Secondary Logo

Journal Logo

Original Article

Electrophysiological tests in intensive care

Botteri, M.*; Guarneri, B.

Author Information
European Journal of Anaesthesiology: February 2008 - Volume 25 - Issue - p 174-180
doi: 10.1017/S0265021507003201
  • Free



Electroneurography is the technique that enables us to study electric conduction in sensory and motor nerves. In combination with electromyographs they provide an indispensable diagnostic approach for recognizing and localizing peripheral nerve pathologies [1-8]. Methodology and recording techniques consist of stimulating the nerve using electrodes, normally surface ones, positioned along the nerve trunk or on the area it innervates, and measuring the potentials of the sensory response (along the nerve pathway) or motor response (from the belly of the innervated muscle). This enables us to measure motor and sensory conduction speed and to analyse the parameters of the response potential (size, amplitude and duration) [9].

Motor conduction speed

The potentials evoked in muscle are usually measured using surface electrodes (and very occasionally with needle electrodes) positioned on the belly of the muscle (active electrode) or on the muscle tendon (reference or indifferent electrode). The stimulation electrodes are placed along the nerve at specific anatomical points with the anode 2-3 cm proximal compared to the cathode so that the active electrode is nearer to the recording electrodes.

The negative charge produced by the electric stimulus tends to depolarize the underlying nerve, causing an action potential to be fired, which spreads in both the orthodromic and antidromic directions along the axons.

Square and rectangular wave impulses are delivered, whose duration varies from 0.1 to 1.0 ms, usually at a frequency of 1 Hz, and sufficiently intense to excite all the fibres present in the nerve (known as the supramaximal level, which may be more than 200 V). The voltage applied to reach supramaximal level obviously depends on the duration of the impulse and tissue resistance.

The muscle potential (M response) obtained from the stimulation of the motor fibres in the mixed nerve or action potential in the compound muscle (compound muscle action potential, CMAP) reflects electrical activity in both the nerve component and the muscle. In normal conditions, the amplitude is quite high (from 5 to >10 mV) and varies according to the number of underlying fibres excited [8].

A reduction in CMAP amplitude can show axonal neuropathy or a primary muscular problem. However, even a myelin type of neuropathy can present with reduced CMAP because of temporal dispersion of the potentials caused by the different conduction speeds of the fibres (desynchronization of the potential) or because of the secondary axonal suffering (tardive).

Motor conduction speed can be obtained by stimulating the nerve trunk at two different points: the difference between the two latencies (proximal and distal), measured in ms, constitutes conduction speed along the segment of the nerve being tested. If the length of the segment tested is measured in mm, we can easily measure motor conduction speed expressed in m s−1. The distal latency alone does not give a real indication of conduction speed along the nerve because this includes synaptic delay due to transmission by the neuromuscular end-plate (about 0.5 ms).

The most representative motor nerves, and those most frequently studied in intensive care are:

Peroneal nerve:

  • derivation: Pedidio muscle;
  • stimulation points: dorsum of foot (distal), peroneal head (proximal).

Tibial nerve:

  • derivation: abductor hallux muscle;
  • stimulation points: internal malleolus (distal), polpliteal fossa (proximal).

Median nerve:

  • derivation: abductor brevis muscle of the thumb;
  • stimulation points: wrist (distal), elbow-biceps sulcus (proximal).

Ulnar nerve:

  • derivation: abductor muscle of little finger;
  • stimulation points: wrist (distal), elbow-epitrochlea-olecranon (proximal).

Sensory conduction speed

Sensory nerves are stimulated by applying electrodes (ring or metal plate) along the pathway of the nerve under investigation. The sensory nerve active potential (SNAP) is measured by superficial or subcutaneous electrodes positioned away from the stimulation electrodes along the nerve pathway [8,9]. The amplitude of the sensory action potentials is greatly reduced (10-15 μV) so it is easy to confuse them with the background noise. To see them, we need to use a special technique called ‘averaging' (the total and then average of more than one event), which enables us to isolate the evoked response from the background noise that has the same latency (temporal interval) and the same morphology.

