Treatment Resistance: A Time-Based Approach for Early Identification in First Episode Psychosis.

Although approximately 1/3 of individuals with schizophrenia are Treatment Resistant (TR), identifying these subjects prospectively remains challenging. The Treatment Response and Resistance in Psychosis working group defines <20% improvement as an indicator of TR, though its utility in First Episode Schizophrenia (FES) remains unknown. In a prospective cohort of FES ( = 129) followed up for 5 years, we evaluated two improvement thresholds for 'probable TR'; <20% and <50% based on positive, negative, and total symptoms. We ascertained (1) the ecological validity (i.e., the ability to identify an expected subgroup of 1/3rd of patients); (2) the predictive validity (i.e., ability to predict poor global functioning) and (3) the clinical utility (association with clozapine use at the 5th year). Using the criteria of a total symptom reduction of <50% or negative symptom reduction of <20% resulted in 'probable TR' rates of 37% and 33%, respectively. Using <20% positive or total symptoms criteria resulted in very low rates, indicating minimal utility in FES. <50% total symptom criterion best predicted the global functioning over 5 years. Clozapine use was only predicted by positive symptom criterion. Prospective characterization of TRS is possible at 6 months after FES through a time-based approach using a 50% threshold for symptom change in treatment-adherent patients.