Reliability and validity of the longitudinal interval follow-up evaluation for assessing outcome of anxiety disorders.