Calibrated clinical judgment in psychiatric drug trials: Beyond psychometric scales

Psychiatric randomised controlled trials increasingly privilege psychometric rating scales, yet bedside clinical training, remains the clinician's primary integrator of meaning, context, and change. Recent reconceptualisations describe clinical judgment as a three-stage process encompassing data collection through clinical interviewing, interpretationviaclinical reasoning, and decision making (Fava and Guidi, 2026; Fava et al., 2026). We provide a historically informed, conceptually focused analysis of the transition from narrative clinical impression to operationalised criteria and multi-item scales, and examine the emergence of clinician-judgment outcomes, particularly the Clinical Global Impressions (CGI) scales, as a hybrid bridge between these traditions. We argue that modern outcome assessment comprises two dominant families: multi-item symptom scales (e.g., PANSS, MADRS, HAM-A) and global clinician-judgment scales (CGI-Severity, CGI-Improvement). Global ratings can capture clinically salient information not fully represented by item totals, but only if their use preserves independent judgment. We propose "CGI 2.0": minimum standards for rater qualification, construct-focused training, structured justification, calibration/monitoring, and prespecified concordance/discordance analyses with primary symptom-scale endpoints, while cautioning against "industrialised" practices such as deriving CGI scores from multi-item totals. A clinimetric integration of psychometric methods that re-legitimises calibrated clinical judgment alongside rigorous measurement could improve interpretability, clinical relevance, and signal detection in psychopharmacology trials.