As non-neurophysiologists we enjoyed this informative review of electromyography (EMG), and benefit enormously from the service provided by our local colleagues. We suggest, however, that EMG is neither ‘essential’ nor ‘diagnostic’ for motor neurone disease (MND). Rather, for at least the 85% of cases with classical amyotrophic lateral sclerosis, MND is a clinical diagnosis.

The average delay in diagnosis of MND from the onset of symptoms has stubbornly remained around 1 year over the last two decades.1 Over-reliance on investigations, including neurophysiology, excluding implausible mimic conditions may be a contributory factor. The first 12 months is precious time for those with a disease in which the median survival is less than 1000 days from symptom onset.2 The El Escorial criteria (EEC) were developed as a research tool, and a recent study of the performance of the modified ‘Awaji-shima’ version of the criteria demonstrated only 60% sensitivity.3 Regrettably, the EEC and its derivatives continue to cause confusion among patients labelled as ‘possible’ or ‘probable’, rather than ‘definite’ amyotrophic lateral sclerosis, simply on the basis of the number of regions with demonstrable mixed signs, despite the fact that the clinical diagnosis of MND is inescapable.

We have also observed a particular and unnecessary diagnostic delay in patients with regionally isolated corticobulbar involvement (historically referred to as Progressive Bulbar Palsy), because of a lack of detectable denervation on EMG (including sampling of tongue).

Bryan Matthews commented that ‘the EMG is at its most reliable when least needed.’4 We recognise an invaluable role for neurophysiology in the detection of wider occult lower motor neurone involvement in cases presenting with only monomelic signs, just as there is for neuroimaging where there is a plausible alternative explanation anatomically. Neurophysiology is also useful in confirming the suspicion of the rare and treatable mimic disorder multifocal motor neuropathy with conduction block, although a lack of wasting in relation to weakness, predilection for the finger extensors and far slower progression are all important clinical clues.

We urge the avoidance of EMG in those presenting with a primary complaint of fasciculations without detectable weakness, to minimise the relatively high chance of false-positive diagnosis.

It is acknowledged that EMG is currently the only MND biomarker routinely applicable to the single patient, among numerous other candidates emerging from neuroimaging and cerebrospinal fluid.5 The technique of electrical impedence myography may prove to have greater sensitivity.6 Meanwhile we suggest that, particularly during a power-cut, neurologists should not hesitate to make a diagnosis of MND using their clinical acumen alone.