EVIDENCE LIBRARY
Cocaethylene Below the Quantification Limit: Why a Printed Number Needs Qualification
A sweat-analysis paper lists numerical cocaethylene results below its stated quantification limit. The discrepancy calls for clarification, not an assumption that every detection is false.
A numerical result below a method’s stated quantification limit should not automatically be treated as a fully supported quantitative measurement. In a 2024 sweat-analysis paper, the published cocaethylene quantification limit was 10 ng per patch, while the field-results table included smaller numerical values. The paper also said results below that limit would be reported as below the limit.
That inconsistency requires clarification before the smaller numbers are used quantitatively. It does not establish that cocaethylene was absent, that the detections were false, or that the entire analytical method was invalid.
The published method separated detection from quantification
Gomes and colleagues developed a disposable-pipette extraction procedure with gas chromatography–mass spectrometry for psychoactive substances in sweat. Table 1 lists a cocaethylene limit of detection of 5 ng per patch and a lower limit of quantification, or LLOQ, of 10 ng per patch.
The methods define the detection limit using an empirical signal-to-noise approach. The quantification limit additionally requires acceptable precision and accuracy. These are different analytical questions: whether target material can be detected, and whether its amount can be measured with the validated quantitative performance.
Section 2.9 says amounts below the LLOQ would be expressed as below that limit. However, Table 6 presents cocaethylene values including 7.800 ng per patch for sample 2S, 8.400 for 3S and 5.000 for 31S. These printed values are all below the stated 10-ng quantification limit.
What can be concluded from the discrepancy?
The published presentation does not follow a single, clearly reconciled reporting rule. The field table, the stated LLOQ and the description of below-limit reporting need to be considered together.
Table 2 also contains low-level performance entries. Their presence should be acknowledged when evaluating the method; it would be too strong to say that the publication contains no low-level analytical information. But those entries do not explain why Table 1 specifies 10 ng per patch as the LLOQ or why the field-results table uses exact numbers below that stated limit.
The appropriate scientific response is to seek clarification of the intended limit and reporting policy. Possible explanations should not be asserted as established facts. The paper does not demonstrate whether the discrepancy arose from a table error, a reporting convention or another cause.
A detected result can retain meaning without an exact amount
Where a signal satisfies identification and detection criteria but falls below the validated quantitative reporting range, a laboratory may distinguish detected material from a reliably quantified amount. The particular reporting language must follow that laboratory’s documented procedure.
One should not simply replace such a finding with zero. Conversely, extra decimal places do not establish quantitative accuracy. The number of displayed digits is a formatting property, not evidence of measurement performance.
The case review therefore needs the actual chromatographic identification, calibration range, low-level controls and reporting rules. A journal’s tabulated number cannot substitute for those records in a different laboratory’s case.
Patch mass is not a blood alcohol concentration
The study’s units were ng per patch: the mass associated with the collection device. They were not ng/mL in circulating blood, serum or plasma. The investigators collected 30 authentic sweat samples using patches worn for 24 hours.
Cocaethylene is associated with cocaine–ethanol coexposure, but it is not ethanol itself. Even a well-supported quantitative patch result would not establish a driver’s BAC, the amount of alcohol consumed or the exact time of consumption through this study.
The validation did not connect below-limit patch results with measured driving performance. It also did not establish a statistical probability that a particular low-level detection meant impairment. Analytical identification, quantitative reliability and behavioral interpretation remain separate steps.
Questions to resolve before using a low numerical result
Confirm which specimen and analyte the report concerns. Identify the method’s detection limit, quantification limit, calibration range and reporting cutoff. Then determine whether the numerical result is a validated measurement, an explicitly qualified estimate or a detected-but-not-quantified finding.
For this paper specifically, clarification should reconcile Tables 1, 2 and 6 with the reporting statement in Section 2.9. Until that is resolved, the below-LLOQ numbers should not be cited as though the publication unambiguously supports their exact quantitative use.
The site’s analysis of the same method’s freeze–thaw experiment addresses a different question: preservation under specified handling conditions. Passing a stability criterion at higher control levels does not resolve an inconsistent low-level reporting convention.
Primary source
Gomes NC, Cabrices OG, De Martinis BS. Innovative disposable pipette extraction for concurrent analysis of fourteen psychoactive substances in drug users’ sweat. Journal of Chromatography A. 2024;1730:465136. https://doi.org/10.1016/j.chroma.2024.465136.