Outcome after first episode psychosis is heterogeneous, but knowledge about the distribution and predictive factors is limited.

To investigate medication and rehospitalizations for five years after first episode hospital treated psychosis.

Swedish population registers were used to select a nation-wide cohort of 962 cases (589 or 61% men) with a first hospitalization for psychosis at ages between 16–25 years. Cases were categorized year by year for 5 years after the initial hospitalization with regard to rehospitalizations and dispensations of antipsychotics and other medications.

The 5-years mortality was 4% (n = 39) with suicides in 16 cases (1.6%, 11 of which were men). Additionally, 139 cases (23% of women and 10% of men) had hospitalizations for suicide attempts within 5 years. A bimodal distribution of years with medication was found indicating two different trajectories of outcome. One peak was seen for cases with dispensations of antipsychotics 5 of 5 years (40% of the cohort). Another peak was seen at dispensations during at most 1 of 5 year (30%). During year 5, 514 (56% of 923 cases surviving 5 years) had dispensations of neuroleptics and 257 (28%) were hospitalized, whereas 356 cases (39%) had no dispensation of neuroleptics or hospitalization.

The population of young cases with first episode psychosis is heterogeneous with at least two clearly separable trajectories based on medication and hospitalizations. The high mortality and high incidence of suicide attempts during a five-year period demonstrate a need for careful monitoring of these patients.

The authors have not supplied their declaration of competing interest.