Can Finger-Prick Blood Measure CBD Like Venous Blood or Plasma? A Five-Patient VAMS Study

Short answer: A 2020 proof-of-concept study measured cannabidiol (CBD) in 30-µL finger-prick blood samples from five people receiving prescription purified CBD for Dravet syndrome. The laboratory’s volumetric absorptive microsampling (VAMS) and UHPLC–MS/MS procedure produced measurable CBD results for all five, alongside results from venous whole-blood VAMS and separated plasma. The authors found no statistically significant difference in their small comparisons (reported p = 0.69). That is encouraging evidence of analytical feasibility—not proof that finger-prick blood and plasma concentrations are identical, interchangeable for every patient, or useful as individual impairment measures.

The original research is Pigliasco and colleagues, Cannabidiol Determination on Peripheral Capillary Blood Using a Microsampling Method and Ultra-High-Performance Liquid Chromatography Tandem Mass Spectrometry with On-Line Sample Preparation, Molecules 2020;25(16):3608, doi:10.3390/molecules25163608. Its central question was whether a low-volume dried whole-blood sampling method could support therapeutic drug monitoring for people prescribed cannabidiol—not whether cannabidiol proves intoxication or how a cannabis-related incident occurred.

What did the investigators actually compare?

The clinical component involved five patients treated with pharmaceutical CBD oral solution (Epidiolex) at a specialist center in Italy. Their ages were 6, 8, 12, 17, and 26 years: despite the publication’s general description of a pediatric group, the sample included one 26-year-old adult. There were four males and one female. Each person contributed a finger-prick capillary whole-blood sample, a venous whole-blood sample prepared on a VAMS tip, and a plasma sample produced from the venous blood. Those are three paired specimen types for five people, not fifteen independent patients.

The investigators reported individual CBD doses of 10, 17.5, or 20 mg/kg/day, according to the patient. The specimens were obtained during clinical monitoring at assumed steady state and described as drug-fasting draws. The paper did not randomize route, conduct a driving trial, or establish a within-person CBD concentration–impairment relationship. The venous draw supplied blood for both VAMS and centrifuged plasma, whereas the finger-prick draw provided capillary blood collected directly onto a microsampling tip.

This paired arrangement makes the paper especially relevant to questions about biological specimen matrices: different collection sites and processed specimens can yield related—but not necessarily identical—reported concentrations.

How does a 30-µL VAMS device work?

Volumetric absorptive microsampling uses an absorbent tip designed to take up a specified small volume of liquid. In this experiment, the device nominally collected 30 microliters of whole blood. The operators allowed venous-derived tips to fill and then air-dried them for one hour at room temperature. Finger-prick tips collected capillary blood. The lab used technical replicate preparations, but those repeats were not additional biological participants.

The dried tip was placed in a tube, extracted with methanol containing deuterated internal standards, incubated at approximately 37°C for ten minutes, sonicated for ten minutes, and centrifuged before injection. The reported instrument method incorporated online sample cleanup, chromatographic separation, and tandem mass-spectrometric detection. Drying, solvent extraction, and cleanup are not cosmetic steps: they determine how much drug becomes available to the detector and whether interfering matrix components are sufficiently controlled.

A low-volume tip is therefore a sample collection and preparation format, not a biological matrix identical to plasma. A VAMS result reports the validated drug concentration for the whole-blood material loaded onto the tip. Plasma has had cellular components removed by centrifugation. The distinction matters even when the numbers happen to look similar in one small dataset.

What were the five observed CBD concentration sets?

Every patient in the clinical comparison had a reported CBD concentration for all three specimen preparations. Values below are reproduced as numerical findings from the original Table 3; all concentrations are in micrograms per liter (µg/L), numerically equivalent to nanograms per milliliter (ng/mL). These are patient-specific observations in a treatment-monitoring setting, not statutory THC thresholds.