Sensory conduction can also be recorded using the orthodromic (direction of centripetal physiological transmission) or antidromic (proximal stimulation and distal derivation) techniques. Sensory conduction velocity (VCS) is calculated in m s−1 and is measured by dividing the distance in mm between the stimulation and registration points by the latency time.

The most representative sensory nerves and those most frequently used in intensive care are:

Sural nerve:

  • derivation: external submalleolar area;
  • stimulation points: extension of triceps tendon at a distance of approximately 13 cm from the derivation electrodes.

Tibialis posterior nerve:

  • derivation: internal submalleolar area;
  • stimulation points: ring electrodes positioned on hallux.

Median nerve:

  • derivation: wrist;
  • stimulation: ring electrodes positioned on index and middle fingers.

Ulnar nerve:

  • derivation: wrist;
  • stimulation: ring electrodes on little finger.

Table summarizing electroneurographical parameters

Motor nerve:

  • motor conduction velocity (MCV), measured in m s−1;
  • proximal and distant latency (ms);
  • amplitude of action potential (CMAP) measured in mV.

Sensory nerve:

  • sensory conduction velocity (VCS), measured in m s−1;
  • latency of sensory action potential (ms);
  • amplitude of SNAP, measured in μV.

From the observation and analysis of the electroneurographical data obtained, three major types of alteration were seen.

  • Reduced amplitude of CMAP or SNAP, with normal or slightly reduced conduction values: this is seen when there is a partial lesion of the nerve, neurpraxia or axonotmesis (before distal degeneration sets in) when conduction is blocked in the affected axons or when there is axonal neuropathy with loss of axons and consequent reduction in the motor and/or sensory action potential. Nerve conduction speed is often normal or only slightly reduced because of degeneration of the thicker and therefore faster fibres.
  • Normal amplitude and reduced conduction values: this is found with segmental demyelination, affecting a majority of the nerve fibres, as in some acute inflammatory polyneuropathies, leading to significant reduction in conduction velocity and/or dispersion of the action potential (multiple focal blocks).

This slowing down of conduction speed, which is so important, is typical of entrapment syndromes that present initially with signs of myelin damage and only later, as the noxious stimulus persists, with signs of axonal damage.

  • Absence of response: when there is total interruption of electric conduction (axonotmesis, neurotmesis).

Conduction velocity measurement has the advantage of being a quantitative measure, which does not rely on patient collaboration or the examiner's subjective impressions. There are, nonetheless, many possible sources of error, particularly in intensive care, and only careful observance of certain rules (cleaning the skin, careful positioning of the electrodes, very precise measurement) enables us to avoid errors while carrying out or evaluating the tests [10]. Conduction velocity is closely linked to body temperature and therefore it is important to control this properly using a lamp if necessary. Johnson and Olsen [11] showed that a 1°C change in body temperature corresponds to changes in conduction values of about 5%.

F response

Conduction in proximal or deep segments of nerves or nerve roots can be studied by recording the ‘F' wave. If we apply supramaximal stimulus after the ‘M' response (motor action potential), we can see a lower amplitude second response with a latency of about 25-30 ms. This is the alpha motorneurone spinal response after antidromic activation (the impulse travels in an antidromic direction as far as the cell body of the alpha motorneurone and then travels back in an orthodromic direction along the same axon) [8,9].

The latency of the ‘F' wave may be longer in patients with peripheral neuropathy, nerve root lesions or nerve entrapment. With the Guillain-Barrè syndrome, which hits proximal nerve segments first, there may be increased latency in the F wave before there is any evidence of nerve conduction velocity slowing down [12,13].

H reflex

Like the F wave, the H reflex can be used to evaluate electric conduction velocity in proximal segments of nerve roots or in peripheral nerves (radiculopathies and peripheral neuropathies). Unlike the ‘F' response, the ‘H' reflex, by definition, is a response which is evoked by activation of a monosynaptic spinal pathway (diastaltic arc). It is often used to study the function of the S1 nerve root (stimulation of sciatic-internal popliteal in popliteal fossa and recording of reflex response in soleus muscle) [8].

