Insight has emerged as a potential predictor variable in cognitive behavioral therapy for psychosis (CBTp). However, previous research has produced mixed results. The present study aimed to clarify whether symptom type is a moderating variable. A group of psychotic patients (n 1⁄4 44) were assessed through pre- and post-treatment in a CBTp specialty track in a partial hospital-based program in the USA. The Insight Scale was used to measure insight, and psychotic symptomatology was assessed using the Mini-International Neuropsychiatric Interview and the Behavior and Symptom Identification Scale. Patients showed a significant decrease in psychotic symptom scores over the course of treatment [t(43) 1⁄4 3.59, p , .001, Cohen’s d 1⁄4 .64]. Furthermore, illness awareness was specifically associated with a decrease in psychotic symptoms for patients who endorsed visual hallucinations (r 1⁄4 2.68, p , .01), auditory hallucinations (r 1⁄4 2.49, p 1⁄4 .01), and/or ideas of reference (r 1⁄4 2.66, p , .01). Insight did not confer additional benefit for patients with paranoid delusions, mind reading, or thought insertion symptoms. These results are discussed in relation to treatment implications within the current US health care delivery system.
