Proximal and distal motor nerve conduction in obturator and femoral nerves

1166

Proximal and Distal Motor Nerve Conduction in Obturator and Femoral Nerves Burhanettin Uludag, MD, Cumhur Ertekin, MD, A. Bulent Turman, MD, PhD, Dilek Demir, MD, Nefati Kiylioglu, MD ABSTRACT. Uludag B, Ertekin C, Turman AB, Demir D, Kiylioglu N. Proximal and distal motor nerve conduction in obturator and femoral nerves. Arch Phys Med Rehabil 2000;81: 1166-70. Objective: To study the proximal and distal motor conduction properties of obturator and femoral nerves. Design: For evaluation of distal motor conduction properties, obturator and femoral nerves were stimulated at the inguinal ligament, and M responses were recorded with needle electrodes from gracilis and rectus femoris muscles. Upper lumbar roots were stimulated with needle electrodes inserted between L1-L2 vertebral laminae. Participants: Sixteen healthy adults, eight of each gender, age 22 to 52 years (mean 37.5). Main Outcome Measures: Description of a method for assessing motor conduction along the obturator nerve and evaluating proximal motor conduction measurements obtained with stimulation of obturator and femoral nerves. Results: Distal motor conduction latencies were 3.9 ⫾ 0.7 msec for gracilis and 4.6 ⫾ 0.5msec for rectus femoris after stimulation of obturator and femoral nerves, respectively. Proximal conduction times from lumbar vertebral level were 10.4 ⫾ 0.3msec for the obturator nerve and 10.8 ⫾ 0.4msec for the femoral nerve. Conduction velocities of proximal segments of both nerves were similar, 62m/sec for the obturator nerve and 65m/sec for the femoral nerve. Submaximal stimulation of both nerves evoked H-reflex responses from their associated muscles. Conclusions: Motor conduction properties of the obturator nerve can be assessed by the method described, particularly to differentiate between peripheral, plexus, or radicular lesions that involve the obturator nerve. Key Words: Obturator nerve; Femoral nerve; Nerve conduction; Electric Stimulation; Lumbar root stimulation; H-reflex; Rehabilitation. r 2000 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation

electrodiagnostic investigations and studies.2 In several previous clinical studies on ON blocks, the ON was stimulated by needle electrodes at the level of the obturator canal and the muscular contractions of thigh adductors were either observed clinically3 or recorded by surface electromyogram (EMG) electrodes.4,5 In these studies, however, the values of motor conduction time were not identified and conduction measurements on the proximal segment of the ON were not attempted. Another nerve of the same region, the femoral nerve (FN), also originates from L2, L3, and L4 spinal nerves. It is formed from the posterior divisions of these nerves and emerges beneath the inguinal ligament to innervate the thigh flexors, including the rectus femoris (RF) muscle. The relatively superficial anatomic location of this nerve around the inguinal ligament has made it possible to be studied electrophysiologically since Gassel first described the electrodiagnostic method for the FN in 1963.6 Gassel investigated in detail the electrophysiologic features of distal motor conduction of FN and the M responses of thigh muscles.6 However, systematic investigations of the more proximal part of FN have not been possible because of anatomic difficulties in reaching this nerve in the pelvic and lower back regions. In certain clinical conditions, such as in patients with unilateral quadriceps muscle weakness and wasting, it may be necessary to delineate the lesion level, ie, at lumbar roots or lumbar plexus, or to differentiate the distal FN-ON involvement. This study investigated an electrophysiologic method for the assessment of the ON in healthy human subjects and evaluated proximal motor conduction measurements that were systematically carried out for both ON and FN. METHODS

T

Subjects Sixteen healthy adult volunteers (8 women, 8 men) participated in the study. The subjects were excluded if they had any history of lower extremity injuries, radiculopathies, or other lesions that could affect findings. Their age ranged from 22 to 52 years, with a mean age of 37.5 years. The study was conducted with the approval of the Ethics Committee of Ege University Hospital.

