A Case–Control Study of Demographics and Clinical Predictors of Treatment-Resistant Schizophrenia in Patients 12 Years After First-Episode

Abstract

Background: This is a retrospective case–control study with an aim to explore the pattern and predictors of treatment-resistant schizophrenia with a large cohort of first-episode psychosis patients with more than 12 years of illness duration. The functioning, cognitive functioning, and clinical outcomes of treatment-resistant (TR) patients will be compared with that of non-TR patients. Methods: Clozapine prescription was considered as proxy indication of treatment resistance in the current study for screening purpose. A detail screening of clozapine prescription history of all 1400 patients, who were presented to mental health service from January 1, 1998, to August 31, 2003, was conducted. Treatment-resistant status of patients was further verified with clinical and medication information collected from the medical records based on the existing operational criteria and were considered as TR group for comparison. Patients in the control group were identified randomly from those who were not prescribed with clozapine, matched with the diagnosis of patients with clozapine in a 1:2 ratios. Operational criteria were used to identify those who were TR without prescribing clozapine within this group. The rest was considered the non-TR (Non TR) group. After data cleaning, the total sample used for analysis was 469, with 160 clozapine patients and 309 control patients. Face-to-face interview and detail medical record review were conducted for all patients. Results: Treatment-resistant schizophrenia (TRS) was found to be 11.79%. TRS patients were found to have poorer social and occupational functioning, cognitive functioning, and clinical outcomes. Binary logistic regression was performed and the model of age of onset, years of education, number of relapse in the first 3 years, duration of first episode, Clinical Global Impression (Severity) Scale positive symptoms (CGIp) at the end of first month, substance abuse history, and premorbid adjust scale (adult) significantly predicted treatment resistance status (Omnibus chi-square = 47.86, df = 7, P < .0001). The model accounted for between 15.9% and 21.4% of the variance in treatment resistance status. Conclusion: The results supported that both neurodevelopmental and early treatment outcomes might be related to the development of treatment resistance and thus may be used as indicators in the future. Also, the importance of early treatment outcomes including number of relapse and duration of first episode has suggested the specific importance of clinical care during the first-episode psychosis.