Welcome and Introduction of Kate Reynolds
Lia Mix: Welcome to the Delphi Insight session. We’re thrilled to host Kate Reynolds, Director of Research at Rocky Mountain Poison and Drug Safety (RMPDS). Kate’s background spans public health and pharmacoepidemiology [study of drug effects in populations]. She has led national projects on medication safety, from cough and cold medicine to antivenom effectiveness and pediatric overdose prevention.
For the past three years, Kate has pioneered psychedelic data-collection initiatives to ensure these treatments are studied in real-world settings. Today’s talk on understanding real-world psychedelic use through a comprehensive national survey introduces a large-scale tool tracking psychedelic use across the United States. It captures state-level prevalence, use motivations, and mental-health profiles to inform policy, clinical trials, and product development. This kind of foundational work creates the safety net an emerging field needs.
About Rocky Mountain Poison and Drug Safety
Kate Reynolds: Thank you, Lia. I’m Kate Reynolds, Director of Research at RMPDS, a division of Denver Health and Hospital Authority, the safety-net hospital for the City and County of Denver. Though housed within a hospital system, RMPDS specializes in monitoring the safety and effectiveness of drugs in the real world, work we’ve done for more than twenty years.
We operate a poison helpline (founded in 1956), provide medical information services and pharmacovigilance [drug safety monitoring], and lead epidemiologic research and public-health surveillance. Our research division has published more than 150 papers on drug use and its impacts. Portions of today’s work were funded by the Substance Abuse and Mental Health Services Administration (SAMHSA). I also want to recognize my colleague Dr. Kari Rockhill, who helped originate this project and the first version of this presentation.
Why a New Psychedelic Survey Was Needed
Kate: Psychedelic research presents distinct challenges. Use spans decades with deep cultural and historical context that can invite stigma and bias. Patterns of use vary (recreational, therapeutic, spiritual, underground), and existing data systems weren’t built to measure that complexity. We designed fit-for-purpose tools to close this gap, integrating them with our long-standing substance-use monitoring in a “mosaic” approach: multiple data sources assembled to create a fuller picture of risk and benefit.
Our population-survey stream includes NSIHT, the National Survey Investigating Hallucinogenic Trends. Development followed a rigorous roadmap:
- Focus groups with diverse individuals with lived psychedelic experience to validate constructs.
- Pilot testing (soft and hard launch) to refine instrument length and wording.
- Content validity checks by asking similar questions in different ways and probing the respondent’s understanding.
- Test–retest reliability to confirm stable responses over time.
- External comparisons against established surveys (methods papers submitted; results forthcoming).
Recruitment is critical to avoid bias. Rather than “going to where psychedelic users are,” we sample from a large commercial survey panel agnostic to substance use (participants might answer TV or travel surveys). We include users and non-users, then strategically oversample psychedelic users. We statistically link and weight these datasets to form an enriched, nationally representative sample.
Substance identification is a core quality feature. We use standardized and slang names, images, and form-specific prompts to distinguish pharmaceutical, recreational, and state-approved products. The instrument flexes as new products or contexts emerge.
To ensure data quality by design, we employ:
- Calibration weighting [statistical adjustment to match U.S. population benchmarks].
- Careless-response detection (e.g., patterned responding).
- Interleaved sections to reduce fatigue.
- Adaptive questions/skip logic [question branching] to shorten irrelevant paths.
- Randomized item ordering to distribute any fatigue effects.
- Large sample sizes to support rare-event estimation.
Beyond prevalence, we measure: reasons for use; set and setting [context such as location, support, intention]; source of acquisition; perceived risks/benefits (via validated scales); co-use patterns; and rotating modules (e.g., preparation, integration, microdosing). We also offer sponsored modules for specific stakeholder questions.
What the 2024 NSIHT Data Shows
Kate: In 2024, we fielded two waves: 57,000 total surveys with 4,329 psychedelic-user completes. Respondents represented all states and Washington, DC. Every psychedelic class was captured; roughly half used two or more psychedelics in the past year. Twenty-nine percent identified as non-white; twenty-two percent were veterans or first responders.
Prevalence highlights: cannabis led at about 19.7%. “Any psychedelic” was 4.8%. Psilocybin was the most common specific psychedelic at 2.1%, higher than several illicit comparators listed in our deck.
Reasons for use vary by substance:
- Psilocybin: predominantly fun/enjoyment (~68%).
- Prescription ketamine: slight majority for treating a medical condition; fun/enjoyment still about one-third.
- MDMA: recreation dominant (~two-thirds), with notable sexual-enhancement motives.