From the Departments of Clinical Neurophysiology and Neurology, Ege University Hospital, Bornova, Izmir, Turkey (Uludag, Ertekin, Demir, Kiylioglu); and the Department of Biomedical Sciences, University of Sydney, Australia (Turman). Accepted December 23, 1999. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. Reprint requests to Cumhur Ertekin, MD, Ege University Hospital, Department of Neurology and Clinical Neurophysiology, 35100, Bornova, Izmir, Turkey. 0003-9993/00/8109-5726$3.00/0 doi:10.1053/apmr.2000.6972

Obturator Nerve Stimulation With the subject lying down, the ON was stimulated percutaneously at the level of the pubic tubercule. Once the pubic tubercule was palpated by the examiner, bipolar stimulation electrodes (Medelec 16893a) were placed 1 to 2cm inferior and 1 to 2cm lateral to the tubercule. The position of electrodes was close to the lateral fold of skin of the pubic region and slightly medioinferior to the inguinal ligament. To accomplish maximum stimulation of the ON, it was necessary to press firmly the stimulation electrode over the lateral fold of skin of the pubic region. At the beginning of the study, needle electrodes were attempted for ON stimulation, but this was not tolerable for subjects. During ON stimulation, the volume-conducted potentials

HE OBTURATOR NERVE (ON) is formed from the ventral divisions of L2, L3, and L4 spinal nerves and provides the motor innervation of hip adductors, including the gracilis (GS) muscle. Although lesions of the ON are uncommon, it is vulnerable to entrapment, particularly where it passes through the obturator canal.1 Along its course, the ON follows a deep anatomic route and is therefore not readily accessible for

Arch Phys Med Rehabil Vol 81, September 2000

MOTOR CONDUCTION IN OBTURATOR NERVE, Uludag

arising from the muscles innervated by the FN were generally picked up by surface-recording electrodes (Medelec 17981a) (fig 1). Therefore, to record the activity of GS muscle without such contamination needle electrode (Medelec 53158a) recordings were employed. Before insertion of the disposable concentric needle electrode, subjects were asked to voluntarily contract their thigh muscles and the GS muscle was palpated during internal rotation of the knee. The insertion points were at the belly of these muscles, and distance between points of stimulation and recording was about 16 to 19cm for the GS muscle and about 19 to 21cm for the RF. Once the insertion point was identified, the electrode was introduced into the relaxed GS muscle. A second needle electrode was inserted into the RF muscle and connected to a separate channel of the EMG apparatus. The electrodes were positioned to ensure that when the subject was asked to contract the GS muscle there was no interfering EMG activity arising from the RF muscle, and similarly the GS muscle was also required to be silent during voluntary contractions of the RF muscle. The ON was stimulated by means of rectangular electric pulses (0.2–0.5msec duration) and the stimulus intensity was increased in a stepwise manner until a supramaximal M response was obtained from the GS muscle. Usually, the maximum stimulation of the ON at the pubic region was achieved at around the maximum output levels of the stimulator. If the ON stimulation evoked an M response in the RF

Fig 1. Electromyogram recordings obtained with bipolar (A,B) surface electrodes, and (C,D) needle electrodes, in response to femoral nerve (FN) and obturator (ON) nerve stimulation over the inguinal ligament. Note the contamination by volume-conducted activity with surface recordings. RF, rectus femoris; GS, gracilis; St, stimulation.

