Archives ofDisease in Childhood 1993; 68: 215-218

An improved clinical method for detectingmeningeal irritation

Johny Vincent, Kurian Thomas, Omana Mathew

AbstractNeck stiffness is the most important sign ofmeningitis. When the neck is flexed, theinflamed nerve roots and meninges of thecervical region get stretched. This causesprotective muscle spasm manifesting as neckstiffness. Kernig's sign represents similarphenomena involving the distal spinal cordand related nerves. A manoeuvre that stretchesthe neural elements of the whole length of thespinal canal simultaneously will be a moresensitive test for meningeal irritation. Elicitingneck stiffness while the patient sits up withknees extended achieves this. This method is *more sensitive, specific, and amenable toobjective assessment.

(Arch Dis Child 1993;68:21548)

Beyond early infancy, neck stiffness is the mostimportant sign of meningitis, and meningealirritation is seldom suspected in its absence.' Inpatients with meningitis, neck stiffness wasobserved to become more marked in the sittingposture. This study evaluates neck stiffness indifferent postures with an aim to delineating theideal method for eliciting neck stiffness and Figure 2 Postre 2 with the patient sitttries to formulate objective criteria for assessing of the couch with legs hanging doum, hipmeningeal irritation. flexed, and absent neck stiffness. The phthe same patient and was taken at the sa

Subjects and methodsNeck stiffness was assessed in postures 1, 2, and3 (figs 1, 2, and 3 respectively). Neck stiffness isproduced by spasm of the extensors of neck.The degree of passive neck flexion possible(angle ABC, fig 4) will be a measure of themuscle spasm and hence that of neck stiffness

Department ofPaediatrics,Medical CoilegeHospital,Round East,Tnchur-1,Kerala,India 680001 Figure 3 Posture 3 with the patien sinJohny Vincent knees extnded and neck stiffness clearlyKurian Thomas that the angle between the plane of the rDepartment of \ >~and vertical is less than 450, and that thPaediatrics, support on the knees and back to avoid I

Benziger Hospital, posture. Note also the tense muscles on t)Quilon,.- Rotate the picture clockwJise through 91Kerala, right side down. Viewming the figure in thIndia w ~ 7 ~- imagine the patient's back resting on a hOmana Mathew Now, one can easily visualise his neck a

Figure I Posture I with the patient supine and mild remaining in positions equivalent to elicitiCorrespondence to: neck stiffness. The photograph is of a patient recovering and Kernig's sign simultaneously in a sulDr Vincent. from pyogenic meningitis and was taken on the fifth The photograph shows the same patient cAccepted 2 September 1992 day of treatment. at the same time.

ting on the edgeps and kneeswtograph showsame time.

ting up withevident. Notenape of neckke patient needsthe tripodAe neck.r to bring itshis position,orizontal couch.and lower limbsing neck stiffnesspine patient.and was taken

215

on July 3, 2020 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.68.2.215 on 1 February 1993. D

ownloaded from

Vincent, Thomas, Mathew

Angle ABC in postures 2 and 3

A

Angle ABCin posture 1

Figure 4 The angle ABC represents the degree of neckflexion. The line BC running parallel to the nape ofneck connects the vertebra prominence and the externaloccipital protuberance. Line AB passing through thevertebra prominence represents the horizontal plane inposture I and vertical plane in posture 2 and 3.

also. For accurate measurement of neck stiff-ness, one can note the degree of flexion achievedfor a given force applied to the head, with theneck fully relaxed. However, the application ofsophisticated gadgets to the head to measure theforce applied will frighten the child and jeo-pardise relaxation of the neck. Also, whileflexing the neck in the supine posture, theweight of the head (which cannot be measuredaccurately) will have to be overcome; in theerect posture, this weight acts to favour neckflexion. Thus it is not practicable to measureneck stiffness by purely objective methods.

