Sensitivity to change and convergent validity of the Tinnitus Functional Index (TFI) and the Tinnitus Questionnaire (TQ): Clinical and research perspectives

Abstract

BackgroundComplete alleviation of tinnitus perception is rarely achieved by current therapeutic interventions. However, adequate therapy may induce a small or large decrement of the tinnitus. The assessment of the therapeutic effect is challenging due to large interindividual variability. Multiple tinnitus questionnaires are available to describe functional effects of tinnitus pre and post-therapeutic intervention, of which the Tinnitus Functional Index (TFI) and Tinnitus Questionnaire (TQ) are two very commonly used questionnaires by clinicians and researchers. ObjectiveTo recommend either TFI or TQ as an outcome measure for the investigation of tinnitus treatments by investigating the consistency between the TFI and TQ, as well as comparing the responsiveness of both scales after therapeutic intervention. MethodsData from the validated Dutch versions of the TFI and the TQ of 100 chronic subjective tinnitus patients who underwent 6 sessions of neuromodulation at the Antwerp University Hospital were included. The patients reported their perceived effect and filled out the TFI and TQ at the pre-therapy, post-therapy and follow-up visit (i.e. ± seven weeks post-therapy). The intra-class correlation (ICC) was determined, measuring consistency between the TFI and TQ. Furthermore, the responsiveness was compared by use of mixed effects analysis. Finally, the agreement between the total scores of the questionnaires and the perceived effect was evaluated by use of their receiver operating characteristic (ROC) curve. ResultsThe total scores of the TFI and TQ showed a good agreement at pre-therapy (ICC = .82) and post-therapy visit (ICC = 0.82). The ICC indicated a lower, but still good agreement at follow-up (ICC = 0.79). Furthermore, the total scores of the TFI and TQ significantly changed over time (pTFI < .05; pTQ < .05). In the patients who verbally reported at follow-up visit that their tinnitus complaints decreased after treatment, the percentage of patients showing a clinically significant decrease on the TFI (i.e. a decrease of ≥13 points) and the TQ (i.e. decrease of ≥12 points) was 50% and 33% respectively, however, not statically significant different (p = .13). Finally, the analysis of the agreement between the perceived effect and the clinically significant improvement on the questionnaires showed a kappa-value of 0.36 for the TFI and 0.29 for the TQ. ConclusionThe TFI and TQ have high convergent validity and are both suitable as an outcome measure to quantify a change in the self-perceived tinnitus burden. The TFI is shorter and shows a slightly higher agreement with the self-reported perceived effect. Therefore, this comparison is in favor of the TFI for research purposes.