15Patient Selection and the Evidence — A Cautious Algorithm

FThe central principle: no target, no procedure

Across this chapter one rule recurs. Surgery and interventional procedures help tinnitus when, and essentially only when, there is an identifiable structural lesion that the procedure can correct — a pulsatile venous-sinus abnormality, a dural fistula, a glomus tumour, a dehiscent canal, a deafferented cochlea suitable for implantation, or a discrete neurovascular conflict [2014].

The vast majority of tinnitus is subjective, non-pulsatile and centrally maintained, with no surgically correctable substrate. For those patients the evidence-based path is the conservative one — education, sound therapy, hearing-aid amplification and cognitive behavioural therapy — not the operating theatre [2013].

TThe evidence base is mostly weak

It is important to be honest about the quality of evidence underpinning interventional tinnitus care. With the partial exception of cochlear implantation for single-sided deafness, most procedures rest on case series and small uncontrolled cohorts. Microvascular decompression, invasive neuromodulation and several venous procedures have no randomised controlled data [2017].

The major clinical practice guideline reflects this restraint: it recommends against routine imaging and against many medical and surgical interventions for non-pulsatile subjective tinnitus precisely because benefit is unproven and harm is possible [2014]. Reading interventional outcome papers requires attention to selection bias, the absence of controls, and the powerful placebo effect that any dramatic procedure exerts on a subjective symptom.

CThe danger of operating on subjective tinnitus

The harm of an inappropriate procedure is not theoretical. Operating on subjective tinnitus without a target can leave the tinnitus unchanged or worse, and adds surgical morbidity — hearing loss, CSF leak, infection, cranial-nerve injury — to a previously non-life-threatening complaint [2014].

There is also psychological hazard. A patient with high distress and unrealistic expectations who undergoes surgery and is not helped may deteriorate, having invested hope in a definitive cure that the procedure could never deliver. The ethical anchor is ‘primum non nocere’: for a benign symptom, the burden of proof sits firmly on the intervention [2013].

TA cautious selection algorithm

A defensible pathway runs in steps. First, characterise the tinnitus: is it pulsatile or non-pulsatile? Pulse-synchronous tinnitus mandates a vascular work-up because a treatable lesion may be present. Second, look for a structural target with appropriate imaging only when the history points to one [2014].

Third, match the target to its established procedure — venous-sinus or fistula intervention for pulsatile vascular causes, canal repair for dehiscence, cochlear implantation for the deafferented or single-sided-deaf ear with disabling tinnitus, MVD only for the carbamazepine-responsive typewriter phenotype. Fourth, where no target exists, return the patient to evidence-based conservative care and reserve invasive neuromodulation strictly for research protocols [2013]. Throughout, multidisciplinary review and validated severity measures (THI, TFI) guard against over-treatment [2014].

CCounselling and shared decision-making

Even when a target exists, counselling must be candid. Cochlear implantation for single-sided deafness, the best-supported indication, still leaves a minority without tinnitus relief, and a few may worsen. Venous and fistula procedures carry their own risks. The patient should understand the probability of benefit, the chance of no change, and the real possibility of harm before consenting [2013].

The clinician’s job is therefore as much about declining to operate as about operating. For most people with tinnitus, the most skilled intervention is to explain why a procedure is not indicated, and to redirect energy toward therapies that reduce distress without surgical risk [2014].