Medical Cannabis Safety Reporting: What a Physician Survey Found
| Audience | Patients, caregivers, clinicians, and cannabis-science readers interested in pharmacovigilance, adverse drug reaction reporting, and medical cannabis care |
| Primary Topic | medical cannabis safety reporting among physicians |
| Source | Read the full source |
Medical Cannabis Safety Reporting: What a Physician Survey Found
A cross-sectional survey of 253 physicians in Poland found frequent clinical exposure to medical cannabis but limited formal education and uncommon cannabis-specific adverse drug reaction reporting. Knowledge of pharmacovigilance and institutional practice context were strongly associated with general reporting. The nonprobability, self-reported survey identifies a safety-system gap, not a causal effect or an adverse-event rate.
| Study Type | Cross-sectional analytical physician survey |
| Setting | Clinically active physicians practicing in Poland |
| Data Collection | Anonymous electronic questionnaire, January through April 2025 |
| Invitations | 1,200 initial invitations through overlapping channels; unique recipient count and valid response rate unavailable |
| Participants | 253 complete submissions |
| Medical Cannabis Exposure | 205 of 253 physicians, 81.0% |
| Adequate Formal Education | 34 of 253 physicians, 13.4% |
| Cannabis-Specific ADR Reporting | 19 of 253 physicians, 7.5% |
| General ADR Reporting | 148 of 253 physicians reported the behavior |
| PV Knowledge Association | Adjusted OR 24.71; 95% CI 9.40 to 64.96 |
| Specialization Association | Adjusted OR 9.97; 95% CI 3.66 to 27.20 |
| Hospital or University Association | Adjusted OR 5.72; 95% CI 2.10 to 15.55 |
| Published | September 10, 2026 |
| DOI | 10.3389/fpubh.2026.1915903 |
Most respondents, 205 of 253, reported using or prescribing medical cannabis in clinical practice. Only 34 respondents considered their formal education in medical cannabis therapy adequate, and every respondent expressed willingness to receive additional training.
That contrast describes the surveyed group, not all physicians in Poland. Open recruitment and chain referral could overrepresent clinicians with particular interest in cannabis or pharmacovigilance, and the study could not calculate a valid response rate.
Only 19 physicians, 7.5% of the sample, reported ever submitting an adverse drug reaction report related to medical cannabis. Because the questionnaire asked about reporting behavior rather than verified case files, the result does not reveal how many suspected reactions occurred or how many reports should have been filed.
The gap could reflect underrecognition, uncertainty, limited exposure to reportable events, or features of the survey sample. The paper supports better reporting systems, but it does not permit an estimate of underreporting prevalence.
In the adjusted model for general adverse drug reaction reporting, knowledge of pharmacovigilance purpose, having a medical specialization, and working in a hospital or university setting retained positive associations. The estimates were large, with wide confidence intervals that deserve attention.
These associations should not be read as proof that training or institutional employment caused better reporting. Professional context, access to reporting systems, workload, prior experience, and unmeasured differences may still influence the pattern.
The paper argues for education that explains what should be reported, when suspicion is enough, where a report goes, and how the reporting process works. That is more useful than relying on a general sense of professional obligation.
For medical cannabis, training should also support product identification, cannabinoid profile, formulation, dose, route, co-medications, timing, and clinically relevant outcomes. Better documentation cannot prove causality in an individual case, but it can make safety reports more interpretable.
Pharmacovigilance is especially important when products and exposure patterns are heterogeneous. Useful reports connect a suspected reaction to the actual product, route, dose, timing, co-medications, and clinical course while acknowledging uncertainty about causality.
The survey also points to an organizational question. Reporting behavior may be easier where formal safety systems are visible and integrated into workflow, while clinicians in private outpatient care may face fewer prompts and less support.
The clinical lesson is not that medical cannabis is unusually unsafe. The lesson is that a rapidly expanding treatment area needs reporting habits and systems capable of detecting safety signals without waiting for perfect certainty.
I would make adverse-event documentation routine and product-specific. Ask what changed, when it changed, what else the patient took, whether symptoms resolved after stopping or reducing exposure, and whether the event belongs in a formal report.
How to Interpret This Medical Cannabis Safety Reporting Among Physicians Evidence Without Overstating It
A useful evidence report should let the signal breathe without inflating it.
The right question is not whether the paper is positive or negative, but what kind of decision it can responsibly support.
A Four-Step Reading Frame
Evidence type
Start by identifying whether the paper is a randomized trial, review, meta-analysis, observational study, or protocol.
Population
Ask whether the studied population matches the patient or clinical scenario involving pharmacovigilance, adverse drug reaction reporting, and medical cannabis care.
Outcome meaning
Look at what actually changed, how it was measured, and whether the change would matter in daily life.
Safety and uncertainty
Read limitations and adverse effects as part of the result, not as a footnote.
Eight Ways to Read the Safety-Reporting Gap
Patient, clinical, methods, safety, caregiver, evidence, ethics, and research perspectives
Your Experience Can Improve the Safety Record
Patients often notice adverse effects outside the clinic, including changes in alertness, balance, mood, heart rate, digestion, memory, or medication tolerance. The survey does not measure patient events directly, but it shows why clear communication matters when medical cannabis products vary widely in dose, formulation, and cannabinoid content.
Bring the exact product label, dose, route, timing, and medication list to clinical visits. Describe what happened and when, including improvement after holding or changing the product. A suspected reaction does not need to be proven before it is discussed or considered for formal reporting.