Repetitive stimulation

This is the method most widely used to test function in the neuromuscular end-plate whenever neuromuscular block treatment has been used or when myasthenic or related pathologies are suspected [8,9].

The most significant transmission defects are:

  • defects in transmitter synthesis (acetylcholine);
  • defects in transmitter release (botulin toxin);
  • competitive block (curare);
  • depolarizing blocks (decamethonium, suxamethonium).

The method (Desmedt test) involves stimulating a peripheral nerve (usually the ulnar nerve at the wrist) and recording the response in the corresponding muscle (little finger abductor for ulnar nerve). Ten impulses are delivered at supramaximal intensity. In a healthy subject, repetitive stimulation at low frequencies (usually 3 Hz) does not cause any significant reduction in the amplitude of the response potential, whether at rest or under stress.

In patients treated with neuromuscular block, the initial CMAP, which is of practically normal amplitude, tends to reduce in amplitude after only a few repetitive stimuli (decreasing response).

In patients with myasthenia gravis, there is an initial decrease in CMAP (until the 4th or 5th stimulus), followed by a slight increase after continued stimulation (decreasing-increasing response). These alterations reflect transmission problems of a post-synaptic nature (reduction in the number of receptor sites for acetylcholine).

With the Lambert-Eaton syndrome (myasthenic syndrome), where the deficit is at the pre-synaptic level (deficit release of acetylcholine by the synaptic vesicles), repetitive stimulation facilitates transmission and leads to progressive increase in the amplitude of CMAP. This is demonstrated only with repetitive high-frequency stimulation (at least 50 Hz).


Electromyography (EMG) is a neurophysiopathological method which, by means of coaxial needle electrodes inserted into the muscle belly, studies electrical activity in the muscle at rest and during contraction, and gives information about function, structure and the recruitment of motor units to the highest level of activation (interferential pattern) [8,9].

Physiology of the motor unit

The motor unit is made up of a motor nerve fibre and all the muscle cells it innervates; it acts as the contractile functional unit in that all the cells within it contract synchronously when the motor fibre is excited [14,15].

The cell bodies of the motor nerve fibres (alpha motor neurones) are situated in the ventral horn of the spinal cord. The axon exits via the ventral route and reaches the muscle via the peripheral nerve (usually mixed) and as it gets close to the nerve fibre it gradually loses its myelin sheath. The force of a muscle contraction depends on the number of motor units recruited, their type, the frequency of firing and the velocity of contraction. The number of motor units varies greatly from one muscle to another. The muscle fibres in the motor unit are spread out within the muscle, and the fibres innervated by the same cell in the anterior horn are generally not contiguous. There are two types of motor unit: those with type I fibres, which contract slowly for continuous, prolonged activity (energy supplied by oxidative mitochondrial metabolism), and motor units with type II fibres, which have rapid conduction and are used for more intense physical effort and rapid movements (energy is mediated by anaerobic glycolysis).

Morphology patterns in EMG recordings

Execution and recording according to the standard method require:

  • looking for any spontaneous activity at rest (insertion, fibrillation, positive slow waves, fasciculation potentials, myotonic activity and pseudomyotonic firing);
  • analysing individual motor unit potentials (MUP) (morphology, amplitude and duration);
  • assessing the voluntary recruitment of motor units to reach maximal force (interferential pattern).

Spontaneous activity

In a healthy muscle, at rest, there is no electrical activity to record (electrical silence). However, when the electrode is in an end-plate zone, there may prove to be some activity in isolated fibres or small groups of fibres even in a healthy subject (end-plate noise or activity). The insertion of a needle electrode into a muscle can also cause action potentials of short duration and low amplitude to fire, probably due to mechanical excitation by the electrode (insertion activity). When there is muscle pathology, either primary or secondary due to denervation, the electrical activity lasts for longer at rest and there are also associated polymorphous patterns of spontaneous activity.