In patients with meningitis, an initial variablerange of neck flexion will be observed to be freeof muscle spasm. This is followed by a secondphase characterised by a sharp and progressiveincrease in resistance to flexion. If forcibleflexion is continued, the third phase is reached.This is characterised by increasing musclespasm and pain and will be indicated by thechild's facial expression. Also, to avoid pain, thechild actively tries to extend the neck, and oftenovercomes the flexion already attained. Withincreasing neck stiffness, the muscle spasmcharacteristic of phase 2 begins earlier, so thatphase 1 gets progressively shortened. AngleABC was measured in patients and controlswhile keeping the neck passively flexed up tothe beginning of phase 2, using an instrumentsimilar to dividers with thin, rigid, straight, andblunt tipped arms. Its joint was placed on thevertebra prominence and one of its arms kept onthe nape of the neck, parallel to the imaginaryline BC (fig 4). The other arm was adjusted tooverlie the imaginary line AB, by keeping ithorizontal for posture 1, and vertical for posture 2and 3. The accuracy of alignment of this arm

was checked visually by comparing it with theplane of the couch for horizontal plane, andwith a freely hanging straight wire suspendednear by for the vertical plane. The instrumentwas then transferred on to a protractor to readoff the angle between the arms. While measur-

ing the angle in posture 1, the chest was notallowed to get lifted, and in postures 2 and 3,the tip of acromion was brought to remain verti-cally above the highest point of iliac crest. For

uniformity all the measurements were madebyJV.

Children aged 2 to 12 years with symptomssuggestive of meningitis, and neck stiffnessevident in posture 3, with or without neckstiffness in the other postures as per thesubjective assessment of JV, were subjected tolumbar puncture. Those with pleocytosis and/orneutrophils detected in the cerebrospinal fluidwere taken up for the study. Thus cases ofpyogenic as well as aseptic meningitis wereincluded. Children offering insufficient co-operation for eliciting neck stiffness, thosehaving paralysis of limbs, decerebrate or decor-ticate posturing, or taking more than two weeksto be cured were excluded. Seventy five caseswere studied. There were 22, 34, and 19 cases inthe 2 to 5, 5 to 8, and 8 to 12 year age groups,respectively. Angle ABC was measured simul-taneously in postures 1, 2, and 3 at the time ofadmission to hospital, and on subsequent days,whenever a change was appreciable in thedegree of neck stiffness.As a control, 900 children attending the

outpatient department were studied for thenormal free range of neck flexion possible inposture 3. All were examined while sitting on afirm couch, without tight clothing. They formedtwo groups with 450 cases in each, and eachgroup had 150 cases in the 2 to 5, 5 to 8, and 8 to12 year age groups. The first group (afebrilegroup) included patients who were unlikely tohave any restriction of neck flexion because ofdisease; for this purpose, children with feverduring the previous three days (a history ordocumented axillary temperature of more than37-2°C), central nervous system disease, evidentdeformities of spine, and painful conditions inrelation to the neck, trunk, or lower limbs wereexcluded. The second group (febrile group)included children with fever during the preced-ing 24 hours due to any cause, with or withoutother problems, but without neck stiffness inposture 3, as judged by the subjective assessmentof one of the authors (JV).

ResultsNeck stiffness was noted to be more marked inposture 3 than in 1 and 2. While sitting up fromsupine posture, patients with marked neckstiffness avoided kyphosis of the back, and inextreme cases kept the spine lordotically curved.They used hands on the bed for the tripodsupporting posture, and could not sit erectunless the knees were drawn up (Amoss' sign2).While measuring neck flexion in posture 3,these patients were permitted just sufficientflexion of hips and knees, so that they could situp to bring the tip of acromion vertically abovethe highest point of the iliac crest.During recovery neck stiffness disappeared

first in posture 1 and 2, but it persisted inposture 3 for another 1-2 days (or more)depending on the speed of recovery. Neckflexion measured simultaneously in the threepostures on 178 occasions while the patients hadvarying degrees of neck stiffness ranged between170 and 60° in posture 1 with a mean (SD) 42 56(10.67)0, 240 and 760 in posture 2 with a mean

216

on July 3, 2020 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.68.2.215 on 1 February 1993. D

ownloaded from

An improved clinical methodfor detecting meningeal imrtation

Figure Thi's photograph shows the same patient one

week after stopping treatment for meningitis. As a

normal child, his angle of neck flexion in posture 3 is

well above 45'. Note the kyphosis of the back; compare

this figure with figure 3.