Make Reporting Part of the Cannabis History
The surveyed physicians reported frequent exposure to medical cannabis, yet cannabis-specific adverse drug reaction reporting was uncommon. A practical response is to incorporate safety documentation into routine cannabis histories instead of treating reporting as a separate task reserved for dramatic events.
Record the product, cannabinoid concentrations, route, dose, titration, indication, onset, duration, concurrent substances, and relevant laboratory or clinical findings. Ask whether symptoms changed after dose reduction or discontinuation. This survey cannot show that a particular workflow improves reporting, but it identifies a plausible operational gap worth addressing.
Large Odds Ratios Do Not Remove Design Limits
Knowledge of pharmacovigilance purpose had an adjusted odds ratio of 24.71 for general reporting, while specialization and hospital or university work also retained positive associations. These estimates are striking, but they arise from a cross-sectional convenience sample and carry wide confidence intervals.
The model adjusted for six prespecified covariates, yet residual confounding remains possible. Age could not be included with experience because all respondents older than 40 reported the general outcome, creating complete separation. Association strength should therefore prompt further study, not a causal claim.
Reporting Frequency Is Not Event Frequency
Only 19 respondents reported a cannabis-related adverse drug reaction, but that number cannot be converted into a safety rate. The denominator is physicians, not exposed patients, and the survey did not verify patient counts, product exposure, suspected events, or submitted reports.
The appropriate safety conclusion is narrower: the observed reporting behavior may leave real-world information underdeveloped. Better surveillance should improve case detail and signal detection without assuming that every symptom is caused by cannabis or that a low number of reports means low risk.
Observed Changes Deserve Precise Documentation
Caregivers may be the first to notice sedation, confusion, falls, appetite changes, anxiety, or functional decline after a cannabis product is started or adjusted. Their observations can be clinically valuable when the patient has difficulty describing timing or severity.
A brief log can capture product, dose, time of administration, symptom onset, other medicines, and what happened after the next dose was held or changed under clinical guidance. The study does not validate a caregiver tool, but its reporting-gap finding supports making observations specific enough for clinicians to assess.
Safety Surveillance Complements Clinical Trials
Trials can characterize common adverse events under controlled conditions, while spontaneous reports and registries can detect patterns in broader populations, longer exposure, unusual products, and combinations not fully represented in trials. Each evidence stream answers a different question and carries different biases.
This physician survey examines the reporting system itself, not product efficacy or adverse-event incidence. Its contribution is to show where surveillance capacity may be weak. It should be read alongside trial safety data, registries, and validated adverse-event databases rather than treated as a substitute for them.
Safety Learning Requires Trust Without Blame
Patients may hesitate to mention adverse effects when they fear judgment about cannabis use, while clinicians may hesitate to report when causality feels uncertain. A safety culture works best when both parties can describe uncertainty honestly and focus on preventing harm.
Reporting a suspected reaction is not an accusation against a patient, clinician, or product. It is a structured contribution to collective learning. Ethical communication should also avoid implying that this Polish convenience sample represents all physicians or that uncommon reporting proves widespread neglect.
Link Reported Behavior to Verified Systems Data
Future studies should use probability-based sampling where feasible, record valid response denominators, and connect clinician knowledge with verified reporting activity. Prospective designs could test whether targeted training, workflow prompts, or simplified digital tools increase complete and useful reports.
Medical cannabis research should also capture patient exposure denominators, product composition, dose, route, indication, co-medications, seriousness, and clinical course. Larger samples are needed for cannabis-specific multivariable analyses because only 19 events were available here. Replication outside Poland will determine whether the same organizational pattern appears elsewhere.
Join the Conversation
Have a question about how this applies to your situation? Ask Dr. Caplan
Want to discuss this topic with other patients and caregivers? Join the forum discussion
Frequently Asked Questions
What did the Polish physician survey examine?
It examined pharmacovigilance knowledge, self-reported adverse drug reaction reporting, clinical exposure to medical cannabis, and cannabis-related safety attitudes among clinically active physicians in Poland.
How many physicians participated?
The analytical sample included 253 complete electronic questionnaires collected from January through April 2025.
How many respondents had medical cannabis experience?
Two hundred five physicians, 81.0% of the sample, reported using or prescribing medical cannabis in clinical practice.
How many considered their formal medical cannabis education adequate?
Thirty-four physicians, 13.4% of respondents, considered their formal education in medical cannabis therapy adequate.
How common was cannabis-specific adverse reaction reporting?
Nineteen respondents, 7.5% of the sample, reported ever submitting an adverse drug reaction report related to medical cannabis.
Does 7.5% mean adverse reactions were rare?
No. The survey measured physician reporting behavior, not the number of exposed patients or the incidence of adverse reactions.
Which factors were associated with general adverse reaction reporting?
In the adjusted model, knowledge of pharmacovigilance purpose, having a medical specialization, and working in a hospital or university setting retained positive associations with general reporting.
Did the study prove that education improves reporting?
No. The cross-sectional design can identify associations but cannot establish that education or workplace caused better reporting behavior.
Can the findings represent all physicians in Poland?
Not reliably. Recruitment used an open, nonprobability convenience sample, unique recipients and a valid response rate were unavailable, and self-selection may limit generalizability.
What information helps when reporting a suspected cannabis reaction?
Useful details include the exact product, cannabinoid content, formulation, route, dose, timing, indication, co-medications, symptom course, seriousness, and changes after reducing or stopping exposure.