- LSD: mixed motives, with fun/enjoyment highest.
The takeaway: motives are heterogeneous across psychedelics, and recreation remains predominant for many substances. That matters for policy design and public perception.
Deep Dive Into Psilocybin Patterns
Kate: State maps show higher “any psychedelic” prevalence clustering in the West; psilocybin is similar with some Midwestern emergence (e.g., Nebraska, Illinois). Alaska appears as an outlier with wide confidence intervals; we’re cautious in interpreting that.
Mental-health profiles: lifetime diagnosed anxiety or depressive disorders are more common among psilocybin users than among other-psychedelic users or non-users—suggesting treatment-seeking behavior. For current symptoms, we used GAD-7 [seven-item generalized anxiety scale] and PHQ-9 [nine-item depression scale]. Among psilocybin users, a smaller share reported moderate/severe current anxiety than had an anxiety diagnosis, while depression showed the reverse pattern (moderate/severe current symptoms exceeding diagnosed depression). This underscores the need to measure both diagnosis and current symptomatology and to consider access criteria based on symptoms, not only diagnoses.
Within psilocybin users, those “treating medical symptoms only” represented a small fraction (about 2.6%). Roughly one quarter used psilocybin for both medical and other reasons; the majority did not use it for medical symptoms.
- Polysubstance psychedelic use in the past year was common across groups and especially likely among “medical-only” users.
- Concomitant [same-time] use of alcohol or cannabis was much lower in the “medical-only” group and higher among those using for mixed or non-medical reasons.
- Self-reported improvement in mental-health symptoms was high among medical-only users and even higher among those using for both medical and other reasons; little change was reported among non-medical users.
Implications: medical users often use multiple substances and may also use for non-medical reasons. Cannabis co-use is common among non-medical users and warrants attention. Differences in self-reported improvement by use pattern merit further study.
Strengths, Limitations, and Uses
Kate: Strengths include broad substance coverage, careful measurement, recruitment strategy, bias-mitigation tools, and scalable sample sizes. Limitations include cross-sectional design (no causal inference), underrepresentation risks (e.g., Indigenous communities), and self-report bias.
We publish public data reports profiling major psychedelics and offering immediately usable prevalence and pattern data. You can also sign up to receive updates on methods, publications, and future waves.
Q&A with Kate Reynolds and Lia Mix
Q1: How Does NSIHT Compare With NSDUH?
Lia: Many people know the National Survey on Drug Use and Health (NSDUH). How does NSIHT compare, and how do your estimates measure up?
Kate: In general, our estimates are similar but not identical by design. We use different methods and a more scalable approach, which involves tradeoffs in what can be captured. We also reach more psychedelic respondents, roughly twice as many per year, and include a more detailed psychedelic list. For example, some NSDUH measures report only lifetime psilocybin use, while NSIHT captures more recent use.
Q2: Where Government Surveys Miss Nuance, What Does NSIHT Add?
Lia: In areas where NSDUH may not tell the full story (analytically, not competitively), how does NSIHT strengthen understanding?
Kate: NSDUH anchors national prevalence and basic demographics. NSIHT adds depth: behavior, co-use patterns, motives, and context, and it adapts quickly as the environment changes. That ongoing, fine-grained monitoring helps interpret the topline numbers and track shifts over time.
Q3: How Do These Data Differ From Clinical-Trial Evidence?
Lia: How is NSIHT’s real-world data different from clinical-trial data?
Kate: Trials answer focused questions under controlled conditions and often exclude people with comorbidities, which can mask real-world complexity. Real-world data contextualize current practice, including populations and combinations not represented in trials. That helps assess population-level impact and risk when therapies roll out in state programs or clinical care, where contraindicated combinations may still occur.
Q4: Did You Capture Diagnoses Beyond Anxiety/Depression and State-Regulated Models?
Lia: Did your survey differentiate mental-health diagnoses or symptoms beyond anxiety and depression, and did you capture data specific to state-regulated models?
Kate: Yes. We include several additional mental-health conditions (details appear in our public report and forthcoming profiles). We can distinguish state-regulated versus other forms of use. Precision may require oversampling [intentionally recruiting more of a subgroup] as state programs scale; we’ll evaluate that over time.
Q5: Most Surprising First-Year Finding?
Lia: What was the most surprising finding from the first year, personally or for Rocky Mountain Poison and Drug Safety?
Kate: Psychedelic tourism. We saw meaningful travel to Colorado and Oregon, especially Colorado, often among people with higher rates of mental-health concerns seeking facilitated contexts and more support. That suggests strong demand for state programs and that unmet needs are already being addressed regionally.
