Brain imaging in psychiatry - a study of 435 psychiatric in-patients at a university clinic

Objective: The use of brain imaging in psychiatry still lacks clear guidelines. We investigated the referral practice, outcome and predictive factors of neuroimaging in a Swiss psychiatric university clinic. Method: Medical files were reviewed retrospectively for 435 consecutively hospitalized patients who were subjected to neuroimaging. The association between the sociodemographic and clinical characteristics and the scan results was analyzed using bivariate and multivariate analyses. Results: Of overall examinations, 69.4% were normal, 16.3% equivocal and 14.3% abnormal; 2.9% of scans ordered for screening only showed pathology. Neurologic signs and advanced age of patients predicted abnormal scan findings, whereas other variables such as EEG results showed no significant association. Conclusion: Our results support the need for clear indications for using brain imaging in psychiatric in-patients. Focal neurologic signs and advanced patient age seems to predict abnormal scan results. However, these criteria are not sufficiently sensitive to predict significant scan findings in all patients.