British Journal of Rheumatology 1989;28:528-533

CLINICAL PRACTICEA STUDY OF UPPER LIMB PAIN AND PARAESTHESIAE

FOLLOWING NECK INJURY IN MOTOR VEHICLEACCIDENTS: ASSESSMENT OF THE BRACHIAL

PLEXUS TENSION TEST OF ELVEYBY J. L. QUINTNER

SlJohn of God Medical Centre, 175 Cambridge Street, Wembley WA 6014, Australia

SUMMARYThe brachial plexus tension test (BPTT) appears to offer a means of examining the extensibility and mech-anosensitivity of the neural tissues related to an upper limb. This test was used to evaluate possible cervicalor brachial plexus neural involvement causing arm pain syndromes in 37 patients presenting for assessmentfollowing neck injury in motor vehicle accidents. The BPTT was considered positive in 55 of the 61 sympto-matic arms. There were no false-positive responses in the 13 asymptomatic arms although a slight loss ofextensibility was evident in five arms.

Twenty patients without clinical evidence of current or previous neck pathology were similarly exam-ined. There were no reports of pain on BPTT in this group. In 36 of the 40 arms a full range of extensibilitywas present. In the other four arms the loss of extensibility was slight.

This study suggests that arm pain and paraesthesiae which follow neck injury in motor vehicle accidentsarise from irritable cervical neural tissues. The study also highlights the persistent nature and widespreaddistribution of the pain in these patients.

KEY WORDS: Cervical spine, Neuropathic pain, Brachial plexus tension, Motor vehicle accident.

THE assessment of patients with persistent painsyndromes of the head, neck and arm followinginjury to the neck in motor vehicle accidents hasbeen a difficult area of musculoskeletal medicine[1—3]. Upper limb pain and paraesthesiae arepart of the complex symptomatology in thesepatients [1, 2,4-7].

Elvey [8] developed the Brachial Plexus Ten-sion Test (BPTT) to help the examiner differen-tiate between local arm conditions causing painand cervical/brachial plexus neural conditionswith pain referral into the arm. The test utilizesmanoeuvres which, used step-wise, increase ten-sion within neural tissues related to the upperlimb.

Cadaver studies demonstrated that movementof the arm in certain directions resulted in move-ment and tension of cervical nerve roots andtheir investing sheaths, and of anatomically re-lated dura [9]. The greatest movement tookplace at C5 and C6 with some effect at C7. Theposition of the upper quarter that placed maxi-mum tension on the cervical nerve root com-plexes combined glenohumeral joint abduction(110 degrees) and external rotation to maximumrange with the arm behind the coronal plane,

Submitted 16 February; revised version accepted 13May 1989.

Correspondence to Dr. Quintner.

with elbow extension, wrist extension, supina-tion of the forearm, shoulder girdle depressionand lateral flexion of the neck to the con-tralateral side [10] (Fig. 1).

Further cadaver studies have shown thatshoulder abduction to 105 degrees markedlyincreases tension in brachial plexus componentsand confirm that the other manoeuvresdescribed by Elvey [8] result in additional ten-sion [11, 12].

Kenneally et al. [13] documented the normalresponses at end range of the BPTT in more than400 asymptomatic volunteers. These indicatedthat C5, C6 and C7 nerve roots were implicatedin the test to a greater degree than C8 and Tlroots. The symptoms most consistently evokedwere a deep stretch or ache sensation in the cub-ital fossa extending down the anterior and radialaspects of the forearm into the radial side of thehand, and a definite tingling sensation in thethumb and the first three fingers.

A clinical validation study of the BPTT foundthat the test had moderate to high intra-exam-iner reliability (0.825) and that it could be usedin the clinical situation to discriminate betweenthe presence or absence of a brachial plexusinvolvement in patients with upper limb symp-toms [14].

