Table 2

 Performance of tests

CMAP, compound muscle action potential; DL, distal latency; FMNCV, fastest motor nerve conduction velocity; NAP, nerve action potential; NCS, nerve conduction studies; NMT, neuromuscular transmission; RNS, repetitive nerve stimulation.
Motor NCSTemperatureA low FMNCV or long DL may relate to low skin temperature. It should always be recorded and cool limbs heated
F wavesNo pathway length allowanceBecause of the long pathway, normal values have to be related to limb length, pathway measurement or body height
Motor NCSWaveform measurementsIf the onset of the CMAP is positive—that is, a downward deflection—the active recording electrode is misplaced away from the motor point and spurious latency and amplitude values will occur
If an initially positive going waveform is only seen in proximal median nerve stimulation there is a likelihood of an anatomic anomaly in the forearm
Motor and sensory NCSInadequate numbers of nerves studiedIn order to determine whether abnormalities are focal, generalised, length dependent, a sample of nerves in upper and lower limbs usually need to be studied
Even in simple entrapment neuropathies, studies should be bilateral and include at least one other nerve not under suspicion to exclude an underlying generalised process
Motor and sensory NCSInadequate range of stimulus sitesDependent on the clinical picture stimulation may be required proximally to detect focal proximal abnormality. However, particularly with root and plexus stimulation, care in ensuring supramaximal stimulation is vital
Motor NCSElectrode positioningIf optimal positioning not used, the indifferent electrode may contribute significantly to amplitude measurements
Motor NCSNerve length measurementA flexible ruler/tape is necessary to follow the nerve path accurately
Some measurements around joints are affected by flexion/extension of the joint. For example, the least errors are found in ulnar NCS if the elbow is flexed to >100°
Motor NCSNot supramaximalIf the CMAP is not recorded supramaximally at all sites, the measured FMNCV will be in error. Spurious conduction block may be seen if proximal nerve stimulation is not supramaximal
Motor NCSOver stimulationIf a nerve is stimulated with too high a current/voltage, adjacent nerves may be recruited and produce spurious results
Sensory NCSStimulus artefactsParticularly in the lower limb nerves the stimulus may cause an artefact which alters the NAP shape. This must be allowed for in calculation if it cannot be abolished
Sensory NCSAbsent potentialsIn some nerves, using the orthodromic technique, the 95% confidence limits may include zero microvolts. Thus an absent potential may not be “abnormal”
Motor and sensory NCSNormal valuesFailure to use age specific normal values in the very young and very old may lead to error
RNSInadequate positioning and movement restrictionIf the recording electrodes do not lie right over the motor point, a spurious decrement may be created as the muscle contracts repeatedly. Care over positioning and ideally splintage of the muscle so that it remains isometric helps
Inadequate facilitationPatients may need continuous encouragement to obtain 30–30 second maximal voluntary contraction
Too short an analysis periodRNS trains should be followed for at least 3 minutes (preferably 5) to pick up the occasional post-facilitation exhaustion
Limited sitesThe sensitivity of RNS in NMT disorders is greatly increased by studying both proximal (nasalis, trapezius) muscles as well as distal (abductor digiti minimi).
Wrong testAs a rule of thumb, a patient who has an unequivocally normal edrophonium test will frequently also have a normal distal and proximal RNS test. In these circumstances the authors go straight to single fibre electromyography of an intermediate and proximal muscle as a more sensitive test.