparison studies, the device results were viewed by "three laboratory assistants." Their specific qualifications (e.g., experience level, professional certification) are not detailed.
  • For the lay-user study, 260 "lay persons" interpreted the results themselves without expert intervention, as the intent was to evaluate usability by the general public.

4. Adjudication Method for the Test Set

• No explicit adjudication method (e.g., 2+1, 3+1) is mentioned for the interpretation of the device's results in the analytical or comparison studies. Each of the three operators' results is presented separately for the comparison study, and discordance tables show individual viewer discrepancies with GC/MS. This suggests that the operators made independent interpretations, and their individual results were compared to the GC/MS ground truth.
• In the lay-user study, each lay person interpreted their own single test device.

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

• No, an MRMC comparative effectiveness study where human readers improve with AI vs. without AI assistance was not done.
• The study described is a comparison of the device's performance against a reference method (GC/MS) when read by individual "laboratory assistants" (human readers). There is no AI component or a comparison of human performance with vs. without AI assistance.

6. Standalone Performance Study (Algorithm only)

• Yes, a standalone performance study was implicitly done. The device itself is a qualitative rapid test, effectively an "algorithm" (immunochromatographic assay producing a visual result) that produces a "positive" or "negative" outcome. The "Performance Characteristics" section details the device's intrinsic analytical capabilities:
  • Precision: How consistently the device performs with known concentrations.
  • Cut-off: Verification of the detection limit.
  • Interference & Specificity: How the device reacts to other substances.
  • Effect of Urinary Density and pH: Stability across different urine conditions.
• While human "operators" or "laboratory assistants" read the visual results, the studies establish the inherent capability of the test mechanism (the "device" or "algorithm") to detect the analytes at specified cut-offs and react appropriately to various conditions. The "Comparison Studies" further validate this by comparing the device's output (as read by humans) directly to GC/MS.

7. Type of Ground Truth Used

• The primary ground truth used for validation was GC/MS (Gas Chromatography/Mass Spectrometry).
  • For the comparison studies, "unaltered clinical samples" had their ground truth established by GC/MS.
  • For the precision, cut-off, interference, specificity, density/pH, and lay-user studies, known concentrations of target drugs were spiked into drug-free urine, and these spiked concentrations were confirmed by GC/MS. This ensures that the true concentration of the target analyte was known for these prepared samples.

8. Sample Size for the Training Set

• The document describes a submission for a medical device (a rapid diagnostic test), not a machine learning model. Therefore, there is no explicit "training set" in the context of typical AI/ML development.
• The device's underlying "learning" or "calibration" would have occurred during its initial design, manufacturing process development, and internal R&D validation (e.g., optimizing antibody concentrations, membrane properties), which typically isn't detailed as a "training set" in a regulatory submission.

9. How Ground Truth for the Training Set Was Established

• As explained above, since this is a rapid diagnostic test and not an AI/ML device, the concept of a "training set" in the machine learning sense does not apply directly.
• The "ground truth" during the device's development (analogous to how a training set might be used) would have been established through a combination of:
  • Known concentrations of pure drug analytes or their metabolites.
  • Reference laboratory methods (like GC/MS) to verify concentrations of prepared samples or evaluate early prototypes.
  • Chemical and biological assays to optimize the components (antibodies, conjugates, membranes) for specific binding and detection characteristics.

Ask a specific question about this device