The aim of the present study was to assess the

528

at South Metropolitan A

rea Health Services on A

ugust 20, 2013http://rheum

atology.oxfordjournals.org/D

ownloaded from

QUINTNER: UPPER LIMB PAIN AND PARAESTHESIAE 529

FIG. 1.—Brachial plexus tension test of Elvey. The shoulder girdle is gently depressed. The arm is then positioned inshoulder abduction (approximately 110 degrees) and external rotation, forearm supination, wrist extension and fingerextension. In this figure, contralateral flexion of the cervical spine is also shown. The range of elbow extension is thencarefully assessed and the subjective response to testing is noted. (Courtesy of Robert Elvey).

response to BPTT, first in patients in whomthere was little likelihood of cervical or brachialplexus neural abnormality, and then in a cross-section of patients who presented with a historyof neck injury in a motor vehicle accident fol-lowed by complaints of persistent neck paintogether with arm pain and/or paraesthesiae.

PATIENTS AND METHODSA study of the response to BPTT was under-

taken in patients with no current or past historyof neck pain. To be suitable for inclusion as a'control', each patient was required to have fulland pain-free cervical movements in all direc-tions, and a full range of passive movements atshoulder, elbow and wrist. There were 20patients in this group. In each patient thesymptomatic response to tension testing and therange of extensibility possible was noted for eacharm. The relevant musculoskeletal diagnosiswas recorded as well as details of age, sex andoccupational status. No attempt was made to

match this group for age and sex with the neckinjury group.

Forty patients with complaints of arm painand/or paraesthesiae as a result of neck injury ina motor vehicle accident who were referred tothe author between January 1986 and December1987 comprised the injury group. Three patientsknown to have neck pain prior to their accident,were excluded. All were pursuing claims forcompensation for personal injury. Apart fromone referred by an insurer, all had been sent byfamily doctors. Eight had been examined by theauthor on one occasion, six on two occasions and23 patients on three or more occasions.

A full medical history was obtained. The timeof onset of the arm symptoms after the accidentwas determined. Spinal and lower limb symp-toms were also noted. Physical examination wasdirected toward local pathology in the upperlimb which, if present, could be a cause ofsymptoms.

Brachial plexus tension testing [8] was per-

at South Metropolitan A

rea Health Services on A

ugust 20, 2013http://rheum

atology.oxfordjournals.org/D

ownloaded from

530 BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXVIII NO. 6

formed on each arm of each patient in bothgroups. The author, sequentially, positioned theshoulder joint in abduction and external rotationand then added shoulder girdle depression, fore-arm supination, wrist and finger extension and,finally, elbow extension. The position of lateralneck flexion to the contralateral side was onlyused in the examination of the symptomaticarms if the other manoeuvres were of full rangeand failed to provoke a symptomatic response.This latter position was used in the examinationof all asymptomatic arms. The technique ofadministering the test included sufficient varia-tions of the anatomical positions to make itunlikely that a false-positive response could beobtained. A test result was considered clinicallyrelevant (positive) if symptoms were reproducedand there was loss of the normal full range of ex-tensibility of the arm [13]. If the result was equiv-ocal, testing was repeated at a futureconsultation and if symptoms were not repro-duced, the test was considered to be negative.

The examiner was not blinded but an attemptwas made to determine the sensitivity of theBPTT by its ability to reproduce the arm symp-toms in the injury group. The specificity of thetest was reflected by the extent to which a normalresponse was obtained in an asymptomatic arm.The values for sensitivity and for specificitycould vary between 0 and 1.

Cervical radiology was available for 31patients. Electrodiagnosis testing of median andulnar nerve function had been performed on twopatients.

RESULTSControl group

There were 11 females and nine males, meanage 42 years (range 14-75 years). Twelve wereemployed, three had retired, three were house-wives and two were at school. Lower limb ar-thropathy or enthesopathy was present in 11.Localized synovitis involved the hands in twopatients. Two patients suffered from tenniselbow and two from tenosynovitis.

In 36 of the 40 upper limbs, a full range of ex-tensibility was possible. The other four limbslacked full extensibility by less than 20 degrees atthe elbow. The most frequently reported sub-jective response was a stretch or pulling sensa-tion in the cubital fossa (19 upper limbs); in 12limbs no symptoms were felt. The remainingupper limb responses consisted of a pulling sen-sation deep within the forearm (four), a pullingsensation in the palm of the hand (three) and a

stretch or pulling sensation in the shoulderregion (two). These responses were consistentwith those in an asymptomatic population [13].