(SD) 61-07 (9.80)0, and 50 and 500 in posture 3

with a mean (SD) 31-35 (jj.55)0. The difference

between the angles of neck flexion in postures

and 3 ranged between 40 and 220, and this

difference was highly significant (p<0-001,t test).

Once neck stiffness was fully relieved, angleABC in posture 3 became more than that in

posture 1; this gave the impression that a

greater degree of neck flexion would be possiblein posture 3 in normals. This is false, and is due

to the kyphosis generally noticeable when a

patient free of meningeal irritation sits up (fig

5). This kyphosis causes the upper thoracic and

consequently the cervical spine also to assume a

forward inclination. This inclination, togetherwith neck flexion, accounts for the exaggeratedangle ABC in posture 3 in cases without

meningeal irritation. In posture 1, as the back

rests on the cot, there is little kyphosis and theangle is due to neck flexion alone, and is hence

less than that in posture 3.

Neck flexion was most free in posture 2. The

difference between the angles of neck flexion inpostures and 2 ranged between64 and 322, andthis was also highly significant (p<0001 0t test).

Among the afebrile control children, neck

flexion in posture 3 varied between 500 and 1040

with a mean (SD)78n81 (7.88)0. Neck flexion infebrile controls ranged between 48,and 970 witha mean (SD)74k27 (9.66)0h

Discussion

The genesis of meningeal signs is best explainedon the basis of mechanical factors.'u The spinalcanal being posterior to the vertebral bodies, the

former lengthens, and its contents get stretchedwhen the spine is flexed. As the brain stem is

relatively immobile the cord gets pulled upwardswhen stretched by neck flexion. The resultant

tension is transmitted along the cauda equina to

the femoral and sciatic nerves. The femoral

nerve passes anterior to the hip. joint and the

sciatic posteriorly. Hence, the sciatic gets

stretched by hip flexion, and femoral by exten-

sion beyond the neutral position of the thigh.The opposite movements relax them. As thedownward continuation of the sciatic is along

the back of the knee it gets stretched by exten-sion and relaxed by flexion at the knee.However, the femoral nerve is little disturbedby movements at the knee as its downwardcourse is along the medial aspect of that joint.Both the nerves are maximally relaxed by keep-ing the hips and knees flexed (posture 2 and thetripod posture with knees drawn up). Here,though hip flexion stretches the sciatic, it getsrelaxed by flexion at knee. In the neutral posi-tion of the lower limbs (posture 1 and the stand-ing posture) the lower spinal roots also remainrelaxed. Flexion of the hip with extension of theknee (straight leg raising and Kernig's sign, andposture 3) stretches the sciatic and lower spinalroots maximally, though the femoral remainsrelaxed. Manoeuvres that stretch the inflamedneural elements and meninges of the spinalcanal induce pain and protective muscle spasm(for example, neck stiffness consequent topassive neck flexion) and lead to posturesdesigned to minimise tension on the inflamedstructures2 4 (for example, on attempting to situp, the patient adopts the tripod supportingposture with back and neck extended, and hipsand knees flexed). These phenomena character-ise all meningeal signs.The augmentation of neck stiffness occurring

in posture 3 can be explained as follows: in thisposture the lower limbs, which remain flexed athips and extended at knees, are in a postureequivalent to bilateral elicitation of Lasegue's(straight leg raising) or Kernig's signs. Thiscauses the distal spinal cord and related nervesto remain stretched and tethered down. Sittingwith a kyphosis stretches the cord at its middle.If the neck also is flexed now, the cord getspulled upwards also. This simultaneous upwardand downward pull on the cord causes moresevere pain and muscle spasm than if the pullwere from one end only, in which case the otherend could allow for some relaxation. Brudzinski'sneck sign substantiates this integration betweenmovements of neck and lower limbs in meningi-tis. Here, neck flexion causes flexion of hips andknees. The latter is a compensatory movementof the lower limbs to relax the cord stretched byneck flexion.

