CGI-S and CGI-I: Clinical Global Impression Scale Defined

What CGI-S and CGI-I Measure in Ketamine Research

CGI-S (Clinical Global Impression, Severity) and CGI-I (Clinical Global Impression, Improvement) are two linked scores a trained clinician assigns during a ketamine or esketamine trial visit, each on a 1-to-7 scale. CGI-S captures how ill the patient appears right now, judged against the clinician's experience with similar patients. CGI-I captures how much that same patient has changed since the trial's baseline visit. Both numbers come from one clinical judgment rather than a symptom tally, which is why trials usually pair them with item-based tools such as the MADRS score explained in ketamine clinical trials.

The 1-to-7 Scale, Anchor by Anchor

Both scores use the same seven-point range, but the anchors mean different things depending on which scale is in use. On CGI-S, a score of 1 means the patient is not ill at all and a 7 means they are among the most severely ill patients the clinician has treated. On CGI-I, a score of 1 means very much improved since baseline and a 7 means very much worse, with 4 marking no real change. A patient can move from a CGI-S of 6 (severely ill) to a CGI-S of 3 (mildly ill) while earning a CGI-I of 2 (much improved), since CGI-I is always measured against that patient's own starting point rather than an absolute standard.

How Ketamine and Esketamine Trials Use These Scores

CGI-S and CGI-I are almost always reported as secondary or supportive measures rather than a trial's primary endpoint. In the pivotal esketamine (Spravato) trials, the Montgomery-Åsberg Depression Rating Scale (MADRS) served as the primary efficacy measure, while CGI-S added a clinician's overall severity judgment at each visit. According to the FDA's 2019 approval announcement, esketamine nasal spray was cleared for treatment-resistant depression based on trials that reported MADRS score change as the primary outcome, with CGI-S included among the supportive measures reviewed.

Ketamine's dissociative effects create a practical wrinkle for CGI-I in particular. Because active ketamine produces noticeable short-term dissociation, see ketamine dissociation meaning, a rater can sometimes guess which patients received active drug versus a saline or midazolam placebo, even in studies designed to keep raters blinded. Researchers reviewing ketamine trials note this as a limitation when weighing how much of a CGI-I improvement reflects the drug's effect on the glutamatergic system, the excitatory signaling network behind ketamine's effects, versus expectation or rater bias.

Reading CGI Scores as a Patient or Caregiver

If you're reading a published ketamine study or an esketamine trial summary, a CGI-I of 1 or 2 ("very much improved" or "much improved") is the figure most often cited as a response, though trials formally define "response" using MADRS percentage change rather than CGI alone. A CGI-S score at baseline tells you how severely ill the study population was to start, which matters when comparing results across different routes of administration, including the IV ketamine vs oral ketamine literature. For a broader walkthrough of how to interpret these figures in published research, see how to read ketamine depression research studies. This entry is educational, not medical advice; talk with a licensed clinician about what any trial's outcome measures mean for your own treatment decisions.