The importance of the orientation of the electrode plates in recording the external anal sphincter EMG by non-invasive anal plug electrodes
N. R. Binnie, B. M. Kawimbe, M. Papachrysostomou, N. Clare and A. N. Smith
University Department of Surgery, Western General Hospital, Edinburgh, UK Accepted: 1 November 1990
Abstract. Two non-invasive anal plug electrodes of similar size have been compared, one with the electrode plates orientated circularly in the anal canal and the other with the plates in the long axis of the anal canal. There was a significant increase in the amplitude in the EMG signals recorded at rest and during squeeze from the external anal sphincter with a longitudinally placed electrode in 117 patients. Inappropriate contraction of the external anal sphincter when straining at stool was more readily detected using the longitudinal electrode in 52 patients investigated for intractable constipation. The longitudinal electrode detected the amplitude of the response to the elicitation of a pudeno-anal reflex more readily than the circular electrode. When in 12 of the 117 the pudeno-anal reflex EMG signal was either absent or not detected with the circumferential plug electrode, the longitudinal electrode detected the presence of a low amplitude response in 11 of these. When the non-invasive longitudinal electrode was compared to invasive fine wire stainless steel electrodes, a correlation was found for external anal sphincter resting EMG (r=0.99, p<0.01), voluntary squeeze EMG (r=0.99, p<0.001) and strain EMG (r=0.91, p<0.01). The longitudinal anal plug electrode thus facilitates surface acquisition of EMG activity.
Resume. Deux electrodes anales non invasives de contact d’un diametre similaire ont ete comparees, l’une avec les plaques d’electrode orientees circulairement dans le canal anal et l’autre avec les plaques suivant le grand axe du canal anal. II existait une augmentation significative de l’amplitude des signaux electriques enregistres au repos et durant Ia contraction du sphincter anal externe pour les electrodes placees lonitudinalement chez 117 patients. Une contraction inappropriee du sphincter anal externe durant Ia defecation etait plus facilement detectee en utilisant des electrodes longitudinales chez 52 patients explores pour constipation irreductible. L’electrode Iongitudinale detectait l’amplitude de reponse au declenchement du reflexe pudendo-anal plus facilement que l’electrode circulaire, mais chez 12 des 117 malades le signal electrique du reflexe pudendo-anal etait soit absent, soit n’etait pas detecte avec une electrode circonferentielle tandis que l’electrode longitudinale detectait Ia presence d’une reponse de basse amplitude chez 11 d’entre eux. La comparaison entre l’electrode longitudinale non invasive et l’electrode invasive par fil d’acier inoxydable fin mon-trait une correlation pour l’EMG de base du sphincter anal externe (r=0.91, p<0.01), le trace de contraction volontaire (r=0.99, p<0.001) et le trace de defecation (r=0.91, p<0.01). Ainsi l’electrode longitudinale par plaque ameli ore l’enregistrement de l’activite electromyographique.
Anal plug electrodes usually have two circular electrode plates circumferential to the anal plug which thus lie parallel to the external anal sphincter muscle fibres. Due to the increased longitudinal conductivity of muscle, it is accepted that bipolar surface electrodes for recording striated muscle EMG are usually placed in the direction of the muscle fibres . Each electrode is connected to either side of a balanced amplifier while a third electrode connects the patient to ground . Because of this theoretical implication, an anal plug electrode was constructed with two electrode plates placed along the long axis of the anal plug, the electrodes being equally separated on the circumference. The electrode plates were thus separated along the length of the external anal sphincter muscle fibres. A series of anorectal electrophysiological investigations was done to compare the two types of anal plug electrodes which differed with their circular and longitudinal arrangements. These included recording the integrated EMG activity of the external anal sphincter at rest, during voluntary squeeze and during straining, and while eliciting the pudendo-anal reflex latency response. Patients with anismus  who have an inappropriate contraction of the external anal sphincter can be recognized by detecting a rise in the external anal sphincter EMG activity when straining at stool. The normal response to straining at stool should be relaxation of the pelvic floor and external anal sphincter with a reduction in the EMG activity. The pudendo-anal reflex latency assesses the integrity of the reflex arc from the dorsal genital nerve to the S234 component of the sacral cord and the efferent pudendal nerve conduction time to the external anal sphincter .
A further aim of the study was to compare the more sensitive of the two types of non-invasive anal plug electrode with invasive fine wire stainless steel electrodes.
One hundred and seventeen patients participated in the comparison of the orientation of the plates of the two anal plug electrodes and had studies done at rest and during squeeze and strain manoeuvres. These 117 patients were 94 females (mean age 55.4±9.3 years) and 23 males of mean age 31±6.2 years. These patients, in addition, all had a pudendo-anal reflex study. Of the 117 patients, 52 patients showed, during straining, possible anismus effects. The patients who were being investigated for obstructive defaecation were represented by 38 females of mean age 48.2±11.5 years and 14 males (mean age 32.4±5.7 years).
Eight consecutive subjects attending the ano-rectal laboratory participated in a comparison of the more sensitive of the two plug electrodes with the fine-wire method. All patients gave informed consent for the procedures involved.
