The aim of this study was to evaluate the applicability and validity of a clinical global impressions-aggression (CGI-A) scale in retrospective trials. CGI-A scores were derived from clinical records, BPRS scores were assessed prospectively in 101 patients with first-episode psychosis. CGI-A showed highly significant correlations to the brief psychiatric rating scale (BPRS) excited component (BPRS-EC; r=0.32; p=0.001) and BPRS resistance (r=0.25; p=0.011), to aggression at admission (r=0.70; p<0.001) and to involuntary treatment (r=0.55; p<0.001). The CGI-A retains high construct and face validity in retrospective trials.
