Medicinal cannabis for psychiatry-related conditions: an overview of current Australian prescribing

Abstract

Objective: Evidence is accumulating that components of the Cannabis sativa plant may have therapeutic potential in treating psychiatric disorders. Medicinal cannabis (MC) products are legally available for prescription in Australia, primarily through the Therapeutic Goods Administration (TGA) Special Access Scheme B (SAS-B). Here we investigated recent prescribing practices for psychiatric indications under SAS-B by Australian doctors. Methods: The dataset, obtained from the TGA, included information on MC applications made by doctors through the SAS-B process between 1st November 2016 and 30th September 2022 inclusive. Details included the primary conditions treated, patient demographics, prescriber location, product type (e.g., oil, flower or capsule) and the general cannabinoid content of products. The conditions treated were categorized according to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition, text revision (DSM-5-TR). Trends in prescribing for conditions over time were analyzed via polynomial regression, and relationships between categorical variables determined via correspondence analyses. Results: Approximately 300,000 SAS-B approvals to prescribe MC had been issued in the time period under investigation. This included approvals for 38 different DSM-5-TR defined psychiatric conditions (33.9% of total approvals). The majority of approvals were for anxiety disorders (66.7% of psychiatry-related prescribing), sleep-wake disorders (18.2%), trauma- and stressor-related disorders (5.8%), and neurodevelopmental disorders (4.4%). Oil products were most prescribed (53.0%), followed by flower (31.2%) and other inhaled products (12.4%). CBD-dominant products comprised around 20% of total prescribing and were particularly prevalent in the treatment of autism spectrum disorder. The largest proportion of approvals was for patients aged 25-39 years (46.2% of approvals). Recent dramatic increases in prescribing for attention deficit hyperactivity disorder were identified. Conclusion: A significant proportion of MC prescribing in Australia is for psychiatry-related indications. This prescribing often appears somewhat "experimental", given it involves conditions (e.g., ADHD, depression) for which definitive clinical evidence of MC efficacy is lacking. The high prevalence of THC-containing products being prescribed is of possible concern given the psychiatric problems associated with this drug. Evidence-based clinical guidance around the use of MC products in psychiatry is lacking and would clearly be of benefit to prescribers.

Keywords: Australia; anxiety disorders; medicinal cannabis; medicinal cannabis use; prescribing; psychiatry.

FIGURE 1

Associations between age, product schedule and type, and indication. Correspondence analyses between age and indication (A), indication and product schedule and type (B), and age and product schedule and type (C). Deviation from independence described by the dimensions on each axis (Dim 1 and Dim 2), with the scaled contribution to the overall variance depicted by the inertia*1,000 (red to blue color gradient). See Supplementary Table S6 for related statistics.

FIGURE 2

Number of SAS-B approvals per month for psychiatric indications from November 2016 to September 2022 (n = 100,666). The solid line represents the best fit, with shading depicting standard error of the mean (SEM).

FIGURE 3

Approvals per month in psychiatric indication categories with >100 approvals followed different patterns of prescribing growth. Approvals over time for anxiety disorders [(A), n = 67,133]; sleep wake disorders [(B), n = 18,321], trauma- and stressor-related disorders [(C), n = 5,799] neurodevelopmental disorders [(D), n = 4,450]; depressive disorders [(E), n = 4,003]; neurocognitive disorders [(F), n = 428]; bipolar and related disorders [(G), n = 212]; disruptive, impulse-control, and conduct disorders [(H), n = 155]; and substance-related and addictive disorders [(I), n = 126]. Solid lines represent the best fit, with shading depicting standard error of the mean (SEM).

FIGURE 4

Approvals per month in psychiatric indications with >100 approvals. Approvals over time for anxiety [(A), n = 67,095]; sleep disorder [(B), n = 11,202)], insomnia [(C), n = 6,877); post-traumatic stress disorder [PTSD; (D), n = 5,799]; depression [(E), n = 3,247]; autism spectrum disorder [ASD; (F), n = 2,206]; attention deficit hyperactivity disorder [ADHD; (G), n = 2,078]; mood disorder [(H), n = 736]; Alzheimer’s disease [(I), n = 272]; restless leg syndrome [RLS; (J), n = 239]; bipolar disorder [(K), n = 212]; Tourette’s syndrome [(L), n = 163]; unspecified dementia [(M), n = 132]; behavior disorder [(N), n = 131]; cannabis use disorder [(O), n = 117]. Solid lines represent the best fit, with shading depicting standard error of the mean (SEM).

FIGURE 5

Patients receiving medicinal cannabis for psychiatric indications are predominantly younger and male. Trends in patient sex (A,B) and age (C,D). Approval trends over time showing a recent decrease in the rate of approvals for males (A), but continued growth in young patients, particularly aged 25–39 (C). The proportion of these changes is also shown (B,D), and suggests that while the number of male prescriptions may be decreasing, the relative proportion of prescribing remains relatively consistent. The lines of best fit in panels (A,B) are shown by the solid line with shaded area showing standard error of the mean. The gap in panels (B,D) indicates no applications submitted during this period.

FIGURE 6

Approvals for medicinal cannabis products for the treatment of psychiatric indications are largely THC-containing oil or flower products. Trends in product schedule (A,B) and type (C,D). Approval trends over time showing continued growth of S8 access (A), which is reflected in the proportional access (B). The rate of approvals for the top three product types (oil, flower, and inhaled) all appear to be decreasing (C). However, oil and flower at least seem to have consistent proportional approvals (D). The lines of best fit in panels (A,B) are shown by the solid line with shaded area showing standard error of the mean. The gap in panels (B,D) indicates no applications submitted during this period.