[Study of P 300 in schizophrenia. Some reflexions on methodological difficulties]

We underscore the methodological difficulties in studying cognitive evoked potentials in schizophrenic patients. The main difficulties of such a study include the interruption of medication, the need of cooperation and immobility of subjects during the recordings, as well as of a correct patient-control matching for motivation, intellectual quotient, social and cultural levels. To establish correlations between clinical symptomatology and P 300 amplitude and latency we used both the DSM III-R and clinical rating scales. Clinical and electrophysiological testing must be performed the same day and have an excellent interrater agreement. Otherwise correlations are not reliable. Use of small patient samples is one major drawback which prevents to draw definitive conclusions. These difficulties usually unsaid are inherent to research in biological psychiatry and may partially explain the heterogeneity of results in the literature.