1167

muscle, which sometimes occurred, the recordings obtained from the GS muscle were not included in the study. This was also the case for the RF muscle’s M response during FN stimulation. For both nerves, the distances between the cathode of the stimulation electrodes and the needle insertion points were measured. Femoral Nerve Stimulation The M responses from the RF muscle were recorded in response to FN stimulation at the inguinal ligament level, with the subject in the same position as for ON stimulation. The pulsation of the femoral artery was palpated and the bipolar surface stimulation electrodes were placed lateral to the point of pulsation over the inguinal ligament while the concentric needle electrode was inserted into the RF muscle. As in the case of ON stimulation, it was ensured that there was no interference of activity picked up by the GS muscle electrode during maximum voluntary contraction of the RF muscle. Lumbar Root Stimulation The method for electric stimulation of lumbar roots was previously described in detail by Ertekin and colleagues.7-9 Briefly, after the assessment of distal motor latencies for the ON and FN, the subject was moved to a lateral recumbent position so the leg studied was positioned above the other leg, and a pillow was placed in between them. During this maneuver, care was taken to ensure that the needle electrodes within GS and RF muscles were stable. The sterilized stainless steel needle electrodes (TECA MG-50a ) coated with Teflon, except at the tips, were used for electric stimulation. The longer cathode electrode (4.5cm) was inserted between the spinal processes at L1 to L2, or rarely at L2 to L3 levels. The shorter anode (2cm) was inserted subcutaneously on the midline 2 or 3 spines above the level of the cathode. To position the cathode near the lumbar roots, it was necessary to advance it slowly while electric pulses (duration 1msec; intensity 50–75V) were delivered at a random rate. As the tip of the electrode approached the posterior lamina level of the lumbar spine being investigated, segmental motor contractions from both legs were often observed. After minor adjustments to the position of the electrode until an M response with the highest amplitude and stable configuration could be obtained from RF and GS muscles, the electric stimulus intensity was increased (100–150V) to achieve stable and reproducible responses from these muscles. Recording and Analysis Procedures The M responses from GS and RF muscles were simultaneously recorded onto a 2-channel EMG recorder (MystroMedelec MS-20a ) with the filter settings of low frequency at 10Hz and of high frequency at 10kHz, respectively. The sweep duration was at 20 or 50msec and the gain was adjusted according to the magnitude of M responses. The latency values to the first deflection, whether positive or negative, and the peak-to-peak amplitudes of M responses were measured for analysis. During both lumbar root and peripheral ON and FN stimulations the electric stimulus intensity was adjusted to exclude the possibility of occurrence of an H-reflex and other late responses. For each stimulation-recording trial three to five M responses were collected. The distances between the root stimulation point at the lumbar laminar level and the peripheral stimulation points of ON and FN were measured by means of obstetric pelvimetry. The latency difference between the M responses obtained by lumbar root stimulation and by periphArch Phys Med Rehabil Vol 81, September 2000

1168

MOTOR CONDUCTION IN OBTURATOR NERVE, Uludag

eral stimulation of ON and FN was used to calculate the proximal motor conduction velocities (MCVs) for ON and FN. The mean, standard deviation (SD), and standard error of the mean (SEM) values were calculated and the tests of significance (Student’s t test) applied to the results obtained. RESULTS Reliable and isolated recordings of EMG activity over the RF and GS muscles could not be obtained by means of surfacerecording electrodes during either the ON or the FN stimulation. Figure 1 shows the effect of spread of volume-conducted activity from one muscle to the recording site over the other muscle in response to stimulation at the inguinal ligament level. During the electric stimulation of ON (fig 1B) both surfacerecording electrodes picked up the muscle-evoked activity generated by adductor muscles. Similarly, when the FN was stimulated, the volume-conducted activity of the quadriceps muscle was picked up by electrodes placed over the GS muscle (fig 1A). Therefore, it was necessary to use concentric needle electrodes to avoid such contamination from the activity of distant muscles. Figure 2 shows the EMG activity recorded from RF and GS muscles by means of concentric needle electrodes in response to isolated electric stimulation of the FN (top traces) and the ON (middle traces). The distal motor conduction latency to the RF muscle in response to FN stimulation was 4.6 ⫾ 0.5msec (table 1). This was slightly longer than the latency to the GS muscle after ON stimulation (3.9 ⫾ 0.7msec). The difference in the latencies of these two nerves may be due to the relatively shorter peripheral anatomic route of the ON. However, when the amplitudes of M responses were measured and compared the peak-to-peak amplitude of the M response

Fig 2. Superimposed traces of M responses obtained by stimulation of distal (at the inguinal ligament) and proximal (at the L1-L2 spine level) segments of femoral nerve (FN) and obturator nerve (ON). Needle electrode recordings are from rectus femoris (RF) and gracilis (GS) muscles.

