BACKGROUND Several studies have suggested a declining first-admission rate for schizophrenia. This study examines the care-based incidence of schizophrenia in a Dutch register area.
METHOD Data from Groningen psychiatric case register were used to compare first-admission rates for schizophrenia over 1976-90 with those of other functional psychoses, and to consider various potential biases.
RESULTS Diagnostic habits probably affected time trends in incidence rates. Using a broader definition of schizophrenia, no evidence was found for a decrease in the incidence of schizophrenia. Although first admissions to intramural services showed a (non-significant) decrease, this effect seemed to be neutralised when all mental health services were taken into account.
CONCLUSIONS Care-based studies of time trends in psychiatric disorders should embrace all mental health services. Furthermore, unless a diagnostic classification system with univocal criteria is used, bias caused by changing diagnostic habits cannot be ruled out.