Utility of risk-status for predicting psychosis and related outcomes: evaluation of a 10-year cohort of presenters to a specialised early psychosis community mental health service

Psychosis transition rates by those at clinical high risk have been highly variable and few studies have compared service presenters across the full psychosis risk spectrum with respect to medium-term outcomes. A 10-year service cohort was examined (N=1997), comprising all presentations to an early psychosis service for young people experiencing a recent psychotic episode or at increased risk ('Psychological Assistance Service', Newcastle, Australia). Baseline and longitudinal service data (median follow-up =7.3 years) were used in a series of logistic regressions to examine relationships between psychosis risk-status and subsequent illness episodes, hospital admissions, and community contacts. Six baseline groups were identified: existing (14.5%) and recent psychosis (19.8%); ultra-high risk (UHR, 9.6%); non-psychotic disorders without (35.4%, the reference group) and with psychiatric admissions (8.3%); and incomplete assessments (12.5%). High comorbidity levels were reported by the cohort (psychosocial problems, 61.1%; depression, 54.1%; substance misuse, 40.7%). UHR clients experienced similar psychosis transition rates to the reference group (17.3% vs. 14.6%; 8.9% vs. 9.1% within 2-years) and comparable rates of subsequent non-psychosis outcomes. A 25.9% conversion rate from early psychosis to schizophrenia was detected. However, among transitioning individuals, UHR clients faired relatively better, particularly with respect to changes in comorbidity and mental health contacts. Interventions tailored to current problems, recovery and psychological strengthening may be more appropriate than those based on estimated psychosis risk, which currently lacks clinical utility.