The table demonstrates both the promise and the qualification. The lowest capillary CBD result was 52.19 µg/L, and the highest was 330.14 µg/L. Yet the paired results differ. Patient 4, for example, had 122.31 µg/L in capillary VAMS and 190.21 µg/L in venous VAMS. For Patient 2, the capillary result of 153.15 µg/L exceeded the separated-plasma result of 119 µg/L. Those observations do not disappear because a small-group statistical test was nonsignificant.

Does p = 0.69 mean the specimens are interchangeable?

No. The investigators reported p = 0.69 for comparisons of capillary VAMS against venous VAMS and against plasma. The published Figure 2 depicts those group comparisons. A nonsignificant test with only five paired individuals means the analysis did not demonstrate a difference at the chosen significance threshold; it does not establish a clinically acceptable agreement margin, a zero difference, or equal concentrations for each person.

Agreement for interchangeability requires its own question: How far may a capillary result depart from the relevant plasma result before clinical decisions would change? The study provided the raw five-patient data, but it did not validate an equivalence margin across the population, prospective treatment decisions, or an individual conversion formula. Its chosen statistical analysis was a Mann–Whitney comparison, although the three specimen results within each patient were related rather than independent cohorts. That methodological limitation warrants caution in interpreting the reported p-value.

A laboratory can have good analytical precision and still observe patient-specific differences between capillary whole blood, venous whole blood, and plasma. Analytical precision concerns repeated measurements of a given type of specimen; matrix agreement concerns whether specimens from different compartments can substitute for one another. The original data support feasibility of the technology, not unlimited clinical or forensic substitutability.

Was CBD actually measured in blood, rather than inferred from a dose?

Yes. Unlike a dose-only pharmacological experiment, this study directly quantified CBD in collected human specimens using chromatography and tandem mass spectrometry. The prescribed 10–20 mg/kg/day dose information is separate from the measured 52.19–383.45 µg/L blood or plasma values. A prescribed dose does not mathematically determine any particular patient’s concentration without additional pharmacokinetic, sampling-time, and adherence information.

Measured CBD is also not the same analyte as delta-9-tetrahydrocannabinol (THC). The analytical method was validated for both parent CBD and parent delta-9-THC, but the five-patient clinical table reports CBD concentrations. The paper does not provide five paired clinical THC concentration series, and none should be invented. Nor does the study measure THC-COOH, the carboxy metabolite commonly discussed in cannabinoid urine testing.

How was the chemical identity supported?

The assay used ultra-high-performance liquid chromatography paired with a triple-quadrupole tandem mass spectrometer. Online solid-phase cleanup reduced interfering components before chromatographic analysis. Deuterium-labeled reference compounds were included as internal standards. The paper specified APCI positive-ion detection and monitored transitions around m/z 315.2→193.1 for the native targets and 318.2→196.1 for labeled standards.

CBD and THC are closely related structural isomers and therefore may share a nominal mass transition; it would be a mistake to treat a shared transition alone as full chemical differentiation. The chromatographic method separated their retention times: the original Figure 1 shows approximate CBD and THC peaks at 3.39 and 3.81 minutes. The complete workflow, including retention-time comparison, standards, transitions, calibrators, blanks, and quality controls, supports analyte-specific interpretation. That is why a discussion of confirmatory testing must identify the actual instrument method and specimen, not merely call a test “mass spectrometry.”

How sensitive and precise was the VAMS assay?

The investigators prepared calibration samples by fortifying blood from healthy donors, not by recruiting hundreds of exposed patients. The method’s reported working interval was 1–800 µg/L for CBD and THC, with a lower limit of quantitation (LLOQ) of 1 µg/L for each. The paper reported calibration coefficients near 0.998–0.9995, intra- and inter-day coefficients of variation up to 14%, and accuracy deviations spanning approximately −13% to +14% across tested levels. These numbers describe analytical method performance under the validation conditions; they are not a diagnostic accuracy percentage for identifying cannabis-impaired drivers.

The modest extraction recoveries deserve careful reading. A 25% extraction recovery for CBD does not mean the method automatically underreported every concentration by 75%. Recovery, precision, and accuracy answer different validation questions. Calibrators and internal standards can help a laboratory quantify a compound despite incomplete physical recovery, provided that extraction behavior, matrix effects, and validation remain controlled. Conversely, valid calibrations in this method cannot be borrowed for other instruments or specimens without verification.

