Early Predictors of Ten-Year Course in First-Episode Psychosis.

OBJECTIVE:Identifying patients at risk of poor outcome at an early stage of illness can aid in treatment planning. This study sought to create a best-fit statistical model of known baseline and early-course risk factors to predict time in psychosis during a ten-year follow-up period after a first psychotic episode. METHODS:Between 1997 and 2000, 301 patients with DSM-IV nonorganic, nonaffective first-episode psychosis were recruited consecutively from catchment area-based sectors in Norway and Denmark. Specialized mental health personnel evaluated patients at baseline, three months, and one, two, five, and ten years (N=186 at ten years). Time in psychosis was defined as time with scores ≥4 on any of the Positive and Negative Syndrome Scale items P1, P3, P5, P6, and G9. Evaluations were retrospective, based on clinical interviews and all available clinical information. During the first two years, patients were also evaluated by their clinicians at least biweekly. Baseline and early-course predictors of long-term course were identified with linear mixed-model analyses. RESULTS:Four variables provided significant, additive predictions of longer time in psychosis during the ten-year follow-up: deterioration in premorbid social functioning, duration of untreated psychosis (DUP) of ≥26 weeks, core schizophrenia spectrum disorder, and no remission within three months. CONCLUSIONS:First-episode psychosis patients should be followed carefully after the start of treatment. If symptoms do not remit within three months with adequate treatment, there is a considerable risk of a poor long-term outcome, particularly for patients with a deterioration in premorbid social functioning, a DUP of at least half a year, and a diagnosis within the core schizophrenia spectrum.