Materials and Methods
This study was a prospective, non-interventional, observational trial in patients experiencing tinnitus. The study was approved by the ethics committee of (Name of the hospital blinded for review) Hospital and all participants gave written informed consent. The study was performed in accordance with the ethical principles laid down in the Declaration of Helsinki of 1975, as revised in 2008. We followed the recommendations of the World Health Organization for the translation and adaptation of instruments and those of Guillemin et al. for the cross-cultural adaptation of health-related quality of life measures.[11,12] Moreover, we followed the STROBE statement for cross-sectional studies. [13]
The Spanish version of the TFI (Sp-TFI) was obtained in the following way. Stage I: the original TFI was translated by two independent otolaryngologists - head & neck surgeons who were both bilingual in English and Spanish. Stage II: The Spanish versions obtained were combined in a single document and sent to a panel of six experts in otology at our Otorhinolaryngology Department for comments. It was agreed that there were only very minor (non-meaningful) differences between the two translations and a general consensus was reached on a single version. Stage III: This agreed Spanish version was sent to a third otorhinolaryngology - head & neck surgeon bilingual in English and Spanish for back-translation to English and no meaningful differences were found with respect to the original version. Stage IV: the translated questionnaire was piloted in a group of patients to test its comprehensibility, ensuring that the patients understood the items as intended. Stage V: The panel of experts reviewed the patients’ responses, and as there was a consensus that the patients had understood the questions well, agreed that the questionnaire piloted could be definitive version (Supplementary figure).
The TFI was developed to provide a rating of tinnitus severity and the impact of its severity, and also a responsive measure of change in severity. It is composed of 25 items divided into eight subscales: intrusiveness (I), sense of control (SC), cognition (C), sleep (SL), auditory (A), relaxation (R), quality of life (Q), and emotional distress (E).[5]
The THI is a 25-item self-reported questionnaire that has been widely used in tinnitus clinical practice and research to assess distress and disability associated with tinnitus. In this instrument, the 25 items are grouped into three subscales: functional (disability affecting daily life); emotional (anxiety and mental stress); and catastrophic (hopelessness, loss of control, and inability to cope with the problem). Three response options are provided for each item: yes, sometimes or no, awarded scores of 4, 2, and 0, respectively. Consequently, the THI total score ranges from zero to 100, higher total scores indicating more severe handicap related to tinnitus. The validity and reliability of the Spanish version have been confirmed.[14]
Additionally, the loudness and annoyance related to tinnitus can be evaluated using VAS.[6,15]
To assess the reliability of the Sp-TFI, patients with tinnitus symptoms who came to an otology consultation, between May 2019 and March 2020, were asked to complete this new questionnaire, and also the THI and two VAS. [6,14] Specifically, using the VAS, patients were asked to rate their self-perceived tinnitus loudness and their annoyance on an 11-point Likert scale from 0-10. Lastly, they were asked to indicate which ear was affected by the tinnitus. The same patients were asked to complete the same questionnaires at 3 weeks and again at 6 months after the first consultation.
At baseline, patients were screened for cognitive impairment through an interview and by reviewing their medical records. Data on clinical and demographic characteristics were also obtained from the medical records of each participant. To explore convergent validity, we assessed correlations between scores on Sp-TFI, THI and VAS.
We prospectively included all patients who were Spanish native speakers, older than 18 years old, and had at least a 1-month history of tinnitus. Exclusion criteria were previous otology surgery on the same side as the tinnitus and/or a diagnosis of malignancy of the ear, as well as refusal to participate or insufficient cognitive skills to complete the questionnaires.
Statistical analysis was conducted using R 4.1.2 statistical language software (R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL (https://www.R-project.org/) with standard CRAN libraries. The internal consistency of the Sp-TFI was assessed with standard Cronbach’s α, considering the THI score the gold standard, and the temporal stability of the test with the intraclass correlation coefficient (ICC) type 2,1 in the nomenclature of Shrout and Fleiss. [16, 17] Specifically, ICC (2,1) was calculated in R using the psych package version 2.1.9 (Revelle, W. (2021) psych: Procedures for Personality and Psychological Research, Northwestern University, Evanston, Illinois, USA. https://CRAN.R-project.org/package=psych). [18]
A linear model was used to identify statistically significant influences and interactions between the variables studied, considering THI score the predicted (dependent) variable. The model equation was developed using a stepwise backward/forward (independent) variable selection method based on the Bayesian information criterion.[19]Statistical significance of the interactions detected in the model was evaluated using analysis of variance (ANOVA) Wald Type II tests.
A p value under 0.05 was considered statistically significant for this study. For both Cronbach’s α and ICC (2,1), values of 0.8 or above were considered target values to validate temporal stability (test-retest) and internal consistency of the Sp-TFI. [20, 21]