Psychosis and bipolar disorder risk in child and adolescent mental health services in the UK: population cohort study.

BACKGROUND Current approaches to identifying individuals at risk for psychosis capture only a small proportion of future psychotic disorders. Recent Finnish research suggests a substantial proportion of individuals at risk of psychosis attend child and adolescent mental health services (CAMHS) earlier in life, creating important opportunities for prediction and prevention. To what extent this is true outside Finland is unknown. AIMS To establish the proportion of psychotic and bipolar disorder diagnoses that occurred in individuals who had attended CAMHS in Wales, UK, and whether, within CAMHS, certain factors were associated with increased psychosis risk. METHOD We examined healthcare contacts for individuals born between 1991 and 1998 (N = 348 226), followed to age 25-32. Using linked administrative healthcare records, we identified all psychotic and bipolar disorder diagnoses in the population, then determined the proportion of cases where the individual had attended CAMHS. Regression analyses examined associations between sociodemographic and clinical risk markers with psychotic and bipolar disorder outcomes. RESULTS Among individuals diagnosed with a psychotic or bipolar disorder, 44.78% had attended CAMHS (hazard ratio = 6.28, 95% CI = 5.92-6.65). Low birth weight (odds ratio = 1.33, 95% CI = 1.15-1.53), out-of-home care experience (odds ratio = 2.05, 95% CI = 1.77-2.38), in-patient CAMHS admission (odds ratio = 1.49, 95% CI = 1.29-1.72) and attending CAMHS in childhood (in addition to adolescence; odds ratio = 1.16, 95% CI = 1.02-1.30) were all within-CAMHS risk markers for psychotic and bipolar disorders. CONCLUSIONS A substantial proportion (45%) of future psychotic and bipolar disorder cases emerge in individuals who had attended CAMHS, demonstrating large-scale opportunities for early intervention and prevention within CAMHS.