Injury groupThirty-one were female, six male, mean age

34 years (range 12-63). Twelve patients wereworking and 16 were unemployed as a result oftheir injury. Eight had not been working at thetime of injury and one was still at school.

Seven had been involved in two accidents, atotal of 44 reported accidents. In 30 of these thepatient had been the driver of a vehicle. Twenty-five collisions were rear-end, 10 side-on, fivewere front-end and four were varied in type. Thetime interval between the accident (initial) andtheir referral ranged from 4 to 76 months (mean28 months). Onset of arm symptoms occurredwithin 3 months of accident in 24; in 13, the onsetwas noted 3 or more months later. The meandelay in onset was 12 months (range 3-31months). Nine of these 13 patients were inemployment of a clerical nature at the time ofonset of arm symptoms.

Symptoms are shown in Table I. Spinal painwas generally more severe in the neck, upperback and occipital regions than in the mid-backor low back. There were 61 symptomatic armsand 13 asymptomatic arms. One patientreported unilateral paraesthesiae in the absenceof pain in either arm. Pain in the absence of para-esthesiae was reported in 17 arms, arm paraes-thesiae in the absence of pain in six arms.

Arm pain was more commonly related to com-ponents of the upper portion of the brachialplexus than to those of the lower portion (seeTable II). The distribution of the paraesthesiaewas variable. In 15 they involved the wholehand, in two patients the whole arm and in twothe forearm, wrist and hand. Nine describedthem in all fingers, four in the ring and little fin-gers, three in the ulnar three fingers, one in thethumb, index and middle fingers and one eachinvolving the thumb and the middle finger. In nopatient was a musculoskeletal abnormality of theupper limb detected. All but two patientsdemonstrated painful limited range of active cer-vical movement in at least one direction. Fifteenexhibited cervical movements which were lim-ited and painful in all directions.

Table III shows the degree of correlationbetween the presence of arm symptoms and theresult of brachial plexus tension testing in the 61symptomatic arms. The overall sensitivity valueof the BPTT was 0-9. In the 13 asymptomatic

at South Metropolitan A

rea Health Services on A

ugust 20, 2013http://rheum

atology.oxfordjournals.org/D

ownloaded from

QUINTNER: UPPER LIMB PAIN AND PARAESTHESIAE 531

TABLE ISYMPTOMS IN INJURY GROUP (n = 37)

Site of painNeckHeadUpper back—rightUpper back—leftMiddle backLower back

Right arm painLeft arm pain

Bilateral arm painUnilateral arm pain

Arm paraesthesiaeBilateral paraesthesiaeUnilateral paraesthesiae

Arm pain and paraesthesiaeBilateral both symptomsUnilateral both symptoms

Unilateral paraesthesiae,no arm pain

Patients(«)

372928201623

3025

1917

1318

2413

1

Per cent

100%78%76%54%43%62%

55%45%

51%46%

35%49%

65%35%

3%

arms, the test produced either no response or thenormal response. In five of the asymptomaticarms the test demonstrated a minor (less than 20degrees) loss of the normal range of extensibilityas in four arms of the control group. The speci-ficity value of the BPTT in both the control andthe injury group was 1 but it is not knownwhether a minor loss of normal range could bedue to a subclinical abnormality of neural tissuesrelated to the upper limb.

There were two patients in whom BPTT re-peatedly failed to reproduce their reportedsymptoms. They both presented with unilateralarm pain; in addition, one experienced paraes-thesiae in that arm, the other paraesthesiae inboth arms.

TABLE IINEUROLOGICAL LEVEL OF ARM PAIN (n 55 ARMS)

Level

C5C6C6/7C7C8Total

Arms

912513

30

Right armBPTT

pos

812512

28

BPTTneg

100012

Arms

1011400

25

Left armBPTT

pos

97400

20

BPTTneg

140005

Neurological deficits were found in one arm ineach of five patients. These comprised absentbiceps jerk in two with diminution of C5 sensa-tion in one. Diminished sensation in the C6 andC7 distribution was seen in three. BPTT waspositive in each of these five arms.

