You may have seen recent headlines claiming medicinal cannabis is “ineffective” for common mental health following a published study by The Lancet Psychiatry. While this is a significant piece of research, it reflects only part of a much broader and more complex picture.
To really understand what this means, we need to look a little more closely at how evidence works, and how it is used in real-world care.
1. “Absence of Evidence” is not “Evidence of Absence”
The most important nuance in this study is a classic scientific distinction. The researchers concluded there is no high-quality evidence that cannabis works for certain conditions—but “no evidence” is not the same as “no effect.”
- Media Interpretation: “Cannabis doesn’t work.”
- Clinical Reality: “We haven’t done enough high-quality, large-scale, placebo-controlled trials yet to prove it works to a rigid pharmaceutical standard.”
- The Takeaway: Most studies analysed were small, short-term, or used varying dosages. This paper is a call for better-funded, rigorous research, rather than a final verdict that the plant lacks therapeutic value.
So rather than a final answer, this study is better seen as a signal that we need better research.
It’s also worth remembering that research evidence is just one part of evidence based medicine. Clinical experience of the prescriber and what matters to the patient are just as important.
2. The “Average” vs. The “Individual”
Meta-analyses (like this Lancet study) look at population averages. If 100 people take a medication and 50 get significantly better while 50 see no change, the “average” result can appear statistically insignificant.
- The Nuance: Mental health is highly subjective. While the “average” person in a rigid trial might not show a massive shift in a primary diagnosis like Major Depressive Disorder (MDD), many patients experience Quality of Life improvements that these trials aren’t designed to measure.
- The Reality: Medicinal cannabis often targets the “ripples” of mental health—such as physical tension, agitation, and poor sleep. When these secondary symptoms improve, a patient’s ability to manage their primary condition often improves too.
3. Safety in Context: The “Number Needed to Harm”
The study also looked at safety, using something called the “Number Needed to Harm” (NNTH) . In clinical research the NNTH tells you how many people you need to treat with a specific medication for one additional person to experience a side effect or “harm”.
In reviewing the safety metrics provided in the study:
- The Findings: For cannabinoids, the NNTH was 7. This means for every 7 people who take it, one might experience a side effect compared to a placebo.
- The Context: Crucially, the “harm” or side effects were minor (like dry mouth or fatigue). The study found no higher risk of serious harm or people dropping out of treatment compared to the placebo groups. No treatment is completely free of side effects.
For perspective, many traditional psychiatric medications have much higher rates of side effects that are severe enough to make patients stop treatment entirely.
4. Research vs Real-World Care
Randomised Controlled Trials (RCT) are often called the gold standard in research, and for good reason. They’re designed to tightly control variables so we can be confident that any effect is actually due to the treatment. But they also happen in very controlled environments, with strict inclusion criteria that doesn’t always reflect the complexity of life.
- The Australian Landscape: In Australia, over 1 million applications have been approved via the Authorised Prescriber (AP) and Special Access Scheme (SAS-B).
- Clinical Justification: These pathways exist because the TGA acknowledges that for many patients, traditional treatments have failed. While RCTs provide the essential evidence pillar of the Evidence-Based Medicine triad, holistic care also requires clinical expertise and a deep understanding of the patient’s unique values and history, individual factors that population-wide averages in a meta-analysis are not designed to capture.
RCTs give us essential information, but they are only one part of the picture.
5. Where We Do See Benefits
It’s important to note that the study did find evidence of benefit in areas that often overlap with mental health:
- Insomnia: Low-to-moderate certainty evidence for improved sleep.
- Autism: Reductions in specific symptoms associated with ASD.
- Symptom Management: The researchers acknowledged cannabis’s proven role in treating chronic pain and spasticity, both of which are major contributors to anxiety and depression.
These are all closely connected to mental health. When sleep improves, or pain is better managed, mental wellbeing often improves alongside it.
Putting it all together
This isn’t a simple ‘good’ or ‘bad’ story. The evidence on medicinal cannabis is still emerging, and while research is important, it’s only one piece of the puzzle.
At Astrid, we believe care happens holistically – it’s one part research, one part clinician’s expertise and experience, and one part patient’s needs and values. We advocate for a fuller, more holistic understanding of cannabis.
References
Primary Study
Wilson, J., et al. (2026). Cannabinoids for the treatment of mental disorders and substance use disorders: A systematic review and meta-analysis. The Lancet Psychiatry, 13(3). https://doi.org/10.1016/S2215-0366(26)00015-5
Media & Industry Analysis
Proudfoot, A. (2026, March 18). Lancet clinical review paper: Why absence of evidence is not evidence of absence. Cannabiz. https://www.cannabiz.com.au/lancet-clinical-review-paper-why-absence-of-evidence-is-not-evidence-of-absence/
The Guardian. (2026, March 16). Cannabis is not an effective treatment for common mental health conditions, says review. The Guardian. https://www.theguardian.com/society/2026/mar/16/cannabis-inot-effective-treatment-for-common-mental-health-conditions-review
The University of Sydney. (2026, March 20). Does medicinal cannabis work for depression, anxiety or PTSD? Our study says there’s no evidence. The University of Sydney News. https://www.sydney.edu.au/news-opinion/news/2026/03/20/does-medicinal-cannabis-work-for-depression–anxiety-or-ptsd–ou.html
Thomson, A. (2026, March 13). The $300-a-week medicine Daniel feels better without. The Sydney Morning Herald. https://www.smh.com.au/national/the-300-a-week-medicine-daniel-feels-better-without-20260313-p5oa5j.html
Wilson, J. (2026, March 17). Does medicinal cannabis work for depression, anxiety or PTSD? Our study says there’s no evidence. The Conversation. https://theconversation.com/does-medicinal-cannabis-work-for-depression-anxiety-or-ptsd-our-study-says-theres-no-evidence-278303
Clinical Context References
Australian Medical Association (AMA). (2026, March 17). Lancet paper released on medicinal cannabis effectiveness. https://www.ama.com.au/articles/lancet-paper-released-medicinal-cannabis-effectiveness
Royal Australian College of General Practitioners (RACGP). (2026, March 17). ‘No evidence’ cannabis works for most mental health disorders: Review. newsGP. https://www1.racgp.org.au/newsgp/clinical/no-evidence-cannabis-works-for-most-mental-healthDjuric, A., & Djuric, Z. (2025). Evidence-based medicine: Past, present, future. Journal of Clinical Medicine, 14(14), 5094. https://doi.org/10.3390/jcm14145094
