The Tinnitus Handicap Inventory is a better indicator of the overall status of patients with tinnitus than the Numerical Rating Scale

The Tinnitus Handicap Inventory (THI) and Numerical Rating Scale (NRS) for awareness, annoyance, loudness, and effect on life are two of the most commonly used questionnaires for patients with tinnitus. This study aimed to determine whether these two questionnaires can comprehensively evaluate the patient's condition and which questionnaire is better as a primary endpoint for tinnitus. Data from 90 patients who visited a university hospital with subjective, non-pulsatile tinnitus and without a history of any psychiatric disorders were reviewed between March 2020 and May 2022. The patients' medical histories, audiological profiles, questionnaires including the THI and NRS, Beck Depression Inventory (BDI), Beck Anxiety Depression (BAI), Hospital Anxiety-Depression Scale (HAD), and the Brief Encounter Psychosocial Instrument (BEPSI) were analyzed. The THI had a moderately positive correlation with the NRS for effect on life, annoyance, and loudness. It also had a low positive correlation with the BDI, HAD-A, BEPSI, and BAI. Considering annoyance and loudness, the NRS did not correlate with the other questionnaires, except for the THI (p > 0.05). The NRS for tinnitus awareness correlated with most audiological profiles. The ROC curve analysis revealed that the THI significantly predicted depression, anxiety, and stress. Conversely, none of the four NRS items significantly predicted these psychiatric problems (p > 0.05). Using the THI as a primary endpoint after interventions rather than the NRS is more reasonable because the THI covers the emotional, functional, and catastrophic aspects of tinnitus, although not the audiological status. To compensate for this, the auxiliary use of NRS for awareness should also be considered. •The Tinnitus Handicap Inventory (THI) correlates with most psychological questionnaires as well as Numerical Rating Scale (NRS) for tinnitus.•The NRS for tinnitus does not correlate with psychological questionnaires.