The Camden Schizophrenia Surveys. III: Five-year outcome of a sample of individuals from a prevalence survey and the importance of social relationships.

Background: Most studies of outcome in schizophrenia have focused on incidence cohorts or samples identified through specialist mental health services; population-based samples provide a more complete picture of the effectiveness of community services.

Aims: To examine whether outcome predictors, derived from studies of selected patients with prolonged schizophrenia, would emerge in a largely community-dwelling population sample.

Methods: A follow-up sample of 114 adults with schizophrenia was identified via two censuses of key informants conducted for two prevalence surveys in North London, five years apart. Symptomatic, clinical and functional outcomes were assessed after five years. A composite score was derived for each individual. Multiple Linear Regression analyses were conducted in two phases to derive a best subset of predictors for global outcome.

Results: After five years, 33% were worse and 62% were better overall. The four best predictors (social isolation, living apart from relatives, longer illness and being an inpatient at first census) accounted for 32% of the variance in outcome of those with schizophrenia and related diagnoses.

Conclusions: Social relationships during the course of illness are an important predictor of overall outcome and relationships with friends and family each seem to make a positive contribution. Policy and service developments should focus on improving participation in community life for people with schizophrenia, particularly their social connectedness.