Documenting an episode of psychiatric illness: need for multiple information sources, multiple raters, and narrative

It is argued that the structured psychiatric interview, which has become the foundation of much modern clinical research, is not a sufficient method of documenting an episode of psychotic illness. This valuable resource needs to be supplemented by a high standard of recorded observation throughout hospitalization by all staff in contact with the patient. The conversion of observations into ratings can involve considerable error and should be left to at least two highly trained research workers. The only data worthy of this effort are the entire body of clinical information obtained from all sources. This requires detailed written records. The advantages of a narrative record, in addition to numerical ratings, is that it can be reinterpreted at any time using different systems of psychopathological or nosological ideas.