Researchers analyzed national repeat cross-sectional surveys from 2023 and 2024. Participants were UK residents aged 16 to 65 who reported cannabis use within the prior 12 months.

The analysis compared medical use with a prescription, medical use without a prescription, and nonmedical use. It examined sources, product types, frequent product use, and a screening threshold for high-risk use.

Medical use without a prescription was more common than prescribed medical use in this sample. Among respondents with prescriptions, only 10.9% reported obtaining all cannabis through a prescription.

After adjustment, prescribed respondents more often reported frequent use of drops, capsules, vape oils, edibles, drinks, solid concentrates, hash or kief, and topicals. They also had a higher probability of screening positive for high-risk use.

The surveys measured exposure and outcomes at the same general time. They cannot establish whether prescribing preceded higher-risk patterns, followed them, or simply identified people with more complex or intensive cannabis use.

Self-report, selection into the survey, differences in health status, symptom burden, product access, and other unmeasured factors may influence the observed associations.

The sourcing result is clinically important because prescribed patients may also use cannabis obtained elsewhere. Product composition, potency, route, frequency, and labeling may differ across sources.

A useful clinical history should therefore cover every cannabis product, not only the prescribed item listed in a medication record.

Prescribing encounters can include structured discussion of treatment goals, total cannabis exposure, product potency, impairment, tolerance, withdrawal, unsuccessful efforts to reduce use, interactions, and adverse effects.

A positive high-risk-use screen is not itself a diagnosis. It is a reason for further assessment, shared decision-making, and proportionate follow-up.

Medical and nonmedical cannabis categories can overlap in real life. People may use multiple products for multiple reasons and obtain them through several channels, which complicates both exposure measurement and clinical counseling.

Future longitudinal research should measure baseline risk, symptom severity, product potency, dose, source, changes over time, and clinically assessed cannabis use disorder before and after prescribing.

The most useful message is not that prescriptions are dangerous. It is that a prescription label can create false reassurance if the clinician never asks what else the patient is using.

A careful medical cannabis visit should connect therapeutic goals with a complete product inventory, an impairment and dependence screen, and a plan for monitoring benefit and harm over time.

A Four-Step Reading Frame