BackgroundMany individuals with chronic or treatment resistant schizophrenia experience multiple relapses or treatment failures during the course of the illness. Some of these relapses are due to poor treatment compliance that is inherent to schizophrenia. Poor insight may be a leading cause for partial- or non-adherence to treatment as a high proportion of individuals with schizophrenia are partially or completely unaware of their mental disorder. The first two years after stabilization are thought to be key for long-term functional and clinical prognosis. In the present study, we first estimated the rate of treatment failure or relapse following clinical stability at one and two years, investigate the time to occurrence of relapse among individuals with chronic schizophrenia and the association between baseline clinical features recorded during the inpatient hospitalization and relapse at one- and two-year time points. We hypothesized that lack of insight, as measured by the PANSS item G12, would be a greater predictor of relapse than other characteristics previously suggested as possible predictors, including other symptoms from the PANSS, age, duration of illness, age at onset of illness, substance use, number of prior hospitalizations, and length of stay of the pre-discharge hospitalization.MethodsA total of 138 participants diagnosed with schizophrenia or schizo-affective disorder were assessed with a comprehensive assessment at one year and two-year following discharge from a long-term psychiatric facility. Regression models were used to determine factors predicting time to relapse and other elements of functioning. Baseline factors examined included PANSS, MCCB, PSP, demographics and treatment variables.ResultsRelapse rates were 56.52% (n=78 of 138) by Year 1, and 69.56% (n=96 of 138) by end of Year 2. The estimated relapse-free period for all individuals at the end of the study was 8.78 months. The backward elimination (–2 log likelihood=189.59, χ2=9.01, df=2, p=.021) showed that the best predictive variables for relapse were lack of insight/judgment as assessed by PANSS item G12 at baseline (B=.20, SE=.09, df=1, p=.011, Exp[B]=1.36), lifetime years of substance use (B=.16, SE=.11, df=1, p=.029, Exp[B]=1.33), PANSS Factor baseline score on Disorganization (B=.15, SE=.12, df=1, p=.031, Exp[B]=1.56), and number of previous hospitalizations (B=.13, SE=.11, df=1, p=.048, Exp[B]=1.23). No other baseline variables were found to be significant.DiscussionPoor insight is a fundamental symptom of schizophrenia that, while not entirely and uniformly expressed in all individuals, is among the most common symptoms across subjects. Our study shows that the numerous negative consequences of lack of insight should lead clinicians and researchers to make insight a high priority for allocation of clinical resources. Effective approaches to managing these predictive characteristics will allow affected individuals and their families and care providers to take part in collaborative treatment and relapse risk-management paradigms.