We’ve Been Assessing Depression All Wrong: A Simple Follow-Up Question for Therapists
By Dr Hilah Kaufman, Clinical Psychologist and Clinical Product Manager at NovoPsych
The Patient Health Questionnaire‑9 (PHQ-9) is one of the most widely used depression measures worldwide. But a new study published in JAMA Psychiatry raises an important question.
What if many people are not interpreting the PHQ instructions the way the test developers intended? And what can clinicians do when they review the results?
Two Words That Change Everything
The PHQ-9 (and all its variants) instructions ask: “How often have you been bothered by the following problems?” followed by a list of depression symptoms.
It is intended to capture how much symptoms bother someone, not simply how often they occur. That distinction is the whole point of the scale.
Instead, most people simply count how often the symptom occurred, ignoring the “bothered by” part entirely. That gap between intended and actual interpretation can mean the difference between the lowest and highest possible score on the same item.
Administer the PHQ-9 in NovoPsych with instant scoring, visual reports, and track your patients’ progress over time.
What the Research Found
Researchers tested more than 800 people across general and clinical samples. In the general population, only 55% interpreted the PHQ as intended. Among people with depression, that dropped to just 16%.
Even more striking: even after the distinction was explained, about 90% of people with depression said they would still answer the same way in the future.
And this matters clinically. The researchers found that people with identical PHQ scores showed meaningfully different levels of depression on a separate validated measure, simply because they had interpreted the instructions differently.
What clinicians should take from this
The takeaway is not that clinicians should stop using the PHQ.
It remains one of the most useful and accessible depression measures available. But it does highlight an important limitation. Self-report scores only mean what we think they mean if people interpret the questions as intended.
A simple follow-up question can be helpful in practice: Were you counting how often the symptom happened, or how often it bothered you?
That clarification can reveal a lot about how a client approached the questionnaire.
A simple follow-up question can be helpful in practice: Were you counting how often the symptom happened, or how often it bothered you?
This study also points to a broader issue in mental health assessment. Psychometric research focuses heavily on reliability and statistical validity, but rarely asks a more basic question: do people actually understand the instructions the way researchers expect them to?
NovoPsych and Assessment Quality
At NovoPsych, we spend a lot of time thinking about how psychological measures are constructed and interpreted. Measurement-based care only works if the tools we use are clearly understood and used thoughtfully.
When we develop measures in-house, instruction clarity and response format are not afterthoughts. They are central to whether a tool actually measures what it claims to measure.
We also create carefully researched profiles for every measure in our library, so clinicians have the context they need to use and interpret results with confidence. You can administer and automatically score the PHQ-9, alongside 150+ other validated assessments, directly in NovoPsych.
Reference: Panayiotou, M., Razum, J., Eisele, G., Wang, S. B., Fried, E. I., & Cohen, Z. D. (2025). Interpretation issues with the Patient Health Questionnaire instructions. JAMA Psychiatry. https://doi.org/10.1001/jamapsychiatry.2025.3796