Can 364 ng/mL Serum THC Coexist With Passing Driving-Skills Tests? A Tourette Case Report

Short answer: Yes, the published observations show that one long-term medical-cannabis user with Tourette syndrome and attention-deficit/hyperactivity disorder (ADHD) scored above a specified German screening criterion on several computerized driving-related tests while his measured serum THC reached 364 ng/mL. That is an observation about one patient under particular testing conditions, not proof that the concentration was safe, that cannabis improved his driving, or that another person with a similar concentration would be unimpaired. The investigators did not conduct an actual on-road driving examination or a randomized untreated-versus-treated trial.

The original primary study is Streetz and colleagues’ 2025 case report in Frontiers in Psychiatry, DOI: 10.3389/fpsyt.2025.1595649. The report combines serial laboratory measurements of three cannabinoids in serum with clinical scales and the Vienna Test System. Its forensic value lies as much in the distinctions among those measurements as in the unusually high numerical result.

What is the narrow scientific question?

The study asks what happened when a 28-year-old man with longstanding Tourette syndrome and comorbid ADHD underwent repeated psychometric testing on a day when he consumed substantial amounts of prescribed inhaled cannabis and on the following day without additional cannabis intake. The researchers measured THC and two metabolites in serum and assessed selected skills considered relevant to driving under a German evaluation framework. This was a single-patient case report, n = 1; the multiple blood samples and repeated tests do not turn it into a multi-participant experiment.

The evidence can answer whether this individual exceeded the specified test criterion at the observed times. It cannot, standing alone, establish an effect of cannabis treatment, prove a universal THC cutoff, estimate accident risk, or determine an individual’s fitness to drive in a separate legal proceeding. Passing a psychometric battery is not interchangeable with completing a real traffic course safely.

Who was studied, and why do his diagnoses matter?

The participant was a 28-year-old male patient with motor and vocal tics and ADHD. He had used several other interventions before being prescribed medicinal cannabis and reported that earlier medications caused sedation or lethargy that interfered with his everyday functioning. The report describes years of ongoing cannabis exposure and a self-reported reduction in tic burden. That history matters because untreated or incompletely controlled neurological symptoms can themselves affect attention, response organization and the demands of driving.

His pattern of treatment was atypically intensive: the authors described daily use of cannabis flower up to approximately 10 grams, involving THC-dominant varieties. That figure concerns plant material consumed; it is not an analytical measurement of absorbed THC in milligrams, an administered pharmaceutical THC dose, or a transferable safe amount. The authors did not independently establish the exact delivered dose from the inhaled material, nor did they compare this person with an unexposed matched control.

Because the participant was a highly experienced user with a specific clinical presentation, extrapolation to an occasional user, a different disability, an older patient or an unmedicated driver would be scientifically unsupported. Neither tolerance nor symptom relief can simply be assumed from the serum concentration.

What did the investigators administer and when?

On the first testing day, the patient smoked or vaporized his ordinary prescribed cannabis according to perceived clinical need. The testing sequence described 2.3 grams before the first period, a further 1.0 gram before the first Vienna Test System assessment, and another 0.8 gram before subsequent testing. The study’s tables report cumulative plant-material amounts of 3.3 grams and 4.1 grams at the two main performance-test occasions. Additional use later brought the reported daily total to 8.8 grams.

On the next day the same assessment sequence was repeated without further cannabis intake that day. The interval since the last inhalation was approximately 13–15 hours, depending on which assessment was being discussed. Importantly, this was not an abstinence condition: laboratory samples still contained parent THC. The participant also decided how much medicinal cannabis he needed, meaning that dose and underlying symptoms were not independently randomized or blinded.

These details constrain every comparison. The day-1 and day-2 sessions differ in time since exposure, symptom experience, possible retest learning, residual circulating THC and perhaps other unmeasured conditions. They are not clean experimental estimates of how many points cannabis added to performance.

What did the laboratory directly measure?

The chemical measurements were made on serum using an accredited liquid chromatography–tandem mass spectrometry procedure (LC–MS/MS) with deuterated internal standards, as described by the authors. Seven samples were reported across the two days. The three reported analytes were parent Δ9-tetrahydrocannabinol (THC), 11-hydroxy-THC (11-OH-THC), and 11-nor-9-carboxy-THC (THC-COOH). Those substances are not analytically or pharmacologically identical.

Parent THC is the principal psychoactive compound tested in this series; 11-OH-THC is another psychoactive cannabinoid metabolite, while THC-COOH is chiefly used as an exposure-associated metabolite rather than a direct measure of current psychomotor effect. The article reported values in ng/mL, not percent alcohol concentration, not a breath-derived BAC equivalent, and not an oral-fluid result.

Serum is not venous whole blood. A measured concentration in the serum fraction cannot be substituted numerically for a forensic whole-blood THC result without an adequately supported matrix-specific interpretation. Nor does identifying parent THC in serum by itself establish current impairment. For foundational distinctions among specimen types, see the published biological matrices guide.

What were the seven serum findings?

