Are interviews suitable for diagnosing mental health conditions?
What's the story
A recent study published in JAMA Network Open has raised concerns over the reliability of diagnostic interviews, the most common method for diagnosing mental health conditions. The research was done by Laura Duncan, a psychiatry professor at McMaster University in Ontario, Canada. She noted that while these interviews are often seen as a "gold standard" for assessing mental disorders in clinical and research settings, they don't provide a definitive benchmark demonstrating excellent validity and reliability.
Reliability concerns
Interviews seen as 'gold standard' for mental health assessment
Duncan highlighted that despite mixed evidence on the reliability of diagnostic interviews, they are still widely regarded as the best available method. This could be due to a lack of better alternatives. The study reviewed data from studies on "test-retest reliability" of these interviews from February 2024 to September 2025. It used Cohen's kappa coefficient to estimate their reliability for different mental health conditions, accounting for random coincidences in diagnoses.
Disorder reliability
Reliability better for substance use disorders
The study found that the average reliability was generally better for substance use disorders, with opioid use disorder showing the highest overall reliability. Duncan explained this is because criteria for substance use disorders are mainly behavior-based. For example, it's easier to quantify how many drinks one had in a week than how many days one felt sad or anxious.
Tool transparency
Psychiatrist calls for more specific data on interview types
Dr. Michael First, a psychiatrist and professor at Columbia University who developed Structured Clinical Interview for DSM 5, expressed frustration with some aspects of the study. He agreed that diagnostic interviews vary in reliability and often fail to correctly diagnose people. However, he wanted more information about which specific instruments were most reliable. "It'd be nice to be able to look at this and say: 'Oh, based upon this paper, I should pick this one because of this,'" he said.
Interview classification
Criticism over lumping different types of interviews together
First also criticized the study for grouping "fully structured" and "semi-structured" interviews together. Fully structured interviews are more likely to yield consistent results when repeated, as they follow a strict script without deviation. These are often used for large population studies and designed for people with minimal training. Semi-structured interviews, on the other hand, allow trained clinicians more flexibility in their questioning, which can lead to more accurate diagnoses but also variability in patient responses over time.