  • Fibrillation potentials: these are small potentials of brief duration (0.5-2 ms) and low amplitude (30-150 μV), of biphasic or triphasic morphology. They probably originate from the spontaneous firing of single muscle fibres. The pathogenic mechanism of fibrillation is not very clear. It has been demonstrated, however, that denervated muscle has an abnormally increased sensitivity to acetylcholine. Fibrillation potentials are typical of denervated muscle and motor neuron lesions, that is, the connection between the axon and muscle fibre has been broken, but they can also be seen in other types of myopathy.
  • Positive sharp waves (PSW) or Jasper potentials are of greater duration and amplitude compared to fibrillation potentials. Characteristic of their morphology is an initial positive deflection with sharp peak followed by a slow negative wave, which can last over 10 ms. They most likely show firing in structurally damaged muscle fibres.
  • Fasciculation potentials: these are characterized by the spontaneous and synergic firing of groups of muscle fibres belonging to the same motor unit. They are big enough to produce a visible contraction though not actual movement. These potentials are usually present in degenerative illnesses of the anterior horn (motor neuron diseases, syringomyelia, acute phase of poliomyelitis) but they can also appear with radiculopathies.
  • Myotonic firing: high-frequency firing of action potentials that can appear spontaneously (spontaneous myotonic activity), be evoked by the movement of the electrode or by percussion of the muscle (mechanical myotonic activity) or follow on from a voluntary contraction during the relaxation phase (post-contraction myotonic activity). These are present in myotonic diseases with or without dystrophy. What is unusual about them is the progressive increase in their frequency of firing (from 10 to 150 Hz), followed by a reduction in amplitude and frequency of the activated potentials, as well as the associated characteristic sound effect of a ‘warplane dropping bombs'.
  • Pseudomyotonic firing: high-frequency firing, but unlike the myotonic type, the amplitude and frequency are constant and both the beginning and end are quite brusque. Also defined as ‘bizarre high-frequency firing', these are found with various pathologies like certain forms of neurogenic atrophy (spinal muscular atrophy) and myopathies (poliomyositis).

Motor unit potentials

During voluntary activation with minimal effort, individual motor unit potentials (MUP) can be isolated, which are the sum of all the potentials generated by groups of muscle fibres in almost simultaneous contraction. Morphologically they can be monophasic, biphasic, triphasic or polyphasic (with five or more phases). Their duration varies from between 2 and 10 ms; their amplitude, which varies from 100 μV to 2 mV, can, under pathological conditions, exceed 10 mV.

Both the morphology and size of MUP are significantly altered by disease processes. With peripheral neuropathy, partial denervation and regeneration phenomena often result and this leads to destructuring of the MUP, causing temporal dispersion, polyphasic morphology and an increase in their duration and amplitude.

If there is a primary muscle problem then we see a loss of muscle fibres but, at least in the beginning, there is no reduction in the number of axons. This means that both the amplitude and duration of single MUP are reduced, the morphology is predominantly polyphasic, closely serrated and the sound it makes is an unmistakable ‘dry, crackling noise'.

Voluntary contraction

As the muscle contraction gradually increases, we can see that more and more motor units are recruited until so many of them are superimposed that it becomes impossible to determine their individual characteristics. This is termed the ‘interference' pattern. When there are neurogenic problems, there is a progressive reduction in the recruitment until the picture we get of electric activity is the ‘intermediate' or ‘single oscillation' type. With primary myopathies, recruitment is not compromised and hence, because of the polyphasia of the MUP, we obtain the ‘interference' type of pattern even during weak voluntary activity (premature recruitment) with the distribution frequency of the electrical activity moving towards the high-frequency end of the spectrum.

Limitations of EMG in intensive care patients

For intensive care patients who are often comatose, sedated and mechanically ventilated, electrophysiology enables us to diagnose critical illness myopathy and neuropathy but it does not enable us to see the relative significance of the muscle or nerve component. If we apply the standard method, differential diagnosis between myopathy and neuropathy is extremely difficult if not impossible [16,17].

