7Evidence and Comparative Efficacy

FHow to read a tinnitus trial

Comparing tinnitus treatments is hard for reasons baked into the condition. Tinnitus is subjective, fluctuating and heterogeneous; outcomes rely on self-report questionnaires such as the Tinnitus Handicap Inventory (THI) and Tinnitus Functional Index (TFI); placebo and expectation effects are large; and natural improvement over time muddies any uncontrolled study. A treatment can look impressive in a before-and-after series and then show little advantage over a credible control in a randomised trial.

So the questions to ask of any claim are: was there a control group, was it credible (an active comparator, not just a waiting list), what outcome was measured, and was the change clinically meaningful rather than merely statistically significant? Holding the evidence to that standard reshapes the comparative picture considerably [2014].

TTRT versus masking

The longest-running comparison pits Tinnitus Retraining Therapy (directive counselling plus low-level sound aimed at habituation) against traditional masking (sound set loud enough to cover the tinnitus). A randomised clinical trial by Henry and colleagues found that both produced improvement, with TRT showing an advantage particularly in patients whose tinnitus was more severe at baseline [2006]. This helped establish TRT as more than a relabelled masking programme, while confirming that sound delivered in either philosophy can help.

The more rigorous test came from the multicentre Tinnitus Retraining Therapy Trial (TRTT). When full TRT (counselling plus sound generators) was compared against standard of care and against partial-TRT arms, the picture was sobering: the additional benefit attributable specifically to the sound-generator component, over and above counselling and standard care, was small and not clearly clinically significant [2019]. The counselling and the natural course of the condition accounted for much of the improvement.

TThe strongest signal: CBT

Against this backdrop, the psychological evidence stands out. A Cochrane systematic review concluded that CBT produces a reliable reduction in tinnitus-related distress and improves quality of life and associated depression, while — importantly — not changing the measured loudness of the tinnitus itself [2020]. Earlier meta-analysis of randomised trials reached the same broad conclusion, identifying CBT as the psychological intervention with the most consistent effect on tinnitus-related quality of life [2011].

This is the cleanest result in the whole field: a modality with a clear, replicated effect on the outcome that matters most to patients — how much the tinnitus interferes with life — even though it does nothing to the percept. It is why guidelines give CBT a recommendation that sound-only approaches do not earn [2014].

CThe modest direct effect of sound

Sound therapy is widely used and intuitively appealing, but the direct evidence for it as a stand-alone treatment is modest. A Cochrane review of sound therapy using amplification devices and/or sound generators found the trials small, heterogeneous and at risk of bias, with no convincing evidence that sound therapy is superior to other interventions and only low-certainty evidence overall [2018]. A scoping review of combined amplification-plus-sound-generation devices similarly found supportive but low-quality evidence, unable to separate the contribution of sound from that of amplification and counselling [2018].

This does not mean sound is useless — many patients value it, particularly for sleep and quiet environments, and it underpins habituation-based programmes. It means the honest claim is “helpful adjunct with low-certainty direct evidence,” not “proven primary cure.”

CThe honest conclusion: combine and individualise

Put the strands together and a consistent story emerges. No single modality dominates: TRT and masking both help but neither is a clear winner, the specific sound-generator effect is small, sound therapy has only low-certainty direct evidence, and CBT has the strongest effect — but only on distress, not the percept. The logical synthesis is that the components are complementary, and that the best outcomes come from combining them and matching the mix to the individual rather than crowning one approach [2019].

This is exactly why guidelines recommend stepped, multidisciplinary care: education and reassurance for all, hearing rehabilitation where there is hearing loss, sound enrichment for those who find it helpful, and CBT-based psychological therapy for distress, escalated and combined according to severity [2014]. The evidence does not tell us to pick a favourite; it tells us to assemble a programme.