Diagnostic specificity of the child psychosis-risk screening system with a focus on the differentiation of schizophrenia spectrum disorders and neurodevelopmental disorders

Research on early psychosis has begun to identify psychiatric characteristics of the prodromal period of schizophrenia; however, subclinical characteristics of children in non-psychiatric fields have not been fully investigated. In our previous study, we developed the Child Psychosis-risk Screening System (CPSS).In the present cross-sectional study, we attempted to identify the risk of developing psychosis in pediatric (n = 216) and psychiatric outpatients (n = 120), aged 6– 18 years, with the CPSS.An analysis of variance of CPSS risk was performed in six diagnostic categories to examine specificity for each diagnosis. Receiver operating characteristic (ROC) curve analysis was conducted using the onset of schizophrenia spectrum as the outcome, and the discriminatory power and cut off values of the CPSS were determined. Logistic regression analysis was performed using clinical data to identify factors associated with the risk group (those at high risk of developing psychosis in the future) identified using the CPSS.There were significant differences in risk variance among diagnostic categories (p < 0.001), especially between schizophrenia spectrum disorders (SSD) and neurodevelopmental disorders (p = 0.001). CPSS had sufficient discriminatory power for SSD diagnosis [area under the ROC curve = 0.853 (95% confidence interval: 0.774–0.931)]. The cut off value for the risk of SSD was determined to be 98.1%, achieving the best mean of the sum of sensitivity (90.9%) and specificity (84.0%). Cross-sectional logistic regression analysis showed that along with “SSD diagnosis,” “winter birth,” and “maltreatment” were factors associated with the risk group (odds ratio = 38.05 [p = 0.001], 2.30 [p = 0.016], and 0.12 [p = 0.024], respectively).CPSS may have potential use in the early detection of psychosis and differentiation from neurodevelopmental disorders, but this study was small and further studies with larger sample sizes and longitudinal study designs are required prior to its use in routine clinical practice.