A 2025 survey of 92 head and neck cancer survivors found that three-quarters reported current or former cannabis use, primarily for pain, sleep, and stress, yet fewer than one in five had received any cannabis-related guidance from a healthcare provider. The convenience sample limits prevalence claims, but the information gap between patient interest and clinical counseling is a meaningful signal for survivorship care.

Head and neck cancer survivors carry an exceptionally high symptom burden following treatment, including chronic pain, swallowing difficulty, and psychological distress. Cannabis use appears common in this population, yet clinicians rarely initiate conversations about it, leaving patients to navigate a complex therapeutic landscape without guidance. Understanding how survivors are using cannabis, why they use it, and what information they lack is essential for closing a communication gap that may carry real clinical consequences, including the dangerous misconception among some patients that cannabis can replace evidence-based cancer therapy.

Head and neck cancer survivors face a unique constellation of treatment-related sequelae, including chronic pain, dysphagia, xerostomia, and significant psychological distress, that make symptom management a central concern of survivorship care. Cannabis has emerged as a self-directed therapeutic option among many cancer survivors, yet specific data on its use among head and neck cancer patients have been scarce. This 2025 cross-sectional survey, published in Laryngoscope Investigative Otolaryngology, recruited 92 HNC survivors through the ResearchMatch national registry and 15 Facebook support groups to characterize the rate, mode, purpose, and information landscape of cannabis use in this population.

Among respondents, 46.7% reported current cannabis use and 28.3% reported former use, with edibles (46.4%) and smoking (29.0%) the most common modes. Pain relief (79.4%), improved sleep (76.5%), and stress management (64.7%) were the leading self-reported purposes. Notably, 16.7% of respondents reported using cannabis to “cure” their cancer. Only 20.7% had received any cannabis-related information post-diagnosis, while 45.6% expressed significant or very high interest in receiving such guidance. On multivariate regression, higher household income was associated with lower cannabis use odds (OR 0.115, 95% CI 0.014 to 0.662, p = 0.03). The authors acknowledge that the convenience sampling strategy, the low ResearchMatch response rate (9.3%), and the small sample size limit generalizability and call for larger, clinic-based probability studies.

Three in Four Survivors, Fewer Than One in Five With a Doctor’s Guidance: The Cannabis Information Gap in Head and Neck Cancer

Imagine recovering from surgery, radiation, and chemotherapy for a cancer that changed how you speak, swallow, and breathe, and managing the resulting pain, sleeplessness, and anxiety largely through a substance your oncologist has never asked you about. That is the picture sketched, however imperfectly, by a new survey of head and neck cancer survivors. The paper claims to characterize cannabis use among HNC survivors, and at a descriptive level it delivers a useful snapshot of modes, motivations, and information gaps. What it actually tested, though, was something narrower: the self-reported behaviors of 92 people who responded to an online survey distributed through cancer support groups on Facebook and a research volunteer registry. Before criticizing the study, I want to credit what it does well. The authors employed multiple imputation with random forest methods, reported E-values for unmeasured confounding, and acknowledged their own evidence level as Level 4. That transparency matters. The finding that edibles were the most common mode of use is genuinely relevant in a population where oral and pharyngeal dysfunction is the rule, not the exception. And the observation that nearly 17% of respondents believe cannabis can cure cancer deserves far more clinical attention than the paper itself gives it. That belief, left unaddressed, can be dangerous.

The central methodological problem, however, is straightforward: recruiting from cannabis-interested online support communities to estimate cannabis use prevalence is like surveying members of a gym to estimate how often the general public exercises. You will find impressively high numbers because the venue itself selects for people already engaged in the behavior. The 75% figure will almost certainly circulate as a fact about HNC survivors, but it is a fact about this particular, self-selected sample. The 9.3% response rate from ResearchMatch compounds the issue, and the Facebook response rate is literally unknowable. This matters not because the study is worthless, but because the headline number is the least trustworthy finding, while the most trustworthy finding, the gap between patient interest in cannabis information and actual receipt of clinical guidance, has received less attention. That informational mismatch is plausible across a wide range of sampling scenarios and is consistent with what the broader oncology cannabis literature has shown. It is, in my judgment, the genuinely actionable takeaway.

What would I say to a patient based on this? I would say that many people in your situation use cannabis for pain and sleep, and I want to be someone you can talk to about it honestly, including what we know, what we do not know, and what is not safe to assume. To a colleague, I would say this paper cannot tell us how many HNC survivors use cannabis, but it reinforces what we already suspect: patients are using it and not telling us, largely because we are not asking. To a policymaker, I would say this is not the study to build a program around, but it is one more piece of evidence that survivorship care needs structured cannabis counseling and that clinical research infrastructure for studying cannabis in this population is overdue. The durable lesson here extends well beyond oncology: convenience samples recruited from communities defined by a behavior of interest will always overestimate that behavior’s prevalence. The population in the survey is not the population in the world, and maintaining that distinction is not a technicality. It is the difference between a hypothesis and a conclusion.