Arch Phys Med Rehabil Vol 81, September 2000

Table 1: Summary of Motor Conduction Parameters for Obturator and Femoral Nerves (n ⴝ 16)

Latency from IL (msec) Latency from L1-L2 (msec) IL to muscle distance (cm) L1-L2 to IL distance (cm) Proximal MCV (m/sec) M-response amplitude (mV) IL stimulation M-response amplitude (mV) L1-L2 simulation

Obturator Nerve

Femoral Nerve

3.9 ⫾ 0.7 (2.8-5.3) 10.4 ⫾ 0.3 (7.5-12.4) 17.1 ⫾ 1.4 31.8 ⫾ 0.8 61.8 ⫾ 1.9 (50.0-75.0) 7.2 ⫾ 0.7 (2.3-12.7) 5.9 ⫾ 1.0 (2.0-17.3)

4.6 ⫾ 0.5 (4.0-5.5) 10.8 ⫾ 0.4 (8.0-14.0) 20.0 ⫾ 1.0 29.6 ⫾ 0.3 65.2 ⫾ 3.7 (42.0-88.0) 12.4 ⫾ 1.1 (2.5-20.0) 4.8 ⫾ 0.5 (2.1-10.0)

Values reported as mean ⫾ SEM (range). Abbreviations: IL, inguinal ligament; MCV, motor conduction velocity.

for FN stimulation was higher in comparison to that for ON stimulation; 12.4 ⫾ 1.1mV and 7.2 ⫾ 0.7mV, respectively (table 1). This discrepancy between latency and amplitude measurements reflects the fact that amplitude measurements of responses obtained by needle electrodes are not reliable indicators of the overall EMG activity in response to maximum nerve stimulation. The M responses evoked in the RF and GS muscles in response to stimulation of lumbar roots at the laminar level are shown in bottom traces of figure 2. The onset latencies from the spinal level to both muscles were not significantly different; 10.8 ⫾ 0.4msec for RF and 10.4 ⫾ 0.3msec for GS (table 1). The mean proximal MCVs for the ON and FN were 61.8 ⫾ 1.9 and 65.2 ⫾ 3.7m/sec, respectively (table 1). Their conduction velocity values were within acceptable ranges for fastconducting motor fibers. Overall, because of the anatomic depth of ON at the peripheral stimulation site, to evoke a maximum response from the GS muscle it was necessary to deliver the maximum output strengths of the stimulator. In several subjects (n ⫽ 6) late EMG responses were studied on a longer sweep duration during stimulation of the ON and FN. Submaximal electric stimulation of both obturator and femoral nerves evoked a late response associated with the H-reflex from the GS and RF muscles, respectively. The H-reflex responses obtained from these muscles are shown in figure 3. It can be seen clearly that in both muscles the H-reflex is evoked at submaximal electric stimulus intensities but disappears as the stimulator output increases to levels at which a maximum M response is obtained. The mean latency of the H-reflex response was 17.4 ⫾ 1.4msec for the RF and 16.8 ⫾ 1.6 msec for the GS muscle. DISCUSSION The ON is not readily accessible for routine electrophysiologic studies. The major limitation for assessing this nerve is associated with its deep anatomic route within the pelvis and the thigh.2 However, in this study it was possible to achieve the maximum stimulation of ON at the level of the pubic tubercule by firmly pressing the stimulation electrodes onto the skin. Therefore it was possible, for the first time, to evaluate the motor conduction properties of ON by means of intramuscular EMG activity recordings evoked in response to electric stimulation at peripheral as well as lumbar spine levels. It should be noted that the amplitude measurements of M responses recorded by needle electrodes are not reliable

MOTOR CONDUCTION IN OBTURATOR NERVE, Uludag

1169

Fig 3. The H-reflex and M responses obtained from rectus femoris (RF) and gracilis (GS) muscles in response to gradually increasing intensities of electric stimulation of femoral nerve (left traces) and obturator nerve (right traces) at the inguinal ligament.