The electrode referred to as the circumferential electrode is the Disa 13k 78/79 (Disa Electronics, Bristol) (Fig. I a).
The longitudinal anal plug electrode has two 0.25 cm x 2 cm stainless steel electrode plates placed in parallel on the longitudinal axis of the neck of the plastic anal plug, separated equally on the circumference . The electrode plates are thus separated in the direction of the external anal sphincter muscle fibres (Fig. I b). The plastic plug is machined to the appropriate shape with an overall length of 7 cm, a bulbous distal end 1.5 cm diameter and the neck of the plug being 2 cm in length and 0.5 cm in diameter.
The fine wire stainless steel electrodes [6, 7] were placed in a hypodermic needle with the ends protruding. The wires were insulated with a teflon coating and approximately 0.25 cm of the tips were bared and hooked over to allow the position to be maintained in the muscle on withdrawal of the needle.
All the subjects had investigation of rest, squeeze and strain EMG performed using either the circumferential electrode first and then the longitudinal one or else this was done in the reverse order, the order being varied but not formally randomised.
For the wire electrode study the subjects were placed in the left lateral position and the fine wire stainless steel electrodes introduced into the external anal sphincter in the mid-line posteriorly with a hypodermic needle to a depth of 1.5 cm. The needle was then withdrawn leaving the hooked electrodes in place in the external
anal sphincter muscle. The wires were connected to an isolated EMG integrator (Ormed 4880, MX216) and the recorder calibrated to read the required EMG range (0-50 µV). After insertional EMG activity ceased the resting EMG activity in the external anal sphincter was recorded. The subject was then asked to contract the external anal sphincter and the squeeze EMG was recorded. The subject was finally asked to strain as if at stool and the strain EMG was recorded. This procedure was repeated twice to confirm the recordings. The wire electrodes were now withdrawn and the anal plug electrode to be compared was inserted into the anal canal for the same investigations to be repeated.
The pudendo-anal polysynaptic reflex arc incorporates the sensory dorsal genital nerve, the S234 spinal cord and the efferent pudendal nerve to the external anal sphincter. The Medelec MS92a stimulus triggered response unit records the time delay from skin stimulus to the EMG response in the external anal sphincter. The time to onset of the digitally averaged response to at least 100 stimulus impulses is taken as the pudendo-anal reflex latency. The sensitivity of the incoming signal amplifier is selected to the most appropriate sensitivity for the amplitude of the incoming EMG signal from the muscle. If the amplifier sensitivity is set too high the display store will be overloaded and the signal peaks will be flattened. The amplitude of the pudendo-anal signal response in microvolts (µV) is therefore equal to the volts per division times the screen divisions, and this can be measured electronically within the apparatus. Finally, a paper copy of the pudendo-anal reflex response can be printed out with all that information included.
The anal plug electrode and fine wire electrode tests were compared by deriving p values and a correlation coefficient .
Fig. 4a, b. Mean pudendo-anal reflex EMG response amplitude (µV) showing an
augmented response with longitudinal anal plug electrodes but a barely detectable
response with the circular anal plug electrode
Orientation of electrodes
The rectified EMG amplitudes recorded from the external anal sphincter with the longitudinal electrode were significantly higher than those recorded with the circumferential electrode during rest and squeeze activity (Fig. 2). In 52 patients with obstructive defaecation there was an inappropriate contraction rectified EMG voltage when straining. The inappropriate EMG change was detected more readily by the longitudinal electrode than by the circular one (p<0.01) (Fig. 2). In the 117 patients who had the pudendo-anal reflex response studied, the signal latency obtained from the external anal sphincter was identical with the circular and longitudinal electrodes (45.3±2.3 ms). There was, however, a significantly higher response amplitude when the responses were recorded with the longitudinal electrode (p<0.01) (Fig. 3). In keeping with this, the sensitivity of the incoming signal amplifier of the EMG apparatus could be significantly reduced by a factor of 10. Figures 4a and b show the enhancement of the EMG voltage obtained with the longitudinal electrode compared to the circular one during the elicitation of the pudendo-anal reflex. There were 14 females (age 66.7± 8.6 years) in whom no pudendo-anal reflex could be detected using the circumferential electrode although the response was present with the longitudinal one.
Invasive versus non-invasive
There was a significant correlation between the rectified EMG voltage recorded with the longitudinal anal plug electrode and the fine wire electrode for the external anal sphincter resting EMG (r=0.99, p<0.01), the squeeze EMG (r=0.99, p<0.001) and the strain EMG (r=0.91, p<0.01) (Fig. 5). The direct correlation persisted between the two techniques in both the normal situation with reduction of EMG on straining, and in the abnormal state of anismus when there was inappropriate contraction of the external anal sphincter on straining at stool.
