F Gerhards1, D Brehmer. 1. Forschungszentrum für Psychobiologie und Psychosomatik, Universität Trier, Trier, Deutschland. gerhards@uni-trier.de
Abstract
BACKGROUND: There are different abridged versions of the tinnitus questionnaire (TQ) and reference values for grading tinnitus-related distress, the diagnostic value of which is unclear, as is the relation of sensitivity to specificity, as well as positive and negative predictive values. The present study is aimed at clarifying these issues and at benchmarking the existing abridged TQ versions. PATIENTS AND METHODS: TQ data from 209 outpatients who had tinnitus for a period of at least 4 months were collected at the beginning of treatment. The rate of correct severity ratings, the quality of making a distinction between (extremely) severe and compensated/moderate tinnitus (area-under curve under the receiver-operating-characteristics curve, and rate of correct classifications), as well as positive and negative predictive values were calculated. RESULTS: Applying unspecific reference values (independent of inpatient/outpatient status), the most favourable evaluation was achieved by the shortest TQ version (10 items). With the 12-item version (mini-TQ), better classification was achieved when reference values specific for outpatients were applied instead of unspecific values.
BACKGROUND: There are different abridged versions of the tinnitus questionnaire (TQ) and reference values for grading tinnitus-related distress, the diagnostic value of which is unclear, as is the relation of sensitivity to specificity, as well as positive and negative predictive values. The present study is aimed at clarifying these issues and at benchmarking the existing abridged TQ versions. PATIENTS AND METHODS: TQ data from 209 outpatients who had tinnitus for a period of at least 4 months were collected at the beginning of treatment. The rate of correct severity ratings, the quality of making a distinction between (extremely) severe and compensated/moderate tinnitus (area-under curve under the receiver-operating-characteristics curve, and rate of correct classifications), as well as positive and negative predictive values were calculated. RESULTS: Applying unspecific reference values (independent of inpatient/outpatient status), the most favourable evaluation was achieved by the shortest TQ version (10 items). With the 12-item version (mini-TQ), better classification was achieved when reference values specific for outpatients were applied instead of unspecific values.
Authors: Anna Chi Shan Kam; Arron Pak Pei Cheung; Patrick Yiu Bong Chan; Eric Kwok Shun Leung; Terence Ka Cheong Wong; Michael Chi Fai Tong; Andrew van Hasselt Journal: Int J Audiol Date: 2009-08 Impact factor: 2.117