Objective: There is a lack of consensus on the identification of seriously mentally ill patients (SMI). This study investigates the external and predictive validity of an operationalized definition for the severity and persistency of mental illness applied to a sample of service users attending a community mental health service.
Method: The definition is based on the fulfilment of dysfunction (GAF < or = 50) and illness duration (> or = 2 yrs) criteria. The study was conducted with a two-year longitudinal design. External and predictive validity of the SMI definition were assessed against the diagnosis of psychosis.
Results: Our data show evidence for an overall high predictive and external validity of the SMI definition and high sensitivity in predicting those with high burden of mental illness.
Conclusions: In order to identify people with high levels of psychiatric burden, the SMI working definition seems to be more useful than that simply based on diagnostic criteria.