The familiar hour-glass-shaped anal plug electrode was originally made by Hopkinson for anal muscle stimulation with two parallel circumferential electrode plates 1 cm apart placed at either end of the neck of the electrode at the presumed site for stimulation of the “rectopubalis” muscle . However, the circular external anal sphincter and loop of pubo-rectalis muscle at the anorectal junction are arranged such that their muscle fibres lie in the same circle or loop shape, respectively. The longitudinal conductivity of muscle tissue is 5 to 15 times larger than in the transverse direction . Recording electrodes should therefore be placed longitudinally with respect to muscle [10). This is at variance with the design of the Hopkinson electrode used in the treatment of female urinary incontinence  and the currently available anal plug electrodes used for surface recording of the EMG activity of the external anal sphincter (Disa 13k78/79).
Electrophysiological investigation of the pelvic floor can provide detailed information on the pelvic floor muscle motor unit potential duration and single fibre density . The pudendo-anal reflex latency investigation is well tolerated, is much less invasive and is a useful ano-rectal EMG screening test in patients with neurogenic faecal incontinence . This study shows that an anal plug electrode with longitudinal electrode plates significantly improves the detection of the external anal sphincter EMG signal response on eliciting the pudendo-anal reflex.
When healthy skeletal muscle is completely relaxed it has no detectable EMG activity . The pelvic floor, external anal sphincter and external urethral sphincter are unusual in that they have continuous tonic EMG activity at rest , provided that the reflex arc is intact [14, 15]. The detection of changes in EMG activity in the pelvic floor can be recorded with a concentric needle electrode . The longitudinal electrodes detected these responses better than the circular ones and may be used to measure the reduction in EMG activity during normal defaecation straining, and to detect failure of this in obstructive defaecation or anismus . The use of artefact free fine wire electrodes is to be regarded as the ‘gold standard” for diagnosing inappropriate contraction of the external anal sphincter during dynamic studies while straining as if at stool . The presence of an anal plug within the anal canal might also induce non-physiological responses in the external anal sphincter. However, the significant direct correlation with the invasive fine wire electrode at rest (p <0.01), during squeeze (p <0.001) and straining (p <0.01) shows that this is not so. The inappropriate increases in the rectified EMG during straining in obstructive defaecation or anismus were detected as readily with the longitudinal anal plug electrode as they were with the fine wire electrodes.
Acknowledgements. Mr. N. R. Binnie was supported by a Scottish Hospital Endowment Research Trust Grant (No.720). Mr. B. M. Kawimbe was supported by a Janssen Pharmaceutical Research Fellowship. Dr. Maria Papachrysostomou was supported by a Scottish Home and Health Department Clinical and Biomedical Research Grant (K/MRS/5001202).
1. Lindstrom L, Petersen I (1983) Power spectrum analysis of EMG signals and its applications. In: Desmedt JE (ed) Computer-aided electromyography. Prog Clin Neurophysiol 10:1 -51
2. Lenman JAR, Ritchie AE, Simpson JA (1983) Clinical dectromyography, 3rd edn. Pitman Books, London, p 16
3. Preston DM, Lennard-Jones JE (1985) Anismus in chronic constipation. Dig Dis Sci 30:413-418
4. Varma JS, Smith AN, Mclnnes A (1986) Electrophysiological observations on the human pudendo-anal reflex. J Neurol Neurosurg Psychiatry 49:1414-1416
5. Unived Technologies Ltd (1987) Proceedings under the Patents Act 1977 in The Official Journal (Patents) No.5154; October 14th 1987, p2444
6. Basmajamian JV, Stecko G (1963) A new bipolar indwelling electrode for electromyography. J AppI Physiol 17: 849-852
7. Womack NR, Williams NS, Homefield JHM, Morrison JFB, Simpkins KC (1985) New method for the dynamic assessment of anorectal function in constipation. Br J Surg 72:994-998
8. Martin Bland J, Altman DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet i: 307-310
9. Hopkinson BR, Lightwood R (1966) Electrical treatment of anal incontinence. Lancet 11:297-298
10. Wynn Parry DB (1953) Electrical methods in diagnosis and prognosis of peripheral nerve injuries and poliomyelitis. Brain 76:229-265
11. Eriksen BC, Bergmann S, Mjolnerod OK (1987) Effect of anal electro-stimulation with the ~lncontan” device in women with urinary incontinence. Br J Obstet Gynaecol 94:147-156
12. Swash M, Snooks SJ (1985) Electromyography in pelvic floor disorders. In: Coloproctology and the pelvic floor: pathophysiology and management. Butterworths, London, pp 88-103
13. Parks AG, Porter NH, Melzak J (1962) Experimental studies on the reflex mechanism controlling the muscles of the pelvic floor. Dis Col Rec 5:407-414
14. Bishop B (1959) Reflex activity of external anal sphincter of the cat. J Neurophysiol 22:679-692
15. Gunterberg B, Kewenter J, Petersen I, Stener B (1976) Anorectal function after major resection of the sacrum with bilateral or unilateral sacrifice of sacral nerves. Br J Surg 63:546-554
16. Bartolo DCC, Jarratt JA, Read NW (1983) The use ofconventional EMG to assess external anal sphincter neuropathy in man. J Neurol Neurosurg Psychiatry 46:1115 – 1118
Prof. A. N. Smith
University Department of Surgery
Western General Hospital
Edinburgh EH4 2XU