The original article’s Table 3 provides the following concentration sequence. These are individual sampling occasions in one person, not a distribution of seven different drivers. The cannabis-gram column is the cumulative amount the patient reportedly used before each sample, not a directly measured dose of bioavailable THC.

The highest directly measured parent-THC value was 364 ng/mL in serum. The next-day samples were still 19 ng/mL, despite no fresh use that day. The metabolites did not follow identical numerical trajectories: day-1 11-OH-THC ranged from 11 to 17 ng/mL, while day-1 THC-COOH ranged from 58 to 65 ng/mL. Separate metabolite signals should not be conflated with a quantitative behavioral score.

The apparent swings from 364 to 144 and from 287 to 63 ng/mL require the actual sampling timeline for interpretation; they do not justify estimating a person’s elimination half-life from two chosen points or inferring exact inhaled dose. The investigators did not present these seven observations as a validated individualized pharmacokinetic model. A further limitation for expert use is that the case report does not provide the underlying full chromatograms, batch-specific quality-control files or a quantified analytical uncertainty budget.

Which driving-related skills were tested?

The Vienna Test System used three families of computerized tasks: a choice-reaction-time test (RT), a determination test assessing performance under stimulus load and stress (DT), and an adaptive tachistoscopic traffic-perception test (ATAVT). A missed-reactions score was also reported for the determination task. These measures address components plausibly relevant to driving, but they do not recreate real road geometry, unforeseen traffic conflicts, police observations, or crash outcomes.

The publication applied the cited German testing framework in which a percentile above 15 on the relevant domains met its described screening criterion. The percentile is a comparison with test norms; it is not a percentage of error-free driving and is not an intoxication probability. Whether a person legally may drive is a separate jurisdiction- and circumstance-specific question.

What did the Vienna tests actually show?

All of the reported percentile scores were above the described screening criterion. Some scores were higher during the first day, others differed across tests, and the report offered no separate randomized control condition from which one could estimate a cannabis-caused change. The observed pattern is laid out below rather than compressed into the misleading statement that high THC “improved driving.”

The lowest score anywhere in this table was RT percentile 30 on day 2, still above the test framework’s threshold. RT percentile 74 after the first day-1 interval should not be re-expressed as a 74% chance of reacting correctly in traffic. Likewise, DT percentile 94 is not evidence of a 94% accident-free driving probability.

The serum results were collected at several moments surrounding the test sequence, but the paper does not supply a validated rule linking each task’s percentile to an exact simultaneous serum THC value, even when a close-in-time blood sample exists. The sentence “364 ng/mL produced a passing test” is therefore too causal and too temporally confident. The defensible statement is that the investigators observed high serum THC within a testing day on which this participant met their selected computerized assessment criterion.

How did the clinical tic and ADHD measures differ?

The case report also assessed symptoms of Tourette syndrome and ADHD. The Yale Global Tic Severity Scale total-tic score was 38 on the cannabis-use day and 41 on the following day. Its broader global-severity score was 48 versus 71; the modified Rush video tic measure was 19 versus 26. These measurements describe the observed patient at two sessions, but do not estimate a population treatment effect.

For ADHD, the Qb test reported activity Q-scores of 1.8 and 1.9, impulsivity Q-scores of 2.1 and 2.2, and attention Q-scores of −0.5 and 0.6 on the two days, respectively. The report describes better attention on the cannabis day, but one measurement pair cannot determine whether cannabis, symptom fluctuations, sleep, repetition or other circumstances explain the difference. A patient’s statement that cannabis relieved symptoms is valuable clinical history, not equivalent to an independent trial with blinded allocation.

This clinical context matters in litigation because a symptom-treated patient may perform differently from an untreated patient for reasons unrelated to the drug’s intoxicating effects. Conversely, fewer visible tics do not establish unimpaired divided attention, judgment or crash avoidance. The relevant causal paths must be distinguished rather than collapsed into a single overall label.

Why is this not proof that high serum THC is harmless?

A high concentration without failing these particular tests in one person establishes that serum THC and the selected test scores did not map onto a universal fail rule in that observation. It does not establish that the concentration lacked any effect, that tolerance eliminated every hazard, that the patient would avoid collisions, or that an unrelated driver would perform similarly. No vehicle was driven on public roads for the study, and no collision, actual lane-keeping or emergency maneuvering outcome was observed.

There was no blinded control, placebo or random dosing. Day 2 was not a verified drug-free baseline, as serum THC remained at 19 ng/mL. The participant had extensive previous exposure, comorbid conditions, self-selected dosing and repeated exposure to the same tests. In addition, the data consist of individual descriptive observations: no group mean, reliable population confidence interval, sensitivity/specificity or receiver-operating-characteristic curve can be calculated to validate a THC threshold.

Those limitations apply in both directions. It would be equally inappropriate to claim that this single report proves cannabis improves driving ability, or that a high serum value alone establishes impairment in every chronic medical user. The study demonstrates a need for properly specified matrices, validated task outcomes, complete timelines and clinical-context evidence.

Can serum THC be converted into a whole-blood legal limit?

