Myndgard

Blog · · 5 min read

The PHQ-9 and GAD-7, explained plainly

What the PHQ-9 and GAD-7 measure, how they are scored, and why they screen and monitor rather than diagnose.

By Myndgard Limited

Banner reading: The PHQ-9 and GAD-7, explained plainly

The PHQ-9 and the GAD-7 are among the most familiar questionnaires in mental health care. Many students meet them at a first counselling appointment or a GP visit. They are short and easy to complete. That familiarity makes it easy to forget what they were designed for, and what they were not.

This post is a plain guide for counselling staff, and for students who want to understand the questions they are being asked.

Where they came from

The PHQ-9 (Patient Health Questionnaire, nine items) was described by Kurt Kroenke, Robert Spitzer and Janet Williams in the Journal of General Internal Medicine in 2001. It was tested with 6,000 patients in US primary care and obstetrics-gynaecology clinics, 580 of whom were also interviewed by a mental health professional. Its nine items follow the nine symptoms used to diagnose major depression.

The GAD-7 (Generalized Anxiety Disorder, seven items) followed in 2006 from the same group, with Bernd Löwe, in Archives of Internal Medicine. It was developed with 2,740 adults across 15 US primary care clinics, 965 of whom also had a telephone interview with a mental health professional.

Both were built for busy general medical settings, where a few minutes is all there is.

How they are scored

Both questionnaires ask about the last two weeks. Each item is scored from 0 to 3:

  • 0 = not at all
  • 1 = several days
  • 2 = more than half the days
  • 3 = nearly every day

The PHQ-9 total runs from 0 to 27. Kroenke and colleagues wrote that scores of 5, 10, 15 and 20 "represented mild, moderate, moderately severe, and severe depression, respectively".

The GAD-7 total runs from 0 to 21. Spitzer and colleagues suggested cut points of 5, 10 and 15 "might be interpreted as representing mild, moderate, and severe levels of anxiety".

A score of 10 or more is the most common threshold for "worth a closer look" on both. In the original studies, a PHQ-9 score of 10 or more had 88% sensitivity and 88% specificity for major depression, and a GAD-7 score of 10 or more had 89% sensitivity and 82% specificity for generalised anxiety disorder. A larger 2019 BMJ meta-analysis by Levis and colleagues, using individual data from 58 studies and 17,357 participants, again found that a PHQ-9 cut-off of 10 or above gave the best balance of sensitivity and specificity.

What they are, and what they are not

These questionnaires are good at two jobs:

  • Screening: flagging that someone may be struggling enough to warrant a proper conversation.
  • Monitoring: showing whether symptoms are moving up or down over time, using the same yardstick each time.

They are not a diagnosis. A score is a summary of what someone ticked on a given day about the previous fortnight. It does not know about bereavement, exam pressure, a recent break-up, or a physical illness, and it cannot weigh those things. That is the job of a clinical conversation.

The student evidence makes this point sharply. A 2022 study by Pranckeviciene and colleagues in PLOS ONE gave both questionnaires to 1,368 Lithuanian university students (average age 22.5) and followed up 560 of them with a structured clinical interview. Both scales were internally reliable. But a PHQ-9 score of 10 or more had only 71% sensitivity and 66% specificity for identifying students at increased risk of a mood or anxiety disorder, and the best GAD-7 cut-off (9 or more) gave 73% and 70%. The authors concluded that "due to low specificity and high false positive rates, both scales are recommended only as an initial screening tool."

In other words, among students, a raised score is a reason to talk, not a conclusion. Thresholds tuned in primary care may also not transfer neatly to a younger, high-pressure population.

Item 9, and why it needs a human

The ninth PHQ-9 item asks how often, over the last two weeks, someone has been bothered by thoughts that they would be better off dead, or of hurting themselves in some way.

Research from large US health systems shows why this item matters. In a 2013 study by Simon and colleagues of 84,418 outpatients, the one-year risk of a suicide attempt rose from 0.4% among people who answered "not at all" to 4% among those who answered "nearly every day". The link held after accounting for overall depression severity.

The same research group's 2016 follow-up of over half a million adults showed the other side: 39% of suicide attempts within 30 days of completing the questionnaire were by people who had answered "not at all" to item 9.

Two things follow:

  1. A raised item 9 answer should always lead to a timely, human follow-up. Someone has taken the step of saying it, and they deserve a response.
  2. A zero on item 9 is not an all-clear. It should never replace asking directly and listening.

This is consistent with NICE guideline NG225 on self-harm (2022), which says: "Do not use risk assessment tools and scales to predict future suicide or repetition of self-harm," and asks professionals to focus instead on the person's needs and their safety. Item 9 is a prompt to have that conversation, not a risk score.

In practice

For counselling services using the PHQ-9 and GAD-7:

  • Use them to open conversations and track change, not to label or to ration access.
  • Explain to students what the questions are for, and show them their own scores over time.
  • Be cautious about fixed thresholds with students; treat a raised score as a reason to ask more.
  • Make sure every positive answer to item 9 reaches a named person who will follow up promptly, and never treat a zero as proof of safety.

In Myndgard Share, for example, an answer that indicates risk alerts the student's assigned clinician so that a person can follow up. The questionnaire's role ends there; the clinician's begins.

If you're struggling: Samaritans 116 123, text 50808, or 112/999 in an emergency.

Sources