Urine Fluorescence Testing for Cannabis: Why Prediction Samples Are Not a Population Validation
A fluorescence method distinguished cannabinoid-related signals in prepared urine samples and a very small authentic sample set. That is evidence about analytical feasibility, not proof of reliable screening across every person or of contemporaneous cannabis impairment.
How the signal was separated
The study coupled excitation–emission matrix fluorescence with parallel factor analysis. The algorithm decomposed overlapping optical signals into components, allowing predictions based on calibration mixtures.
The methods describe reference urine from two nonconsumer volunteers with added standards and specimens from two anonymous frequent cannabis users. Those groups should remain distinct: adding a known standard to urine does not reproduce every feature of an authentic user’s specimen.
The laboratory comparison matters
The researchers compared predictions with gas chromatography–mass spectrometry and reported favorable prediction errors under their test conditions. They also described urine extraction and solvent use. The paper’s broad language about a solvent-less optical method should not be interpreted as proving the entire authentic-urine preparation workflow used no solvents.
The paper’s nomenclature for its carboxylated target also requires care: its label THC-COOH should not be casually treated as interchangeable with the plant precursor THCA. Reference-standard identity and the actual validated analyte are essential in any case application.
What larger validation would need
Two authentic donors cannot establish population-wide sensitivity, specificity or robust interference performance. Calibration fit, prediction error and a clinically meaningful error rate are different quantities.
The method also does not measure driving behavior. Before relying on an optical screening result, identify the matrix, preparation, analyte, reporting range and confirmatory procedure. A urine signal, even accurately measured, does not establish impairment during a separate event.
Related analysis
See our discussion of why chemical detection and impairment remain separate.
Primary research
Sheng-Feng Cui and colleagues. Identification and Quantification of the Main Psychoactive Ingredients of Cannabis in Urine Using Excitation–Eemission Matrix Fluorescence Coupled with Parallel Factor Analysis. ACS Omega, 2023. DOI: 10.1021/acsomega.3c04913.