Not automatically. The specimen described by Streetz and colleagues was serum, produced from blood after separation. Many forensic reports and statutes instead address whole blood or a specific analytical definition of blood concentration. An expert should identify the collection container, specimen fraction, analyte, method, calibration, sample interval and the actual controlling legal definition before making comparisons.

Using a generic serum-to-whole-blood ratio to transform 364 ng/mL into an asserted legally equivalent number would add an external assumption absent from this case report. Such a calculation would require scientifically supported matrix-specific evidence and a clear statement of uncertainty. Nor can one infer an exact consumption time or dose by retroactively selecting one serum value from this series.

For the related laboratory distinction between a specifically identified substance and a presumptive result, see the published confirmatory drug testing guide. Here, the reported LC–MS/MS was a specific laboratory assay, but assay specificity does not by itself validate an individualized impairment inference.

What would a litigation-quality review need?

A case-specific review would need the complete original laboratory report, chain of custody, serum collection and processing records, calibration and control acceptance, lower quantitation limit, sample timestamps and method validation. For a performance allegation, it would also require the raw test output, standardized administration details, normative framework and the timing of other medications or clinical symptoms. When available, contemporaneous on-road conduct or appropriately validated simulator evidence would answer different questions than this clinical computerized screen.

It is particularly important to keep a legally defined prohibited concentration separate from a scientific opinion about functional impairment. An applicable per se rule is a legal element, not a conclusion that can be derived from this n=1 case report. Medical authorization, statutory exceptions and jurisdictional treatment of prescribed cannabis likewise cannot be decided by applying a German study’s discussion to a different court.

Documenting the actual specimen collection and timeline is central; the site’s published specimen collection overview explains why labeling, timing and handling are part of interpretation. The Streetz report alone cannot replace original evidence from the matter being evaluated.

What are the source’s most important unanswered questions?

First, because the participant was already using cannabis heavily, the experiment never observed his genuine cannabis-naive or independently drug-free driving baseline. Second, the measurements lack a contemporaneous untreated control person or blinded assignment, leaving treatment, familiarity and test-order effects unresolved. Third, the paper uses high amounts of reported cannabis flower without a verified individual systemic THC dose; smoking and vaporization also differ in delivery dynamics.

Fourth, the computerized battery does not observe real-world driving. Fifth, seven serum samples are insufficient for an individualized concentration–effect model, especially without complete specimen and assay records. Sixth, a single patient cannot define a general rule for Tourette syndrome, ADHD, medical cannabis users or all drivers. Seventh, the authors disclose extensive cannabis-related professional and research relationships for one coauthor, which belongs in transparent source appraisal without assuming that it invalidates the measured numbers.

None of these points erases the documented test results. They define the boundary of what those results can establish. Scientific scrutiny is strongest when the observation is preserved accurately but the untested extrapolations are excluded.

Frequently asked questions

1. Did the report measure 364 ng/mL THC in whole blood?

No. The published method specifies serum analyzed by LC–MS/MS. The concentration should be called a serum result unless supported independent records demonstrate otherwise.

2. How many drivers were tested?

One patient underwent several repeated computerized tests. Seven serum samples and multiple test sessions do not create seven independent participants or a controlled driving cohort.

3. Did he actually drive on a road in the experiment?

No on-road test or instrumented road-driving outcome was reported. The measures were computerized psychometric tasks that address components considered relevant to driving.

4. Does a percentile of 94 mean 94% driving safety?

No. Percentile ranks compare performance with a normative reference distribution. They are not calibrated probabilities of safe operation, collision avoidance or sobriety.

5. Did the second day provide a THC-free comparison?

No. He reportedly used no new cannabis that day, but two serum samples still contained 19 ng/mL parent THC. It was not an independently verified unexposed baseline.

6. Does this case prove chronic users cannot be impaired by cannabis?

No. One person’s favorable results on specified tasks cannot establish a population-wide immunity, person-specific safety guarantee or universal concentration threshold. Other functions and actual road conduct were not tested.

7. Can a lawyer or expert use the paper to decide another driver’s impairment?

It can inform a discussion of specimen matrices, test interpretation and scientific limits, but cannot replace that person’s validated forensic result, examination, conduct evidence, applicable law and complete medical history. Case-specific causation must be evaluated separately.

Bottom line: preserve the observation, reject the overreach

Streetz and colleagues documented a striking combination: serial LC–MS/MS serum THC results as high as 364 ng/mL, continued 19 ng/mL the next day, and passing scores under the reported German psychometric criterion in one very experienced medicinal-cannabis user with Tourette syndrome and ADHD. The responsible answer is neither to discard the observation nor to transform it into a declaration of universal safety or universal impairment. It is a case-specific, descriptive result with important limits on matrix conversion, medication effect, test meaning and driving causation.

Primary research: Streetz CM, Szejko N, Pisarenko A, et al. “Case Report: Effect of medicinal cannabis on fitness to drive in a patient with Tourette Syndrome and ADHD.” Frontiers in Psychiatry. 2025;16:1595649. https://doi.org/10.3389/fpsyt.2025.1595649.