What does four-week stability mean—and what does it not mean?

In the reported stability experiments, spiked VAMS quality-control samples were stored under tested conditions, including +25°C, for intervals through 28 days. Table 2 showed values within the authors’ acceptance framework at the studied levels. For example, at +25°C on day 28, CBD showed 97% accuracy at LLOQ and 91% at high QC; THC showed 97% and 91%, respectively. These are method validation observations on prepared controls.

They should not be recast as a universal rule that every delayed capillary sample remains accurate for four weeks. The paper did not exhaustively examine every temperature excursion, sunlight exposure, wetness event, shipping delay, atypical hematocrit, or unknown storage history. An interpretation of an actual submitted specimen still requires its collection, drying, packaging, transport, chain-of-custody, and laboratory acceptance records. Further reading on specimen collection and handling provides the broader framework, without converting this research validation into an assurance for a different case.

Was the hematocrit question fully tested?

Not across a wide clinical population. A selling point of volume-controlled absorptive tips is to reduce the variable-volume problems associated with some traditional dried blood spots. But the authors stated that the measured hematocrit values in their patients and donor blood ranged from approximately 45% to 47%. Their results therefore cannot by themselves demonstrate unbiased quantitation over the full range of severe anemia, polycythemia, newborn hematocrit, or markedly different erythrocyte distributions.

For a case involving capillary-versus-venous disagreement, a reviewer would want to know the collection technique, calibration-matrix hematocrit, tube or tip type, actual fill quality, and relevant validation data for that person’s measured blood properties. The generalized proposition that VAMS reduces hematocrit effects does not eliminate the need for analyte-specific and matrix-specific method validation.

Did the five patients supply enough data for universal therapeutic monitoring claims?

The study was explicitly a proof of concept. Its clinical comparison had five people with one specific treatment indication, formulation, and monitoring protocol. There was no independent validation cohort large enough to establish a universal capillary-to-plasma conversion equation, no clinical outcome trial showing that VAMS-guided dosage improved seizure control, and no demonstration that its agreement persists across all patient ages, concurrent medications, collection times, or extreme concentrations. Its reported blood values are descriptive observations, not universally recommended therapeutic CBD targets.

The sample included ages 6–26, and the paper described treatment with prescription purified CBD. Findings in this setting need not transfer directly to over-the-counter CBD products with different compositions, routes, excipients, or labeling accuracy. Equally, data on measured CBD in prescribed patients do not prove that a cannabis product contained or lacked THC unless the actual product or a suitable biological specimen was separately tested for THC with a validated method.

How should clinicians and laboratories use the finding?

The defensible practical message is conditional: a standardized 30-µL capillary VAMS method with robust UHPLC–MS/MS analysis may reduce collection burden when a treating service needs actual CBD concentrations and has validated its own procedure. Before replacing venous plasma monitoring, the service should verify clinical agreement limits, analytical controls, expected CBD range, timing conventions, and patient-specific factors. This study does not instruct clinicians to change a dose on the strength of one capillary measurement alone.

For laboratories, the useful lesson is that microsampling validation must separately address selection of matrices, volumetric collection, extraction efficiency, calibration, matrix effects, analyte discrimination, quality controls, stability, and incurred patient-sample comparison. For legal or administrative review, these same categories help distinguish what a measurement establishes from what remains a scientific inference. The fact that an assay is fast, low-volume, or sophisticated does not enlarge its tested scope.

Can this study determine THC-related driving impairment?

No. Its clinical endpoint was CBD concentration agreement, not driving performance. The five monitored patients were not given standardized THC doses or observed during a driving test. The paper did not measure validated roadside performance, time since THC use, case-specific psychomotor impairment, crash risk, or the effect of CBD on a particular witness. The capillary whole-blood measurements here also cannot be assumed equivalent to an entirely different jurisdiction’s required venous whole-blood forensic THC testing.