Radiological changes were reported in 12 ofthe 29 patients with cervical radiology availablefor review. A reversal of the lower cervical lord-osis was found in one patient; disc degenerationin nine; an end-plate fracture at C6 in one andmimimal compression fracture of C6 in another.CT scan of the cervical spine performed in fourpatients confirmed the end-plate fracture of C6,showed a discogenic canal stenosis at C6/7 inanother and did not reveal an abnormality in twopatients.

The nine patients with cervical spondyloticchanges were older (mean age 41 years) than thegroup of patients as a whole (mean age 31 years).Symptoms in these patients were bilateral in fourand unilateral in the others.

As shown in Table IV, BPTT was positive in89% of the symptomatic arms of patients withnormal cervical radiology and in 83% of thesymptomatic arms of those with abnormalradiology.

One patient with bilateral arm pain and para-esthesiae had a normal nerve conduction study;another with bilateral symptoms had carpaltunnel syndrome. Decompression surgery didnot alleviate arm pain but did lessen the severityof paraesthesiae.

DISCUSSIONSciatic pain has been shown to arise from lum-

bar and lumbosacral nerve root tissue whichexhibits hyperalgesia [15, 16]. Straight-leg-rais-ing is therefore used in the routine examinationof patients with lower limb pain syndromeswhich may arise from these tissues [17—21].

This study supports the use of a comparabletest, the BPTT, in a clinical setting where armpain and/or paraesthesiae could reasonably beexpected to arise from hyperalgesic cervical orbrachial plexus neural tissues. The positive re-sponses in the symptomatic arms resulted frommovement and tension imparted by the test tothese tissues. As elbow extension was the lastmanoeuvre performed as part of the BPTT, andneural tissues are the only tissues with an ana-tomical connection between the neck and thewrist, elbow extension is likely to have trans-mitted both movement and tension upwards,particularly along the median nerve, into the

at South Metropolitan A

rea Health Services on A

ugust 20, 2013http://rheum

atology.oxfordjournals.org/D

ownloaded from

532 BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXVIII NO. 6

Symptoms

PainParaesthesiaeBoth symptoms

TABLE IIICORRELATION OF ARM SYMPTOMS WITH BRACWAL PLEXUS TENSION TESTING

Total arms (n)

554461

BPTT positive

48(87%)40 (91%)55 (90%)

BPTT negative

7 (13%)4 (9%)6(10%)

cervical and brachial plexus neural tissues [22].No false-positive responses were noted in theasymptomatic arms of the inj ury group nor in thearms of the control group.

In the injury group, limitation of the normalrange of extensibility of neural tissues betweenthe neck and the wrist was always found whenthe test reproduced the arm symptoms. Thislimitation was interpreted as due to reflex con-traction of the upper limb and cervical paraspi-nal muscles in order to protect the hyperalgesiccervical neural tissues.

The BPTT requires further study in patientpopulations with known (demonstrable) cervicaland upper limb neural pathology in order toestablish its place in clinical medicine. However,there are no examination techniques compar-able to the BPTT which can provide the samepositive and objective confirmation of thepatient's arm symptoms. For example, theshoulder abduction test of Davidson et al. [23]relieved the unremitting cervical radicular painin 15 of 22 patients with cervical extradural com-pressive radiculopathy confirmed on myelogra-phy. This test is purely subjective; the authorspresumed that tension on cervical nerve rootswas lessened when the test position was adoptedby the patient.

Upper limb pain or numbness occurringshortly after a neck injury may correlate with abad prognosis for recovery [24, 25]. Both themechanism of injury to cervical neural tissues inmotor vehicle accidents and the resulting path-ology have been poorly understood. Directtrauma may stretch or compress nerve rootsleading to swelling and vascular congestion. At a

CERVICAL

Radiology

NormalAbnormalNoneTotal

TABLE IVRADIOLOGY AND ARM PAIN (n = 36 PATIENTS*)

Totalpatients

17127

36

Painfularms

27181055

BPTTpos.

24 (89%)15 (83%)9(90%)

48(87%)

BPTTneg.