This study sits at the very beginning of the research arc for cannabis use specifically in head and neck cancer survivorship. Prior work has documented cannabis use across broader oncology populations, and at least one prospective case-matched study has suggested quality-of-life benefits including decreased pain and anxiety in HNC patients. However, no large, probability-sampled study of cannabis use patterns in HNC survivors currently exists. The present work is best understood as hypothesis-generating: it identifies patterns and gaps worthy of rigorous investigation rather than confirming them.

From a pharmacological standpoint, clinicians should be aware that the predominance of edible cannabis use in this population raises specific considerations around absorption variability, delayed onset, and the potential for overconsumption, particularly in patients with altered gastrointestinal anatomy or function following treatment. Drug interactions between cannabinoids and common post-treatment medications, including certain analgesics, antiemetics, and antidepressants, remain poorly characterized. The finding that some patients believe cannabis can cure their cancer underscores the urgency of proactive counseling. The single most actionable step for clinicians today is to incorporate a brief, nonjudgmental cannabis screening question into routine HNC survivorship visits, creating a space for honest conversation that many patients clearly want but are not receiving.

This is an original cross-sectional survey study, designated as Level 4 evidence by the authors themselves. It generates descriptive data and exploratory associations from a convenience sample. It cannot establish causation, demonstrate cannabis efficacy or harm, or provide reliable population-level prevalence estimates. The single most important constraint on inference is the self-selected recruitment strategy, which systematically biases every prevalence figure upward and cannot be corrected through statistical adjustment alone.

The broader oncology cannabis literature has consistently found that patients use cannabis for symptom management at rates that outpace clinical documentation and counseling. General population studies suggest that cancer patients and survivors may actually be less likely to use cannabis than those without a cancer history, which raises questions about whether this survey’s 75% rate is inflated by sampling bias rather than indicative of genuinely higher use among HNC survivors. At least one prospective case-matched study of HNC patients has found quality-of-life improvements with cannabis use, lending some biological plausibility to the symptom management rationale reported here. The income association identified in this survey is consistent with established data showing greater cannabis use prevalence among lower-income populations in the United States. Overall, this study extends the literature descriptively but does not challenge or confirm prior findings with sufficient rigor to shift clinical assumptions.

The most consequential analytic choice was not statistical but structural: the decision to recruit heavily through Facebook HNC support groups, which effectively predetermined a high use rate. No post-hoc statistical adjustment can correct for this. Within the analysis itself, the multivariate regression produced odds ratios with extremely wide confidence intervals, particularly for the non-US residence variable (OR 0.07, 95% CI 0.003 to 0.74). A different model specification, or the exclusion of sparse cells, could easily have rendered that association nonsignificant, as indeed it became in the imputed data model (p = 0.054). The income association is more stable across both complete-case and imputed analyses, but even that result would benefit enormously from replication in a larger, probability-sampled cohort.

The most consequential misreading would be to treat the 75% cannabis use figure as a generalizable prevalence estimate for all head and neck cancer survivors. This number describes a self-selected convenience sample recruited in part through online communities where interest in cannabis is likely a selection criterion for participation. The true population prevalence is unknown and almost certainly lower. A related misreading involves the null SF-12 results: the finding that quality-of-life scores did not differ across cannabis use groups should not be interpreted as evidence that cannabis has no effect on quality of life. The study was far too underpowered to detect such differences, and an absence of statistical significance in an underpowered study is not evidence of absence.

This study provides a useful but methodologically limited first characterization of cannabis use patterns in head and neck cancer survivors. Its prevalence estimates cannot be generalized. What it does contribute is a plausible and clinically actionable signal that many HNC survivors are using cannabis without clinical guidance, and that a meaningful fraction harbor dangerous beliefs about cannabis as a cancer cure. The evidence supports initiating conversations with patients, not changing treatment protocols. Rigorous, clinic-based research in this population is urgently needed.

Do 75% of head and neck cancer survivors really use cannabis?

Not necessarily. That figure comes from a small, self-selected online sample that was partly recruited through cancer support communities on Facebook. People who use or are interested in cannabis were more likely to respond. The actual rate among all HNC survivors is unknown and probably considerably lower, though some degree of widespread use in this population is plausible based on the broader oncology literature.

Can cannabis cure head and neck cancer?

No. There is no clinical evidence that cannabis cures any type of cancer. Some laboratory studies have explored cannabinoid effects on cancer cells, but these do not translate to a treatment recommendation for patients. The finding that nearly 17% of survey respondents believed cannabis could cure their cancer is a concerning signal that highlights the importance of clear, evidence-based communication between clinicians and patients.

Should I talk to my doctor about cannabis use during or after cancer treatment?

Yes. This study’s most important finding is that the vast majority of survivors using cannabis had never discussed it with a healthcare provider, even though many wanted to. Cannabis can interact with medications, and its effects vary depending on the form used. Having an open, honest conversation with your care team allows them to help you make safer, better-informed decisions about symptom management.

Does this study prove cannabis helps with pain or sleep after head and neck cancer treatment?

No. The study recorded what patients said they used cannabis for, not whether it actually worked. It did not measure outcomes before and after cannabis use and found no difference in quality-of-life scores between users and non-users. Whether cannabis genuinely improves pain, sleep, or other symptoms in this population remains an open question requiring controlled clinical trials.

Have thoughts on this? Share it: