GAD-7 vs BAI: Comparing Anxiety Assessment Tools
Compare GAD-7 and Beck Anxiety Inventory (BAI) for clinical use. Understand when each anxiety assessment is most appropriate for your practice.
GAD-7HEDISFree
Generalized Anxiety Disorder 7-item Scale
BAI
Beck Anxiety Inventory
Psychometric Properties
| Metric | GAD-7 | BAI |
|---|---|---|
| Internal Consistency (α) | 0.89-0.92 | 0.92-0.94 |
| Test-Retest Reliability | 0.83 | 0.75 |
| Sensitivity | 89% | 85% |
| Specificity | 82% | 81% |
| Discriminant Validity | Good (less somatic focus) | Moderate (somatic overlap) |
When to Use Each Assessment
Routine anxiety screening in primary care
Brief, free, validated in medical settings, less somatic overlap
Panic disorder assessment
Comprehensive coverage of physical panic symptoms
Generalized anxiety disorder monitoring
Specifically designed for GAD; excellent sensitivity to change
Patients with chronic medical conditions
Less somatic focus reduces false positives from medical symptoms
Value-based care and quality reporting
HEDIS-aligned; counts toward anxiety screening measures
Research requiring established norms
Longer publication history with extensive normative data
Key Differences at a Glance
- GAD-7 focuses on cognitive/emotional symptoms; BAI emphasizes somatic symptoms
- GAD-7 is free; BAI requires licensing fees
- GAD-7 better discriminates anxiety from depression
- BAI is more sensitive to panic symptoms
- GAD-7 is HEDIS-aligned for quality reporting
The Bottom Line
The GAD-7 is the preferred choice for most clinical settings due to its brevity, zero cost, HEDIS alignment, and better discrimination from depression and medical symptoms. The BAI is useful when assessing somatic anxiety symptoms or panic, particularly in research contexts.
When to Use Both
Consider using GAD-7 for routine screening and general anxiety monitoring, adding BAI when somatic anxiety symptoms or panic are prominent clinical features that need detailed tracking.
Choosing between them in practice
The GAD-7 is the practical choice for anxiety screening and progress monitoring: seven items, free, scored 0-21, cut-off of 10. The BAI (21 items, 0-63) is weighted heavily toward physiological arousal, which makes it strong for panic presentations and weaker for generalized worry — and it is licensed rather than free.
Score crosswalk
| Severity | GAD-7 (0-21) | BAI (0-63) | Typical clinical action |
|---|---|---|---|
| Minimal | 0-4 | 0-7 | No action; re-screen routinely |
| Mild | 5-9 | 8-15 | Psychoeducation, self-management, repeat in 2-4 weeks |
| Moderate | 10-14 | 16-25 | Positive screen: further assessment and a treatment plan |
| Severe | 15-21 | 26-63 | Active treatment; assess function, avoidance and comorbid depression |
Crosswalks are approximate. Severity bands from different instruments are not interchangeable, so keep one instrument as the measure of record for a given patient.
Switching instruments mid-treatment
- The GAD-7 tracks worry and its interference; the BAI tracks bodily arousal. A patient can improve markedly on one and barely move on the other, so pick the instrument that matches the target of treatment before you start.
- For panic disorder, a BAI drop is often the earlier signal. For generalized anxiety, the GAD-7 is more sensitive to change.
- Administer both once at the transition visit if you are changing instruments mid-episode, then baseline on the new one.
- Pair either instrument with the PHQ-9. Anxiety and depression co-occur in roughly half of presentations, and treating one while unmeasured symptoms of the other persist is the most common cause of a plateau.
Evidence notes
- GAD-7 validation: Spitzer, Kroenke, Williams & Löwe (2006), Arch Intern Med 166(10):1092-1097. Sensitivity 89%, specificity 82% at a cut-off of 10 for GAD.
- BAI manual: Beck & Steer (1993). Internal consistency 0.92; designed specifically to discriminate anxiety from depression.
- A 4-point GAD-7 change is the commonly used minimum clinically important difference in primary care samples.
Frequently asked questions
BAI vs GAD-7 — which should a clinic standardize on?
The GAD-7 for almost every general behavioral health or primary care clinic: it is free, takes about a minute, is validated for both screening and monitoring, and is accepted for quality reporting. Reserve the BAI for settings where panic and somatic arousal are the main clinical target.
Does the GAD-7 detect panic disorder or social anxiety?
It screens broadly and is reasonably sensitive to panic (74%) and social anxiety (72%), but it was validated for generalized anxiety disorder. A positive GAD-7 should prompt a diagnostic conversation, not a specific anxiety diagnosis.
Is the BAI free?
No. The BAI is copyrighted and licensed per administration. The GAD-7 is free for clinical use, which matters when you are screening every patient at every visit.
How often should anxiety be re-measured during treatment?
Every 2-4 weeks during active treatment is the usual cadence, and monthly during maintenance. Anything less frequent than monthly makes it hard to distinguish real change from ordinary week-to-week variation.
Ready to Use GAD-7 or BAI?
Download printable forms or start tracking patient outcomes with TouchpointHQ.
5 Free Patients Forever
No credit card required • Full feature access