Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The tuning-fork is firmly established in the neurologist’s diagnostic armamentarium, having taken its pride of place for the investigation of hearing long before that of vibration sense. And is it any surprise? Sleek burnished steel, brandished by the neurologist at the faintest mention of hearing loss, is struck exuberantly against the elbow with a satisfying ping, echoing the confidence of the examiner that the ensuing tests will definitively distinguish a conductive from a sensorineural loss. We would argue, however, that the best use of cranial nerve VIII is the examiner’s own eighth nerve to take a history of the patient’s hearing loss, rather than the use of an antiquated technique of limited value.
The tuning-fork’s origins can be traced to the court of King James II. John Shore, the Sergeant Trumpeter, is credited with its invention in 1711 for tuning the lute; he apparently quipped before each of his performances, “I never go anywhere without my pitchfork” (clearly an 18th century Beach Boys predecessor (figure 1).1 Potential medical applications were conceived a century later. Ernst Heinrich Weber observed in 1825 that individuals heard a tuning-fork placed on the vertex better in the worse-hearing ear (ie, that with a conductive loss). Thirty years later, Heinrich Adolf Rinne developed experiments comparing the loudness of perceived bone conduction with that of air conduction on the problematic side. The terms ‘Rinne positive’ (air conduction is louder than bone conduction) and ‘Rinne negative’ (bone conduction is louder than air conduction) have been conventionally used to describe the tests’ results and represent a constant source of confusion for medical students for whom tuning-fork tests are still heavily emphasised.2
From the perspective of a neurologist, both tests have poor diagnostic accuracy and are often incorrectly used to screen for any type of hearing loss, despite having been envisaged only for conductive losses (Rinne) or low-frequency unilateral losses (Weber). Weber’s test can only be reliably interpreted when there is a unilateral hearing loss, but even in this situation an unacceptably high false-negative rate (30% refer to the midline3) limits its usefulness. Of the 70% who do lateralise, about a quarter does so to the ‘wrong’ ear.4 A recent study found a non-lateralising or incorrectly lateralising Weber’s test in over a fifth of 250 patients tested, and the test result did not reliably predict the audiometry results for the whole cohort.5 Rinne’s test is poorly sensitive at low air–bone gaps (50% at 10–20 dB), and the reliability depends markedly on the user’s experience and whether masking was used.6 A normal test result is not useful in ruling out hearing impairment in everyday clinical practice: a systematic review of five studies found low likelihood ratios ranging from 0.01 to 0.85.7 Another often-overlooked limitation is the propensity for falsely abnormal results. These occur when the bone conduction test on the poorer hearing side is heard by the contralateral ear; that is to say, a sensorineural loss in the test ear could lead to an incorrect diagnosis of a conductive loss on that side. Therefore, the clinician needs to mask the non-test ear (eg, by tragal rubbing) in cases where there is gross asymmetry of hearing, not something typically done by the investigating neurologist. The test is highly specific if performed correctly,8 but this is only really of use in confirming a conductive loss, something that is often apparent by looking in the ears and not of particular value to the neurologist who would benefit more from a highly sensitive diagnostic assessment.
If a patient reports hearing loss at a neurology clinic, we would therefore argue that tuning-fork tests are misleading; audiometry is the only way of reliably detecting, distinguishing and quantifying the deficit. We suggest that, if it were possible, all the 256 and 512 Hz tuning-forks (for Weber’s or Rinne’s) would be smelted and re-made as (useful) 128 Hz ones (for vibration sense). This has the advantage of ensuring that the correct-sized tuning-fork is available in neurology clinics. Why does the neuromythology of this matter? Simply this: undergraduates spend too much time being taught useless parts of the physical examination, leading to confusion. As a result, we perceive that this makes them less likely to examine the parts that matter and adds to the misperception that the neurological examination is time-consuming, which is incorrect.9 We agree with Dr McBride who, in 1886, expressed his concerns in the British Medical Journal that “the time-honoured tuning-fork test must now be considered as uncertain”.10 Given the advent of audiometry, it seems remarkable that over 130 years later medical students still get taught this stuff, even though there is not a shred of evidence to support the tests’ ongoing use.
We thank Professor Martin Turner, Oxford, for suggesting the Beach Boys-inspired title.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed. This paper was reviewed by Michael Halmagyi, Sydney, Australia.