The Fear of Tinnitus Questionnaire: Toward a Reliable and Valid Means of Assessing Fear in Adults with Tinnitus.

OBJECTIVES

The purpose of this study was to assess the factor structure, reliability, and validity of the Fear of Tinnitus Questionnaire (FTQ); a brief self-report measure of people's fears about potential cognitive, emotional, behavioral, and social consequences of living with tinnitus.

DESIGN

Five hundred eighty-eight Dutch-speaking adults with tinnitus completed an online battery of questionnaires measuring tinnitus-related distress, fear, catastrophizing, and quality of life. The sample was randomly split into two to perform exploratory and Bayesian confirmatory factor analyses. A subsample of participants (n = 144) completed the battery of questionnaires a second time after a 2-week interval to calculate test-retest reliability and conduct a Bland-Altman analysis. Convergent and concurrent validity of the FTQ was assessed with the complete data set and measures of tinnitus-related distress as the outcome.

RESULTS

Exploratory factor analyses indicated that single- and three-factor FTQ models were both valid solutions. Posterior predictive p values for the Bayesian confirmatory factor analyses ranged between 0.51 and 0.53 indicating that the respective models were an excellent fit for the data. The FTQ showed excellent test-retest reliability (average value intraclass correlation coefficient (ICC) = 0.92; 95% confidence interval, 0.89-0.95) and in the Bland-Altman analysis, satisfactory agreement between participants' scores after a 2-week interval. Furthermore, the FTQ demonstrated good internal reliability (α = 0.83, 95% confidence interval, 0.81-0.85) and added statistically significant amounts of variance to models predicting tinnitus-related distress and interference in daily life.

CONCLUSIONS

The FTQ has good psychometric properties and can be used to assess people's fear of tinnitus in research or clinical settings. Further work to establish the reliability and validity should be conducted and include an examination of a version of the FTQ that uses Likert-type response scales which might offer improved sensitivity.