**
In a revealing new analysis published in JAMA Network Open, researchers have cast a shadow over the reliability of mental health diagnostic interviews, a tool often regarded as the “gold standard” for identifying conditions such as depression, anxiety, and substance use disorders. Led by Laura Duncan, a psychiatry professor at McMaster University in Ontario, Canada, the study highlights significant discrepancies in the reliability of these assessments depending on the specific mental health condition being evaluated.
The Gold Standard Under Scrutiny
Diagnostic interviews have long been the primary method used by clinicians and researchers to assess mental health disorders. However, Duncan notes that these interviews do not consistently provide a robust benchmark of validity and reliability. “While they are often viewed as the best available approach, this may be more a reflection of a lack of superior alternatives rather than their efficacy,” she explains.
The study synthesises findings from various investigations into the “test-retest reliability” of these interviews, covering research conducted from February 2024 to September 2025. Using Cohen’s kappa coefficient, the authors evaluated how consistently patients received the same diagnosis when subjected to the same interview on two separate occasions.
Variability Across Conditions
The results of the study revealed that the reliability of diagnostic interviews varies significantly across different mental health conditions. Notably, the highest reliability was observed in substance use disorders, particularly opioid use disorder. Duncan attributes this to the measurable nature of substance use criteria, which are often based on concrete behaviours. “It’s simply easier for individuals to recall how many drinks they consumed in a week compared to quantifying their feelings of sadness or anxiety,” she notes.
Conversely, conditions like depression or anxiety proved to be less reliably diagnosed, reflecting the subjective nature of these mental health issues. This inconsistency raises critical questions about the effectiveness of current diagnostic practices.
Calls for Improved Standards
Dr Michael First, a psychiatrist at Columbia University and the creator of the Structured Clinical Interview for DSM-5 (SCID), voiced his concerns regarding the study’s findings. While he agrees that the variability in diagnostic interviews is troubling, he believes the analysis lacks specificity regarding which diagnostic tools yield the most reliable results. “It would be beneficial to identify specific instruments that are more effective,” First comments, emphasising the need for clearer guidance in the field.
Duncan acknowledges the limitations of the existing research, stating that the available data often fails to clearly differentiate between fully structured and semi-structured interviews. Fully structured interviews, like the SCID, adhere strictly to a script, which can lead to more consistent outcomes. In contrast, semi-structured interviews allow for greater flexibility, enabling clinicians to explore responses further. This freedom can enhance diagnostic accuracy, but it may also introduce variability in patient responses across different sessions.
The Path Forward
The study reveals a significant gap in rigorous research surrounding psychiatric diagnoses. Duncan highlights the necessity for improved methodologies in assessing mental health conditions. “The lack of comprehensive information on the formats of different interviews points to a broader issue within psychiatric research,” she asserts.
Both Duncan and First express a desire for the future of mental health diagnostics to evolve beyond rigid categories. They envision a shift towards understanding mental health on a spectrum rather than as binary conditions, which could lead to more accurate and compassionate care.
Why it Matters
As the conversation around mental health continues to gain traction globally, this study shines a light on the pressing need for more reliable diagnostic tools. The implications of inaccurate diagnoses can be profound, affecting treatment plans and outcomes for millions of individuals. By addressing these discrepancies and pushing for more research, the mental health community can work towards a future where diagnosis is not just more accurate, but also more reflective of the complex and nuanced nature of human emotions and behaviours.
