Reliability and validity of a tardive dyskinesia videotape rating technique.

Ninety-four psychiatric in-patients, receiving regular antipsychotic medication, were videotaped using a standard procedure. The tapes were rated by blind observers using a simple scoring system for the duration of abnormal movements. Using this combined videotape and rating scale assessment technique the re-rating reliability, inter-rater reliability and test-retest reliability were high. In order to demonstrate the validity of the technique the rating scale scores in a sub-sample of 30 patients, were compared with the assessment of three experienced clinicians on the same patients, and AIMs scores. Central (lip, tongue, jaw and neck movements) scores showed close agreement with the clinicians' assessment, suggesting that clinical diagnosis is based principally on the presence and severity of oro-facial dyskinesia. Total rating scale scores were in close accord with total AIMs scores. When the two scales were carried out on the same patients on the same occasion a diagnostic criterion level of 2 or more on the central score produced a tardive dyskinesia prevalence rate identical to that produced by an AIMS criterion level of 2 or more on the global severity rating. The tardive dyskinesia prevalence rate based on the central score criterion level showed an increase with age.