M125. Predictors of Community Tenure in Patients With Treatment Resistant Schizophrenia Following Discharge From a Social Learning Inpatient Program

• Published in: Schizophrenia bulletin 2017-03, Vol.43 (suppl_1), p.S256-S256
• Main Authors: Palekar, Nikhil, Latoussakis, Vassilios, Farley, Elizabeth, Bloch, Andrew, Katz, Marcie, Parker, Sandra, Amar, Jennifer, Anthony, Donna, Gallego, Juan, Ahmed, Anthony
• Format: Article
• Language: English
• Subjects: Abstracts
• ISSN: 0586-7614; 1745-1701
• DOI: 10.1093/schbul/sbx022.119

Background: Resistance to antipsychotic treatment is a significant clinical problem in patients with schizophrenia. The Social Learning Inpatient Program at New York—Presbyterian, Westchester Division is a specialized inpatient program based on a social learning model which includes interventions such as social skills training, cognitive remediation, cognitive-behavioral therapy (CBT) for Psychosis and other token economy-based interventions. The goal of our study was to determine if treatment at our program was associated with lower re-hospitalization rates after discharge and to look for predictors of successful community tenure. Methods: 58 patients with treatment resistant schizophrenia (TRS) consented to participate in the study and were assessed at the time of discharge and then followed up with a phone call at 1 month, 3 months and 6 months to collect information regarding re-hospitalization. TRS was defined based on 2 failed adequate antipsychotic trials. Assessments at the time of discharge included the Positive and Negative Syndrome Scale (PANSS) and Montreal Cognitive Assessment (MOCA) which were administered by trained clinicians. PANSS ratings were summed into composite scales informed by the 5-factor model of the PANSS—these included Positive, Negative, Disorganization, Excitement/Agitation, and Emotional Distress subscales. We used a binary logistic regression (BLR) model to identify the fewest set of symptom variables that will predict successful community tenure amongst patients discharged from the social learning program. We used an ANOVA to compare the MOCA cognitive scores. Results: Fifty-six percent of the patients were men, 36% were Caucasian, 40% African American, 20% Hispanic and 4% were Asian. At 6 months following discharge, 23% of patients with TRS were re-hospitalized at least once and 77 % of patients maintained community tenure without re-hospitalization. The overall BLR model was significant (−2LogL = 19.54, χ 2 (1) = 6.86, P  = .009). The selected BLR model accounted for 37.9% variance in the data (Nagelkerke R 2  = 0.379) and correctly predicted the hospitalization status of 74% of patients discharged. The BLR identified the Disorganization subscale of the PANSS as the strongest predictor of re-hospitalization within 6 months. Patients who were re-hospitalized within 6 months also had lower memory subscale scores ( F =30.25, P  = .032) than non-re-hospitalized patients. Conclusion: Seventy-seven percent of patients discharg