13Questionnaires — THI, TFI and Friends

FWhy measure with a questionnaire at all?

Tinnitus is a private percept; nobody else can hear it, and its physical correlates — matched loudness, pitch, maskability — correlate poorly with how much it bothers a person. Two patients with identical audiograms and identical matched loudness can be worlds apart: one untroubled, the other unable to work or sleep. The gap between the sound and the suffering is exactly what a validated questionnaire is built to capture.

Patient-reported outcome measures (PROMs) turn a subjective experience into a number that can be compared across visits, across clinicians and across studies [2014]. They serve three jobs: a baseline of severity, a structured profile of which life domains are affected, and a yardstick for change after treatment. A questionnaire score is not a diagnosis, but it is the single most useful number you will record in a tinnitus clinic.

TThe Tinnitus Handicap Inventory (THI)

The THI, published by Newman and colleagues in 1996, remains the most widely translated tinnitus instrument in the world [1996]. It has 25 items grouped into three subscales: functional (11 items, on role, social and physical functioning), emotional (9 items, on anger, frustration and depression) and catastrophic (5 items, on the sense of desperation and loss of control). Each item is answered “yes” (4 points), “sometimes” (2) or “no” (0), giving a total from 0 to 100.

Newman’s grade bands divide the total into no handicap (0–16), mild (18–36), moderate (38–56) and severe to catastrophic (58–100). The THI is quick, robust and easy to score by hand, which is why it endures. Its main weakness is sensitivity to change: it was designed to describe handicap, not to track small treatment effects, so a clinic that wants to demonstrate improvement may find the THI moves sluggishly.

TThe Tinnitus Functional Index (TFI)

The TFI was developed by Meikle and colleagues in 2012 specifically to be responsive — to detect the change in tinnitus severity that an effective treatment produces [2012]. It has 25 items spanning eight subscales: intrusiveness, sense of control, cognition, sleep, auditory difficulties, relaxation, quality of life and emotion. Items are rated 0–10 and the total is rescaled to 0–100.

Because the eight subscales each profile a distinct domain, the TFI tells you not just how bad but in what way a patient is affected — sleep-dominant, cognition-dominant, emotion-dominant — which can steer therapy. Subsequent validation and outcomes work confirmed its strong internal consistency and treatment responsiveness, and a minimal clinically important difference (MCID) of roughly 13 points has been proposed for judging whether change is real rather than noise [2016]. Independent psychometric studies have broadly endorsed the structure while debating the exact subscale model [2016].

CThe older cousins — TQ, THQ and the mini-TQ

Before the THI and TFI dominated, several instruments mapped the same territory. The Tinnitus Questionnaire (TQ) of Hallam and the Tinnitus Handicap Questionnaire (THQ) of Kuk are still encountered, especially in European and cognitive-behavioural settings, and the abbreviated mini-TQ offers a shorter screen. Each carries its own subscale model and norms, so scores are not interchangeable: a “moderate” on one is not a “moderate” on another.

The practical lesson is to pick one instrument and use it consistently for a given patient and clinic, so that follow-up scores are comparable. Mixing instruments between visits destroys the ability to read change. European consensus work has tried to standardise which PROMs are reported, precisely so that data can be pooled across centres [2019].

CUsing the numbers in clinic

A workable routine is to administer the chosen handicap questionnaire at the first visit (the baseline), repeat it at defined follow-up points, and interpret change against the instrument’s MCID rather than treating any drop as success. A 5-point fall on the TFI is within measurement noise; a 13-point fall is meaningful. Pair the questionnaire with brief screens for the comorbidities it cannot fully capture — anxiety, depression and insomnia — because these often drive the total score and respond to specific treatment [2014].

Finally, read the subscales, not just the total. Two patients with a TFI total of 60 may need different care if one is sleep-dominant and the other emotion-dominant. The questionnaire is most powerful when it is used as a profile and a tracker, not merely as a severity stamp.