Clinical Global Impressions Scale (CGI)
Also known as: CGI, CGI-S, CGI-I
The Clinical Global Impressions Scale is a clinician-administered two-part assessment developed by William Guy in the ECDEU Assessment Manual (1976) to provide rapid, global ratings of illness severity and treatment response. Part 1 (CGI-Severity) rates current severity; Part 2 (CGI-Improvement) rates change since treatment initiation. The CGI is among the most widely used global outcome measures in psychiatric research and clinical practice, prized for its brevity, interpretability, and ability to capture clinician expertise and nuanced clinical judgment.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Recommended for: baseline severity and endpoint improvement assessment in antidepressant, anxiolytic, and antipsychotic trials; rapid clinical outcome rating in busy settings (inpatient, outpatient, emergency); complement to objective symptom measures; assessment where clinician judgment is central (complex cases with multiple diagnoses, significant psychosocial factors); treatment monitoring in unblinded, open-label settings where clinician expertise is appropriate.
Strengths & limitations
- Exceptional brevity—two single-item ratings, 1–2 minutes; highly practical for busy clinics and repeated administration
- Captures clinician expertise—integrates clinical judgment, diagnostic acumen, and nuanced observation that symptom checklists cannot fully capture
- Highly responsive to treatment—demonstrates large effect sizes for pharmacological and psychotherapeutic interventions, sensitive to clinically meaningful change
- Widely validated across conditions—used in psychiatry, pediatrics, neurology, rheumatology, and pain medicine; robust across diagnoses and populations
- Free public domain—no licensing restrictions; available in multiple languages
- Complements objective measures—provides clinician perspective alongside patient-reported outcomes and symptom severity scales
- Rater-dependent reliability—lacks structured anchor points; inter-rater reliability varies (r = 0.50–0.85) depending on rater training and shared case experience
- Vulnerable to bias—influenced by rater expectations, patient demographics, treatment allegiance; unblinded raters may inflate improvement ratings
- Global impression may obscure domains—a single severity score does not reveal which symptoms are most problematic or why improvement occurred
- No item content detail—global rating does not specify whether improvements are in mood, anxiety, cognition, function, or other domains; supplementary data necessary
- Ordering effects—raters may anchor on baseline severity and fail to adjust upward or downward when substantial change occurs, particularly in long-term follow-up
- Ceiling and floor effects—severe cases (CGI-S 6–7) may improve substantially but remain at high scores; mild cases may not improve further
Frequently asked
How does CGI differ from a symptom-based measure like HAM-D?
CGI is a clinician's global judgment of overall severity and change; HAM-D is a detailed symptom inventory. CGI is faster (1–2 min vs. 15–30 min) and captures expert judgment, but is less specific about which symptoms changed. HAM-D is more objective and detailed. In trials, both are often used: HAM-D for precise symptom measurement, CGI for clinician-informed global assessment.
Why does my CGI-I rating differ from my colleague's on the same patient?
CGI relies on clinician judgment; different raters weight symptoms, context, and improvement differently. To improve consistency: (1) use standardized interview probes; (2) review symptom measures together before rating; (3) attend rater training workshops; (4) conduct calibration meetings comparing ratings on shared cases.
Can I use CGI-S to diagnose a psychiatric disorder?
No. CGI-S rates severity, not diagnosis. A patient with CGI-S 5 (markedly ill) could have major depression, bipolar disorder, generalized anxiety disorder, or another condition. CGI-S indicates the patient needs treatment; diagnosis requires full psychiatric evaluation.
How do I interpret CGI-I when symptoms improved but patient functioning declined?
Discordance between symptom improvement and functional decline is clinically important. Rating CGI-I ≥5 (improved) based on symptom reduction alone may miss functional deterioration. Use CGI alongside functional measures (Sheehan Disability Scale, GAF) to capture complete clinical picture. Interview the patient: 'Your mood has improved, yet I notice you're withdrawing socially. Tell me about that.'
Sources
- Guy, W. (1976). ECDEU Assessment Manual for Psychopharmacology. Rockville, MD: National Institute of Mental Health, US Department of Health, Education, and Welfare. link ↗
- Busner, J., & Targum, S. D. (2007). The Clinical Global Impressions Scale: applying a research tool in clinical practice. Psychiatry (Edgmont), 4(7), 28–37. link ↗
- Kadouri, A., Corruble, E., & Ly, K. H. (2007). The CGI: assessment of its usefulness in the context of a European multicentric antidepressant drug trial. European Neuropsychopharmacology, 17(6–7), 468–472. link ↗
How to cite this page
ScholarGate. (2026, June 3). Clinical Global Impressions Scale (CGI). ScholarGate. https://scholargate.app/en/clinical-psychology/clinical-global-impressions-scale
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Hamilton Depression Rating ScaleClinical Psychology↔ compare
- Montgomery-Åsberg Depression Rating ScaleClinical Psychology↔ compare
- Patient Global Impression of ChangeClinical Psychology↔ compare
- Patient Health Questionnaire-9Clinical Psychology↔ compare