To identify which patient factors predict psychiatric hospitalization in patients presenting to the emergency department and to examine the role of the dimensional approach to psychopathology in predicting hospitalization.

We consecutively recruited 312 patients undergoing psychiatric evaluation in the emergency room of a hospital in Rome over a 6-month period. Patients were rated on the SVARAD (Scala per la Valutazione Rapida Dimensionale), a scale designed for the rapid assessment of the main psychopathological dimensions. Information about patient history, as well as sociodemographic and clinical variables, was also collected. Univariate analysis was performed to detect the variables associated with recommendation of psychiatric hospitalization. Multiple logistic regression analysis was used to identify independent predictors of hospitalization and compare their strength. A replication study was performed in another hospital on a random sample of 118 patients.

In both studies, patients who were recommended for psychiatric hospitalization showed significantly higher levels of anger/aggressiveness, apathy, impulsivity, reality distortion, thought disorganization and activation. Multivariate analysis identified psychopathological dimensions (reality distortion, impulsivity, apathy), diagnosis of psychotic or mood disorders and proposal for compulsory admission as independent predictors of psychiatric hospitalization. Hierarchical regression analysis revealed that the dimensional evaluation was the strongest predictor.

Our findings suggest that, in emergency setting, a systematic dimensional assessment may usefully complement the categorical assessment. Future research should aim at developing an operational assessment model, including both categorical and dimensional approaches to psychopathology.