It would therefore be scientifically improper to take a particular capillary CBD value from Table 3, convert it to an alleged plasma THC or urine THC-COOH concentration, and infer contemporaneous intoxication. Those are different analytes, different matrices, and different questions. Similarly, a p-value from a five-person CBD comparison cannot validate any individual forensic cutoff or retrospectively reconstruct a person’s drug intake.

What records would matter in a real capillary-versus-venous dispute?

The first question is whether the two laboratory results were obtained from the same person, sufficiently close in time, and reported for the same molecule using appropriately validated methods. Collection site and tissue type should be identified precisely. A finger-prick VAMS specimen remains whole blood, even after drying and extraction; an ordinary venous plasma result comes from separated liquid. A different specimen or analyte invalidates a simplistic one-to-one comparison.

The second question is what the laboratory actually validated: the physical tip and nominal volume, calibration matrix, hematocrit interval, collection quality, analytical instrument, extraction and internal standards, retention-time and transition acceptance rules, limits of quantitation, matrix effects, recovery, repeat precision, and shipping stability. Without that context, a concentration printed on a report cannot establish interchangeability with a concentration generated elsewhere.

The third question is whether the requested conclusion is analytical, clinical, or legal. A validated CBD concentration is evidence that CBD was measured in that specimen under those method conditions. Therapeutic efficacy, timing of administration, impairment, and causation all require evidence beyond the number itself. The study provides helpful technical examples, but it does not resolve those further questions for a particular patient or incident.

Frequently asked questions

1. Did this study test finger-prick CBD in people?

Yes. Five individuals receiving prescription purified CBD provided finger-prick capillary blood alongside venous specimens. Their ages were 6–26 years. Each person contributed paired samples, not three independent participants.

2. Is a VAMS tip the same as a dried plasma spot?

No. The VAMS tips in this study held a measured volume of whole blood from capillary or venous collection. Plasma was separately prepared by centrifuging venous blood. The cellular composition and processing are different.

3. What does a 1 µg/L LLOQ mean?

The lower limit of quantitation was 1 microgram per liter, equivalent to 1 nanogram per milliliter, for parent CBD and parent THC in the validated VAMS procedure. It does not indicate that any human with a level below 1 µg/L was drug-free or unaffected.

4. Does p = 0.69 prove finger-prick and plasma CBD match exactly?

No. The five-person comparisons did not show a statistically significant group difference, but the individual concentrations varied and the research did not establish a universal equivalence margin or conversion equation.

5. Was THC measured and reported in the five patients?

The method was analytically validated for CBD and THC. The published five-patient comparison table reports CBD results; it does not provide a paired patient-by-patient THC dataset, so conclusions about clinical THC presence or absence cannot be constructed from it.

6. Can the four-week VAMS stability result be applied to every mailed specimen?

No. Four-week +25°C stability was demonstrated for specified prepared quality-control samples and conditions. Actual collection, drying, transport, temperature excursions, and specimen acceptance still require documentation.

7. Does the study validate finger-prick drug testing to establish driving impairment?

No. It tested feasibility of low-volume CBD therapeutic monitoring in five treated patients. It contained no controlled driving-impairment endpoint, validated legal THC threshold, case-level impairment assessment, or incident-causation experiment.

Evidence-based conclusion

Pigliasco and colleagues demonstrated that a 30-µL capillary whole-blood VAMS assay could quantify CBD in five people treated with prescription CBD, with a 1–800 µg/L validated assay range, analytical precision and stability data, and paired venous VAMS/plasma comparisons. Their published p = 0.69 findings are consistent with feasibility in this small cohort, but do not prove that capillary whole blood, venous whole blood, and separated plasma are universally interchangeable. The proper forensic takeaway is analytical specificity: identify the actual specimen, molecule, validation range, and measurement question before attempting any inference about exposure, medication effect, timing, or individual impairment.

Primary study: Pigliasco F, Barco S, Dubois S, et al. Molecules. 2020;25(16):3608. doi:10.3390/molecules25163608.