3(11%)3 (17%)1 (10%)7(13%)

*One patient did not report arm pain.

later stage, fibrosis may form within the nerveroot together with adhesions between the rootand surrounding dural sleeve or foraminal tis-sues [1, 4-6]. Irritation of nerve roots as a resultof injury to intervertebral discs, or to interver-tebral joints has also been postulated [4]. Symp-toms occurring soon after injury may be relatedto secondary spasm in the scalene muscles caus-ing irritation to the brachial plexus [2, 26]. Ani-mal experiments support the hypothesis thatthese arm pain syndromes may arise from chron-ically injured dorsal nerve roots or dorsal rootganglia [27, 28].

Nordin et al. [29] demonstrated ectopicimpulse formation in patients with paraesthesiaeresulting from different types of nerve disorders.These firing patterns were provoked using thedifferent limb manoeuvres which also provokedparaesthesiae, thus confirming increased mecha-nosensitivity of these neural tissues. In addition,Asbury and Fields [30] recently provided a hypo-thetical basis for pain of nerve trunk origin.Their hypothesis explains both the neural symp-toms following neck injury and their reproduc-tion on BPTT.

Those patients who have objective signs ofupper limb neurological deficit after neck injuryin motor vehicle accidents have a poorer prog-nosis [7]. In this study, five patients were foundto have neurological deficits and in only two wasthe deficit objective. More sensitive means of as-sessing sensory function were beyond the scopeof my study but may have revealed abnormalityin a larger proportion of patients [31, 32].

ACKNOWLEDGEMENTSThe author thanks Mr. Robert Elvey for his

assistance in the preparation of this paper and formuch fruitful discussion; he also thanks ProfessorLawrence Beilin for his helpful comments andsuggestions.

REFERENCES1. Cammack KV. Whiplash injuries to the neck.

Am J Surg 1957;93:663-6.2. Macnab I. The 'whiplash' syndrome. Orthop

Clin North Am 1971;2:38SM03.3. De Palma AF, Subin DK. Study of the cervical

syndrome. Clin Orthop 1965^8:135-42.

at South Metropolitan A

rea Health Services on A

ugust 20, 2013http://rheum

atology.oxfordjournals.org/D

ownloaded from

QUINTNER: UPPER LIMB PAIN AND PARAESTHESIAE 533

4. Gay JR, Abbott KH. Common whiplash inju-ries to the neck. JAMA 1953;152:1698-704.

5. Seletz E. Whiplash injuries. Neurophysiolog-ical basis for pain and methods used forrehabilitation. JAMA 1958;168:1750-5.

6. Braaf MM, Rosner S. Whiplash injury of theneck: symptoms, diagnosis, treatment andprognosis. New York State J Med1958;58:1501-7.

7. Norris SH, Watt I. The prognosis of neck inju-ries resulting from rear-end collisions./ Bone Joint Surg [Br] 1983;65:608-ll.

8. Elvey RL. The investigation of arm pain. In:Grieve GP, ed. Modem manual therapy ofthe vertebral column. Edinburgh: ChurchillLivingstone, 1986: 530-5.

9. Elvey RL. Brachial plexus tension and thepathoanatomical origin of arm pain. In:Glasgow EF, Twomey L, eds. Aspects of ma-nipulative therapy. Melbourne: LincolnInstitute of Health Sciences, 1979: 105-10.

10. Reid SA. The measurement of tension changesin the brachial plexus. In: Dalzeill BA,Snowsill JC, eds. Proceedings of the FifthBiennial Conference, Manipulative Thera-pists' Association of Australia. Melbourne,1987: 79-90.

11. Ginn K. An investigation of tension develop-ment in the cords of the brachial plexusduring the upper limb tension test. In: Dal-zeill BA, Snowsill JC, eds. Proceedings ofthe Fifth Biennial Conference, ManipulativeTherapists' Association of Australia. Mel-bourne, 1987: appended 1-11.

12. Landers J. The upper limb tension test—theeffect of shoulder abduction on the testresponse. In: Dalzeill BA, Snowsill JC, eds.Proceedings of the Fifth Biennial Con-ference, Manipulative Therapists' Associ-ation of Australia. Melbourne, 1987: 1-12.