indicators of conduction parameters. During ON stimulation, however, attempts to record the M responses by means of surface electrodes were unsuccessful. There are several factors associated with the failure of this method. First, the electric activity in distant muscles contaminated the responses investigated in the target muscle, namely the activity recorded from RF and GS muscles in response to FN and ON stimulation, respectively. As the latency values for RF and GS muscles were close to each other, the activity at the onset point of the M response could have been contaminated by the volumeconducted activity arising from the other muscle. Second, in the majority of subjects, the ON stimulation required the maximum stimulator output. A maximal M response was assured as a stabile M response with consecutive stimuli at same intensity level. If a stable maximal M response could not be obtained with this manner until maximal stimulator output in any subject, this subject dropped out of the study. For instance, rarely in obese subjects was it not possible to reach the maximum M response levels of the GS muscle and these subjects were not included in the study. Because the distance between the nerve stimulation and EMG recording sites were relatively short (17.1 ⫾ 1.4cm for GS and 20.0 ⫾ 1.0cm for RF), this in itself could have been the source of contamination spreading from stimulating electrodes to surface recording electrodes. Third, some adductor muscles, such as the adductor longus muscle, may in some cases receive their innervation from the FN instead of the ON,2 and thus be a potential source of contamination. Therefore, rather than using surface electrodes, which pick up the activity from a relatively larger surrounding area, it is more reliable to use the needle electrode technique that records the activity from a smaller area around its tip within the muscle. The results from latency measurements revealed that the peripheral motor conduction time for FN is within compatible

ranges of measurements reported in previous studies that used surface- or needle-recording techniques.6,10,11 The mean distal latency for the ON was slightly shorter than that of the FN. This significant difference appears to be associated with the relatively shorter peripheral route followed by the ON (17.1 ⫾ 1.4cm), compared with the FN (20.0 ⫾ 1.0cm) (table 1). The electric stimulation of lumbar roots at the laminar level produced a distinct M response with constant latency to thigh muscles in all subjects studied. The mean latency measurement of around 10msec is comparable with values reported for the RF muscle in previous studies.7-9,12-14 The proximal MCV of ON is not significantly different than the conduction velocity of FN (61.8 ⫾ 1.9m/sec for ON and 65.2 ⫾ 3.7m/sec for FN; table 1). This suggests that the lumbar roots, spinal nerves, and proximal nerve fibers of both nerves contain the fastest conducting motor fibers with comparable diameters. Because of their circumferential anatomic course of the ON and FN within the pelvic cavity, the proximal distances measured by pelvimetry might not have been very accurate. However, the MCV values are within acceptable ranges and similar to other nerves for which the proximal MCVs could be measured.15-17 During submaximal electric stimulation of both the ON and FN at the distal stimulation point, a late response associated with the H-reflex could be evoked from the GS and RF muscles, respectively. With increasing stimulus intensities that reached supramaximal levels these responses disappeared and were replaced by M responses (fig 3). The H-reflex has been demonstrated in muscles other than the soleus, including the quadriceps femoris muscle.18-20 The latency measurements of this potential (17.4 ⫾ 1.4msec for RF and 16.8 ⫾ 1.6msec for GS) indicate the values for conduction of impulses to spinal cord on Ia afferents and efferent conduction from the spinal cord to muscles on fast motor fibers. A potential problem with the H-reflex recorded in response to ON stimulation may occur if this late potential is interpreted as a prolonged M response. Arch Phys Med Rehabil Vol 81, September 2000

