Reliability of Andreasen's thought, language and communications disorder scale.

The study of disordered thinking has been hampered by a lack of clinical instruments aimed at the measurement of thought disorder. Andreasen has proposed a rating scale for thought disorder (TLC) and reported on its reliability and preliminary aspects of its validity. In our study, a psychiatrist and physician assistant rated 98 psychiatric inpatients using the TLC and BPRS. Interrater reliability ranged from 0.35 to 0.80 (weighted kappa) on 18 aspects of thought disorder and one global item. The order of reliability for the 18 individual items was strongly correlated between the Andreasen sample and our own (r = 0.77). While the BPRS thought disorder scale was significantly related to the global rating of thought disorder on the TLC (r = 0.71), only half of the variance in thought disorder measured by the TLC was accounted for by the BPRS factor. Thus, this study supports Andreasen's contention that the TLC is a reliable instrument and suggests that the instrument may provide additional information about thinking disorders.