EMGs, by recording spontaneous activity, enable us to identify potentials of acute denervation like fibrillation and PSWs and, although in normal clinical practice activity like this is not usually recorded until between 1 and 3 weeks after the lesion occurs, in intensive care it is documented from the moment it first appears [6,18]. However, the presence of spontaneous denervation activity can be seen in both neuropathies and some types of acute myopathies. On the other hand, reduction in CMAP amplitude alone does not enable us to differentiate specifically between motor axonal neuropathy and myopathy.

Analysis of the MUP is therefore indispensable as is the study of recruitment patterns, although the latter demands collaboration of the patient, which is not usually possible in intensive care.

Direct stimulation of the muscle

With the aim of trying to make a differential diagnosis between myopathy (CIM) and polyneuropathy (CIP) in critically ill patients, we applied a method devised by Rich and colleagues [19,20], and then used by other authors [21,22], which enabled us to measure the amplitude of the motor action potential (CMAP) by stimulating the nerve (neCMAP) and direct stimulation of the muscle (dmCMAP). Patients identified with pathological CMAP on the standard ENG underwent direct muscle stimulation. This involves the placing of a monopolar subdermal needle electrode (anode) on the distal third of the tibialis anterior muscle, referred to a monopolar electrode (cathode) positioned 2 cm laterally.

The muscle is stimulated by gradually increasing the intensity of the electric impulse (0.1 ms duration, 0.5 Hz frequency) until we obtain a palpable muscular twitch (10-100 mA intensity).

Guided by the twitch, the recording electrode, a monopolar subdermal needle electrode (anode) is placed 2-3 cm down from the stimulating electrode and referred to a second monopolar electrode (subdermal or surface), which is placed a few centimetres distal to it. During stimulation, the recording needle electrodes are moved slightly in order to obtain the maximum amplitude of the muscle action potential (dmCMAP). By applying a supramaximal stimulus to the nerve underlying the muscle area under study, on the other hand, we obtain the classic compound action potential (neCMAP). By analysing the amplitude of the two responses from indirect stimulation (of the nerve) and direct stimulation of the muscle under investigation and by working out a mathematical ratio between neCMAP and dmCMAP, we obtain the nerve/muscle ratio, which, according to Rich [20], enables us to deduce the following:

  • when the ratio is greater than or equal to 1, it suggests either myopathy or a normal clinical picture. Differential diagnosis between these two conditions is easy to make based on the absolute value of the dmCMAP amplitude - whether it is normal or reduced (myopathy; loss of electrically excitable muscle);
  • a ratio of less than 1 with reduced neCMAP shows the presence of neuropathy.

To ensure greater reliability in the results obtained, we thought it better to introduce a modification to Rich's original method. Instead of measuring CMAP amplitude from peak to peak, we decided to measure from the isoelectric point to the first peak. The reasons for this were:

  • neCMAP is usually composed of a normally polymorphic and polyphasic response potential whose components are not homogeneous and temporally dispersed, probably because the recording electrode recruits motor units at the edge of the area of muscle fibre under study;
  • the dmCMAP we get is morphologically simplified (bi-, triphasic) and probably made up of more homogeneous components.

We therefore felt that it was better to measure amplitude from the isoelectric point to the first peak (morphologically similar and comparable to the response from direct stimulation of the muscle) to minimize possible error introduced by the summation (temporo-spatial) of late or borderline components recruited by a supramaximal stimulus applied higher up the nerve trunk or, in any case, not directly belonging to the dmCMAP.

To confirm the validity and reliability of direct muscle stimulation, the technique was applied to patients both before and after curarization, thus obtaining stable and reproducible dmCMAP results.

It will be the subject of a future study to apply the technique to healthy subjects in order to obtain absolute normative data as reference values for direct diagnosis of myopathy without needing to use the neCMAP/dmCMAP ratio.