1170

MOTOR CONDUCTION IN OBTURATOR NERVE, Uludag

This may become even more important in subjects with normal than higher thresholds for ON stimulation. Magnetic stimulation is another method for the stimulation of spinal roots and can also be used for assessing ON conduction. However, clinical experience so far suggests that supramaximal stimulation of lumbar roots can not be achieved with magnetic root stimulation.7,9,21,22 The laminar electric stimulation of lumbar roots has no serious risks, apart from a mild unpleasant sensation that may be experienced by some subjects.7,9 Nonetheless, this condition appears to be a minor obstacle because individuals receive only a few electric shocks during the assessment. Overall, the clinical pathologies of ON are uncommon. Nevertheless, traumatic lesions of the lumbosacral plexus attributable to fractures of the pelvis, acetabulum, or femur, or surgery on the proximal femur and hip joint may involve the ON.23-25 A study on 53 cases of lumbosacral plexopathy injuries reported a 3.5% occurrence of ON involvement.26 Because the ON is located considerably lateral to the aorta, it is rarely affected by direct aortic compression. However, hematomas in the psoas muscle may compress it together with the FN.27,28 The ON may also be damaged during labor29 and various operations such as prolonged urologic30 or abdominal surgery.31 In lithotomy position the angulation of the nerve as it leaves the obturator canal or compression between the fetal head and the pelvic wall may damage the ON. Entrapment neuropathies of the ON may have been underestimated, as the majority of cases are reported from clinically manifest conditions with a history of an underlying cause, such as trauma or surgery. Recently, however, obturator neuropathies in athletes with chronic groin pain have been reported in a series of 32 cases.32 The fascial entrapment of the ON where it enters the thigh was the cause in all cases. This finding suggests that sports-related ON entrapments may well be more prevalent than were previously considered. The correct diagnosis of this condition requires reliable and accurate electrodiagnostic tests and measurements. Furthermore, in the clinical setting, it may be necessary to differentiate between peripheral, plexus, or radicular lesions that involve the ON. This technique has some restrictions for measuring ON peripheral conduction velocity due to using concentric needle electrode, but it may prove to be useful in a variety circumstances such as in diagnosing obturator neuropathy and in differentiating between lesions of the root, plexus, and nerve. References 1. Nakano KK. The entrapment neuropathies. Muscle Nerve 1978;4: 264-79. 2. Mumanthaler M, Schliack H. Peripheral nerve lesions diagnosis and therapy. Stuttgart-New York: Georg Thieme Verlag; 1991. 3. Magora F, Rozin R, Ben-Menachem Y, Magora A. Obturatory nerve block: an evaluation of technique. Br J Anaesth 1969;41: 695-8. 4. Atanassoff PG, Weis BM, Brull SJ, Horst A, Kuilling D, Stein R, et al. Electromyographic comparison of obturatory nerve block to three-in-one block. Anesth Analg 1995;81:529-33. 5. Atanassoff P, Weis BM, Brull SJ, Horst A, Ku¨lling D, Stein R, et al. Compound motor action potential recording distinguishes differential onset of motor block of the obturatory nerve in response to etidocaine or bupivacaine. Anesth Analg 1996;82:317-20. 6. Gassel MM. A study of femoral nerve conduction time. Arch Neurol 1963;9:607-14. 7. Ertekin C, Nejat RS, S¸irin H, Selc¸uki D, Arac¸ N, Ertas¸ M, et al. Comparison of magnetic coil stimulation and needle electrical stimulation in the diagnosis of lumbosacral radioculopathy. Clin Neurol Neurosurg 1994;96:124-9. 8. Ertekin C, Nejat RS, S¸irin H, Selc¸uki D, Arac¸ N, Ertas¸ M. Comparison of magnetic coil and needle-electrical stimulation in diagnosis of lumbosacral radioculopathy. Muscle Nerve 1994;17:685-6.

