[Conduction block: a notion to let through]

Historical study of electrodiagnosis indicates that nerve conduction block is an old notion, used as early as the second century by Galien and then early in the 19th by physiologists such as Müller and Mateucci. Although introduced into the field of human pathology by Mitchell in 1872, who used it to study nerve injuries, and then by Erb in 1874 to study radial palsy, the contribution of nerve conduction blocks to electrodiagnosis was not exploited until the 1980s. At that time, attempts to improve early diagnosis of Guillain-Barré syndrome showed that among the electrophysiological consequences of demyelination, conduction block was the most appropriate to account for the paralysis. At the same time, descriptions of neuropathies characterized by conduction blocks led to considering conduction block as a major electrophysiological sign. Why was it so difficult for this sign to be retained for electrodiagnosis? Since the notion is not always associated with anatomical lesions, it doesn't fit easily into anatomoclinical reasoning, but has to be thought of in functional terms. Understanding how an uninjured axon could fail to conduct action potentials leads to an examination of the intimate consequences of demyelinations and axonal dysfunctions. But some of the difficulty encountered in adding this new old sign to the armamentarium of electrophysiological diagnosis was related to the technical precautions required to individualize a block. Several pitfalls have to be avoided if a conduction block is to be afforded real diagnostic value. Similar precautions and discussions are also needed to establish an opposing sign, the "excitability block" or "inverse block".