13. Kenneally M, Rubenach H, Elvey RL. Theupper limb tension test: the SLR of the arm.In: Grant R, ed. Physical therapy of the cer-vical and thoracic spine. New York: Church-ill Livingstone, 1988: 167-94.

14. Selvaratnam P, Glasgow EF, Mathias T. Thediscriminative validity of the brachial plexustension test. In: Dalzeill BA, Snowsill JC,eds. Proceedings of the Fifth Biennial Con-ference, Manipulative Therapists' Associ-ation of Australia. Melbourne, 1987:325-50.

15. Lindahl O. Hyperalgesia of the lumbar nerveroots in sciatica. Acta Orthop Scand1966;37:367-74.

16. Smyth MJ, Wright V. Sciatica and the interver-tebral disc. An experimental study. / BoneJoint Surg [Am] 1958;40:1401-18.

17. Charnley J. Orthopaedic signs in the diagnosisof disc protrusion. With special reference tothe straight-leg-raising test. Lancet1951;i:186-92.

18. Goddard MD, Reid JD.Movements induced

by straight-Ieg-raising in the lumbo-sacralroots, nerves and plexus, and in the intrapel-vic section of the sciatic nerve. J Neurol Neu-rosurg Psychiatr 1965;28:12-18.

19. Breig A, Troup JDG. Biomechanical consider-ations in the straight-leg-raising test. Cada-veric and clinical studies of the effects ofmedial hip rotation. Spine 1979;4:242-50.

20. Falconer MA, McGeorge M, Begg AC. Obser-vations on the cause and mechanism ofsymptom-production in sciatica and lowback pain. J Neurol Neurosurg Psychiatr1948;ll:13-26.

21. Edgar MA, Park WM. Induced pain patternson passive straight-leg-raising in lower lum-bar disc protrusion. A prospective clinical,myelographic and operative study in 50patients. / Bone Joint Surg [Br]1974;56:658-67.

22. Ewing MR. Postoperative paralysis in theupper extremity. Lancet 1950;i:99-103.

23. Davidson RI, Dunn EJ, Metzmaker JN. Theshoulder abduction test in the diagnosis ofradicular pain in cervical extradural com-pressive monoradiculopathies. Spine1981 ;6:441-6.

24. Hohl M. Soft-tissue injuries of the neck inautomobile accidents. Factors influencingprognosis. J Bone Joint Surg [Am]1974;56:1675-82.

25. Greenfield J, Ilfield FW. Acute cervical strain.Evaluation and short term prognosticfactors. Clin Orthop 1977;122:196-200.

26. Wickstrom J, LaRocca H. Management ofpatients with cervical spine and head injuriesfrom acceleration forces. Curr Pract OrthSurg 1975 ;6:83-98.

27. Howe JF, Loeser JD, Calvin WH. Mechano-sensitivity of dorsal root ganglia and chron-ically injured axons: a physiological basis forthe radicular pain of nerve root compress-ion. Pain 1977;3:25-^1.

28. Wall PD, Devor M. Sensory afferent impulsesoriginate from dorsal root ganglia as well asfrom the periphery in normal and nerve in-jured rats. Pain 1983;17:321-39.

29. Nordin M, Nystrom B, Wallin U, HagbarthK-E. Ectopic sensory discharges and paraes-thesiae in patients with disorders of periph-eral nerves, dorsal roots and dorsal columns.Pain 1984;20:231^t5.

30. Asbury AK, Fields HL. Pain due to peripheralnerve damage. Neurology (Cleveland)1984^4:1587-90.

31. Lindblom U, Verrillo RT. Sensory functions inchronic neuralgia. J Neurol Neurosurg Psy-chiatr 1979;42:422-35.

32. Szabo RM, Gelberman RH, Dimick MP. Sen-sibility testing in patients with carpal tunnelsyndrome. J Bone Joint Surg [Am]1984;66:61-4.

at South Metropolitan A

rea Health Services on A

ugust 20, 2013http://rheum

atology.oxfordjournals.org/D

ownloaded from