An essential condition to validate a diagnosis is its stability over time. Since there are no biological markers for psychiatric disorders, the diagnosis relies on clinical expertise, with several consequences in treatment planning, disease burden and disability, affecting outcome and public health.

The aims of this study were:

1. 1. the assessment of long term stability of the diagnosis of psychotic disorders,

2. 2. its implications in patient treatment, and

3. 3. the evaluation of eventual predictors of diagnosis stability.

This was a retrospective study carried out in the Department of Psychiatry of a large University Hospital (Hospital S. João, Porto, Portugal). Patients enrolled were admitted in the inpatient unit from 2000 to 2003 (n=190, 12.41% of 1531 patients admitted), experienced a first psychotic episode, and fulfilled criteria for one of the following diagnosis: schizophrenia, schizoaffective disorder, bipolar disorder, drug induced psychosis, acute and transient psychotic disorders, schizophreniform disorder and psychosis NOS (ICD-10 classification). the diagnoses were extracted from clinical records, and reassessed five years after the initial diagnosis. the analysis focuses on diagnostic agreement over time; clinical and demographic variables were also collected and putative associations with diagnostic shift considered.

The study is now under statistic evaluation.