Clinical Global Impression - Improvement
CGI-I is a clinician rating of change from a previous clinical frame. This page is for clinics. It is not a patient questionnaire.
Overview
What is the CGI-I?
CGI-I asks how much the person's condition has improved or worsened relative to a baseline. Without a shared baseline and rater convention, the rating is hard to interpret.
Development and clinical use
CGI-I accompanies CGI-S in the classic CGI pair used in trials and some services. It is a judgement of change, not a computed difference between two symptom totals.
- Rater
- Clinician only
- Companion
- CGI-S
- Requires
- A baseline frame
- Not a computed delta
- Global judgement
- Clinical role
- Change anchor
- Diagnosis
- Not established by rating
Administration
A focused clinician rating
A clinician who knows the baseline frame rates CGI-I. It should not be emailed to clients as a self-report of improvement.
- One rated item at a time with visible progress
- Keyboard-operable response choices
- Previous, next, and review steps for the rater
- Clear confirmation that the rating is stored for clinic review
The illustration shows a rater control, not the official CGI wording.
Report showcase
Designed to make evidence easy to review
The downloadable synthetic example illustrates report layout using invented response patterns, without patient data or instrument questions.
- Named instrument, version, and provenance
- Item-level response pattern without source wording
- Methodology, references, and clinical boundary
- Responsible clinician review rather than automated conclusions
Clinician review required
Methodology
Understand the measure and its scoring conventions
Official seven-point improvement anchors exist. This page does not independently verify a production scoring rule.
References
- CGI in ECDEUGuy W. ECDEU Assessment Manual for Psychopharmacology. 1976.
Questions
What clinics ask before they evaluate the CGI-I
These answers stay inside educational catalogue scope. They do not replace official manuals or clinician judgement.
Is CGI-I a patient satisfaction item?
No. It is a clinician global impression of change.
Can it replace PHQ-9 change scores?
No. It is a different kind of evidence. They are not interchangeable.
Can I complete CGI-I here?
No. This landing is educational.
What if there is no baseline?
Then an improvement rating has no shared frame. That is a protocol problem, not something software should invent.