Instruments for the assessment of social anxiety disorder

Abstract

Social anxiety disorder (SAD) is difficult to diagnose by clinicians, a fact that stimulates the study and development of instruments that will favor its early and systematic recognition. The objective of the present study was to identify in the literature indexed from 1999 to July 2004 articles related to psychometric studies on instruments for the assessment of SAD. A systematic search was performed in the PsychoInfo, Lilacs and Medline indexing services using the following key words: social phobia scale, social phobia validity, social phobia reliability and social anxiety scale, with 26 papers being identified. For analysis, the studies were divided into two groups: a) Eleven regarding the Liebowitz Social Anxiety Scale (LSAS) and b) Fifteen regarding other scales already validated or in the process of validation. Among the studies with the LSAS, six worked with the hetero-applied version and four with the self-applied version, identifying satisfactory psychometric qualities regarding validity in the original language (English) and in other languages such as French, Hebrew, Turkish and Spanish. In addition, good reliability indices were found, and regarding factorial analysis, the four-factor model was considered to be the most adequate. Among the studies using other scales, eight worked with instruments already validated and seven with new scales, identifying satisfactory values regarding validity and reliability. The main limitations pointed out in the studies surveyed were related to the diversity of the gold standard adopted; sample composition with the absence of clear inclusion and exclusion criteria, and the small number of studies on non-clinical samples, impairing the use of the instruments for general population screening. Future studies are clearly needed to improve the validation of the criterion based on factorial and cluster analysis.