Arch Phys Med Rehabil Vol 81, September 2000

9. Ertekin C, S¸irin H, Koyuncuogˇlu HR, Mungan B, Nejat RS, Selc¸uki D, et al. Diagnostic value of electrical stimulation of lumbosacral roots in radiculopathies. Acta Neurol Scand 1994;90: 26-33. 10. Chopra JS, Hurwitz LJ. Femoral nerve conduction in diabetes and chronic occlusive vascular disease. J Neurol Neurosurg Psychiatry 1968;31:28-33. 11. Ertekin C. Saphenous nerve conduction in man. J Neurol Neurosurg Psychiatry 1969;32:530-40. 12. Bischoff C, Meyer B-U, Machetanz J, Conrad B. The value of magnetic stimulation in the diagnosis of radiculopathies. Muscle Nerve 1993;16:154-61. 13. Maertens de Noordhout CD, Rothwell JC, Thompson PD, Day BL, Marsden CD. Percutaneous electrical stimulation of lumbosacral roots in man. J Neurol Neurosurg Psychiatry 1988;51:174-81. 14. Ugawa Y, Rothwell JC, Day BL, Thompson PD, Marsden CD. Magnetic stimulation over the spinal enlargements. J Neurol Neurosurg Psychiatry 1989;52:1025-32. 15. Liguori R, Krarup C, Trojaborg W. Determination of the segmental sensory and motor innervation of the lumbosacral spinal nerves. Brain 1992;115:915-34. 16. Wochnic-Dyjas D, Glazowski C, Niewiadomska M. The F-wave in the vastus lateralis M. and the segmental motor conduction times for L2/L4 motor neurons. Electroencephalogr Clin Neurophysiol 1996;101:379-86. 17. Zhu Y, Starr A, Su Hwan S, Woodward G, Haldeman S. The H-reflex to magnetic stimulation of lower limb nerves. Arch Neurol 1992;49:66-71. 18. Aiello I, Serra G, Rosati G, Tugnoli V. A quantitative method to analyse the H-reflex latencies from vastus medialis muscle: normal values. Electromyogr Clin Neurophysiol 1982;22:251-4. 19. Jusic A, Baraba R, Bogunovic A. H-reflex and F-wave potentials in leg and arm muscles. Electromyogr Clin Neurophysiol 1995;35:471-8. 20. Kameyama O, Hayes KC, Wolfe D. Methodological considerations contubuting to variability of the quadriceps H-reflex. Am J Phys Med Rehabil 1989;68:277-82. 21. Britton TC, Meyer BU, Hardmann J, Benecke R. Clinical use of the magnetic stimulator in the investigation of peripheral conduction time. Muscle Nerve 1990;13:396-406. 22. Macdonell RAL, Cros D, Shahani BT. Lumbosacral nerve root stimulation comparing electrical with surface magnetric coil technique. Muscle Nerve 1992;15:885-90. 23. Weber ER, Daube JR, Coventry MB. Peripheral neuropathies associated with total hip arthroplasty. J Bone Joint Surg 1976;58: 66-9. 24. Melamed NB, Satya-Murti S. Obturator neuropathy after total hip replacement. Ann Neurol 1983;1:578-9. 25. Siliski JM, Scott RD. Obturator-nerve palsy resulting from intrapelvic extrusion of cement during total hip replacement: report of four cases. J Bone Joint Surg 1985;67:1225-8. 26. Stoehr M. Traumatic and postoperative lesions of the lumbosacral plexus. Arch Neurol 1978;35:757-60. 27. Kubacz GJ. Femoral and sciatic compression neuropathy. Br J Surg 1971;58:580-2. 28. Fletcher HS, Frankel J. Ruptured abdominal aneurysms presenting with unilateral peripheral neuropathy. Surgery 1976;79:120-1. 29. Stewart JD. Focal peripheral neuropathies. New York: Elsevier; 1987. 30. Pellegrino MJ, Johnson EW. Bilateral obturator nerve injuries during urologic surgery. Arch Phys Med Rehabil 1988;69:46-7. 31. Bischoff C, Scho¨nle PW. Obturator nerve injuries during intraabdominal surgery. Clin Neurol Neurosurg 1991;3:73-6. 32. Bradshaw C, McCrory P, Bell S, Brukner P. Obturator nerve entrapment: a cause of groin pain in athletes. Am J Sports Med 1997;25:402-8. Supplier a. Medelec Ltd., Manor Way, Old Woking, Surrey GU229JU, UK.