1. Bolton CF. Neuromuscular conditions in the intensive care unit. Intensive Care Med 1996; 22: 841-843.
2. Lacomis D, Petrella T, Giuliani MJ. Causes of neuromuscular weakness in the intensive care unit: a study of ninety-two patients. Muscle Nerve 1998; 21: 610-617.
3. Coacley JH, Nagendran K, Yarwood GD et al.. Patterns of neurophysiological abnormality in prolonged critical illness. Intensive Care Med 1998; 24: 801-807.
4. De Jonghe B, Cook D, Sharhar T et al.. Acquired neuromuscular disorders in critically ill patients: a systematic review. Intensive Care Med 1998; 24: 1242-1250.
5. Latronico N, Candiani A. Neuromuscular abnormalities in patients with organ failure and sepsis. Intensive Care Med 1994; 20: 612-613.
6. Latronico N, Fenzi F, Recupero D et al.. Critical illness myopathy and neuropathy. Lancet 1996; 347: 1579-1582.
7. Gutmann L, Gutmann L. Critical illness neuropathy and myopathy. Arch Neurol 1999; 56: 527-528.
8. Kimura J. Electrodiagnosis in disease of nerve and muscle: principes and practice. In: Davis FA, ed. Principles of Nerve Conduction Studies. Philadelphia, USA, 2001.
9. Lenman JAR, Ritche AE. Elettromiografia clinica. Raffaello cortina editore, 1983.
10. Van Dijk GJ, Van Benten I, Kramer CGS et al.. CMAP amplitude cartography of innervated by the median, ulnar, peroneal and tibial nerves. Muscle Nerve 1999; 22: 273-389.
11. Johnson EW, Olsen KJ. Clinical value of motor nerve conduction velocity determination. JAMA 1960; 172: 2030-2035.
12. Bolton CF, Laverty DA, Brown JD et al.. Critically ill polyneuropathy: electrophysiological studies and differentiation from Guillain-Barre syndrome. J Neurol Neurosurg Psychiatry 1986; 49: 563-573.
13. Olney RK, Aminoff MJ. Electrodiagnostic features of the Guillain-Barre syndrome: the relative sensitivity of different techniques. Neurology 1990; 40: 471-475.
14. Berne R, Levi MN. Trasmissione sinaptica. In: Fisiologia. Milano, Italy: Casa EA, ed. Ambrosiana, 1989: 52-70.
15. Berne R, Levi MN. Il muscolo come tessuto. In: Fisiologia. Milano, Italy: Casa EA, ed. Ambrosiana, 1989: 402-419.
16. Latronico N. Neuromuscular alterations in the critically ill patient: critical illness myopathy, critical illness neuropathy, or both? Intensive Care Med 2003; 29: 1411-1413.
17. Bednarik J, Lukas Z, Vondracek P. Critical illness polyneuromyopathy: the electrophysiological components of a complex entity. Intensive Care Med 2003; 29: 1505-1514.
18. Tennila A, Salmi T, Pettila V et al.. Early signs of critical illness polyneuropathy in ICU patients with systemic inflammatory response syndrome or sepsis. Intensive Care Med 2000; 26: 1360-1363.
19. Rich MM, Teener JW, Raps EC et al.. Muscle is electrically inexcitable in acute quadriplegic myopathy. Neurology 1996; 46: 731-736.
20. Rich MM, Bird SJ, Raps EC et al.. Direct muscle stimulation in acute quadriplegic myopathy. Muscle Nerve 1997; 20: 665-673.
21. Trojaborg W, Weimer LH, Hays AP. Electrophysiologic studies in critical illness associated weakness: myopathy or neuropathy - a reappraisal. Clin Neurophysiol 2001; 112: 1586-1593.
22. Lefaucheur JP, Nordine T, Rodriguez P et al.. Origin of ICU acquired paresis determined by direct muscle stimulation. J Neurol Neurosurg Psychiatry 2006; 77: 500-506.


© 2008 European Society of Anaesthesiology