Incidence and prevalence of non-affective psychoses in Stockholm County : a register-based study

Abstract

Objective: To estimate the incidence and prevalence of non-affective psychoses including the subgroup schizophrenia in Stockholm County, using health-care records, and to investigate the relationship between psychiatric out- and inpatient care for persons with non-affective psychosis.

Method: Data from the psychiatric out- and inpatient health-care in Stockholm County has been collected at local level and constitutes Stockholm Non-Affective Psychoses-registry (SNAPS). The register comprises 13,561 individuals registered with the diagnosis of non- affective psychosis at some time during the years 1997–2006. SNAPS include information about outpatient visits (1997–2006) and inpatient occasions (years 1973–2006) as well as information about psychiatric diagnoses. Incidence and prevalence based on both out- and inpatient data and incidence based solely on inpatient data were estimated for the year 2005. Furthermore, prevalence over time were estimated, and for the year 2005, different geographical areas in the county.

Results: When data from out- and inpatient care used, the incidence of schizophrenia was estimated to 28/100 000 inhabitants (age range 18–34 years) and non-affective psychosis to 72/100 000 inhabitants (age range 18–44 years). Considerably lower incidences were noted when only inpatient data were used (schizophrenia 13/100 000 and non-affective psychosis 42/100 000). Schizophrenia prevalence was estimated to 0.37% and for non-affective psychosis to 0.67% (age group 18–64). There were large geographical differences in the prevalence of non-affective psychosis in Stockholm County (from 0.35% to 1.32%). The proportion of persons who only had contact with outpatient care during the years 1997–2006, was estimated to 45%. When consideration to previous inpatient care was taken, the proportion reduced to twenty-five percent. For prevalent cases in 2005, the proportion of persons who only appeared in outpatient care was 67%.

Conclusion: Treatment incidence and prevalence of non-affective psychoses can be estimated with the health-care based registry SNAPS. Estimations of psychosis incidence and prevalence based on health-care records after the millennium, should include both out- and inpatient data, to avoid too much underestimation and misclassification of incident cases. Given the development in psychiatric care, with more out- instead of inpatient care, information about diagnoses from both out- and inpatient care will be needed in the future. This is important for health-care planning as well as for research.