Maintenance

How accurate is the GAD-7 for measuring anxiety?

Last updated

Short answer

It is accurate enough to trust as a quick screen. In the study that validated it, a GAD-7 score of 10 or higher flagged generalized anxiety correctly about 89% of the time, and cleared people without it about 82% of the time. Scores of 5, 10, and 15 mark mild, moderate, and severe anxiety.

The research
Spitzer, R. L., Kroenke, K., Williams, J. B. W., & Löwe, B. (2006). A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7. Archives of Internal Medicine, volume 166, issue 10, pages 1092 to 1097.
Sample About 2,740 primary care patients completed the GAD-7, and 965 of them also had a separate interview with a mental health professional that served as the yardstick.
Published 2006
Journal Archives of Internal Medicine
Robert Spitzer, Kurt Kroenke, Janet Williams, and Bernd Löwe built the GAD-7, a seven-item questionnaire a person fills out about anxiety symptoms over the past two weeks. Each item is scored from 0 to 3, and the seven add up to a total from 0 to 21. About 2,740 adults across primary care clinics completed it, and 965 of them also sat for a separate interview with a mental health professional, which served as the yardstick. Against that yardstick, a total score of 10 or more identified generalized anxiety disorder with 89% sensitivity and 82% specificity. In plain terms, when the interview found an anxiety disorder the score caught it about 89% of the time, and when the interview found none the score agreed about 82% of the time. The authors marked scores of 5, 10, and 15 as the points for mild, moderate, and severe anxiety. The scale held together well on repeat testing and lined up with other anxiety measures, and as scores rose, people reported worse daily functioning and more disruption to work and daily life. The authors concluded the GAD-7 is a reliable and valid measure of anxiety severity, brief enough for everyday care.
Read the paper

What does this mean for your clinic?

Accuracy is the right thing to ask about, because your clinic acts on the number. A GAD-7 score decides whether a patient gets flagged for anxiety, whether a plan changes, and whether a case looks better or worse than last month. So the first question is simple: does the number reflect what is really going on?

The validation study says it mostly does. Measured against a trained interviewer, a score of 10 or higher caught generalized anxiety about 89% of the time, and cleared people who did not have it about 82% of the time. That is strong for a form a patient finishes in about two minutes. It is not perfect. It misses roughly one in nine people who do have an anxiety disorder, and it flags about one in five who do not, so a score is a signal to look closer, not a verdict on its own.

The three bands are what make the GAD-7 useful for tracking over time, not just for a one-time check. Because 5, 10, and 15 map to mild, moderate, and severe, a score you collect this month can be compared with the one from last month on the same scale. A number that drifts up is an early sign a patient may be struggling more, and a validated severity measure is what lets your care managers see that drift before it becomes a crisis visit. The study backs the scale up: as totals rose, daily functioning fell, so the number is tracking something real in a person's life.

Now the honest limits. The GAD-7 measures the severity of reported anxiety symptoms and screens for a likely disorder. It does not diagnose on its own, and it only reflects what a person is willing and able to put down on the day they fill it out. A cutoff is a trade-off: set it lower and you catch more real cases but also more false alarms. The headline numbers come from primary care patients in the United States around 2006, and later studies in different settings have sometimes found the same cutoff less accurate, so your own population may read a little differently. Treat the score as a careful estimate that a clinician confirms, not as the final word.

This is also where the gap between visits shows up. A score taken only in the room tells you how a patient was doing on one day every few weeks. The same validated questionnaire, collected between visits, can show a rising score while there is still time to act on it.

The takeaway

Terms in this page
Between-session support Mental readiness

This is the work Asendar™ was built for: collecting the same validated scores between visits and keeping them in front of your team, so a rising number shows up while there is still time to act, not at the next appointment.

Start a 15-day free trial for your clinic

Track CoCM minutes, catch patients at risk of missing the code, export a billing packet. No card to start, $199/mo after the trial.

People also ask

What GAD-7 score means anxiety?

In the validation study by Spitzer and colleagues (2006), scores of 5, 10, and 15 mark mild, moderate, and severe anxiety. A score of 10 or more was the usual cut point for a likely generalized anxiety disorder, measured against a clinician interview.

Does the GAD-7 diagnose anxiety?

No. It measures the severity of reported anxiety symptoms and screens for a likely disorder. Spitzer and colleagues validated it against a clinician interview, which is still the step that confirms a diagnosis. A score points your team to look closer, it does not settle the question.

How is the GAD-7 different from the PHQ-9?

They are companion tools from the same research group. The GAD-7 rates anxiety over the past two weeks, while the PHQ-9 rates depression. Both are brief self-reports scored on fixed thresholds, and many clinics use them together to track anxiety and mood side by side.

Explore more in the Maintenance pillar