Testing for neck stiffness in posture 3 isequivalent to eliciting Kernig's sign and neckstiffness simultaneously-with one augmentingthe other. Eliciting neck stiffness in other pos-tures pulls the cord upwards only, leaving itslower end free. Textbooks of clinical medicine,paediatrics, and neurology either advise neckstiffness to be elicited in the supine posture,' -8or do not specify the posture.2 9 Opinion to thecontrary, that neck stiffness is better elicited inthe sitting posture, has been expressed byWehrle'0 and by Illingworth" independentlyand is based on personal observations (personalcommunications). The present report is inagreement with this view.As a test of neck stiffness it has been

suggested that the ability to approximate thechin to the chest is to be looked for. ' 2 7 9 This,however, would not be a satisfactory criterionfor assessing neck stiffness. Because of theirshort neck, children can approximate the chinto the chest by tilting the face downwards

....:s

:R

217

on July 3, 2020 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.68.2.215 on 1 February 1993. D

ownloaded from

Vincent, Thomas, Mathew

(nodding movement). This movement involvesflexion at the atlanto-occipital joint only, anddoes not cause a true flexion of the whole lengthof neck. Consequently, significant stretching ofthe cord does not occur and it can be carried outdespite mild meningeal irritation. Only trueneck flexion-and not the nodding move-ment-alters angle ABC. Hence, this angle,which reflects true neck flexion and the resultantstretching of the cord, is a better measure ofneck stiffness than the position of chin inrelation to the chest.The angle of neck flexion as obtained in

posture 3 is modified by, and is the aggregateeffect of flexion of neck, flexion of trunk, andflexion of hips with extension of knees. All thesemovements together stretch the neural elementsof the whole length of the spinal canal. Hence,this angle integrates all the important signs ofmeningeal irritation (neck stiffness, Kernig'ssign, and Amoss' sign) and is a composite,single, and more sensitive index of meningealirritation. It also forms a criterion suitable forobjective assessment of meningeal irritation atthe bedside. This method also obviates the needto elicit Amoss', Brudzinski's, and Kernig'ssigns separately.

For these reasons, neck stiffness due tomeningitis, meningism, and disease involvingthe neural elements of the craniospinal axisshould become more marked in posture 3, andneck stiffness due to local painful conditions inthe neck would not show this postural variation.Thus this method scores advantageously in thematter of specificity also.Though one of the authors (JV) could

subjectively appreciate neck stiffness on a fewoccasions in children who could flex the neckbeyond 450 and up to 500 in posture 3, forpractical bedside purposes meningeal irritationcan be assumed to be absent if a child can siterect with knees extended and then allowpassive neck flexion to the extent that the napeof neck forms an angle of more than 450 withvertical. The presence of pain and/or musclespasm within 450 of flexion is indicative ofmeningeal irritation. All control children couldachieve neck flexion beyond this.

Being more sensitive this method helps todetect pyogenic meningitis earlier. Detection ofaspectic meningitis with minimal meningealirritation is also made possible. Naturally, morecases of meningism will also be picked up.However, giving due consideration to all clinicalaspects in evaluating these cases with mild neckstiffness, and observing the evolution of theillness over the ensuing 12-24 hours will help toavoid lumbar puncture on many occasions. Inthis study, neck stiffness that was evident onlyin posture 3 led to the suspicion and subsequentconfirmation of the diagnosis of meningitis in 18patients. As this method combines all theimportant signs of meningeal irritation into oneobjectively assessable and more sensitive test,the absence of neck stiffness in posture 3 will bestrong evidence on firm grounds to excludemeningitis.

Thanks are due to Professor Paul F Wehrle, director ofpediatrics, Los Angeles County, USC Medical Center, California,and the late Professor Ronald S Illingworth, emeritus professorof paediatrics, University of Sheffield, for kind replies to queriesand for permitting us to use the contents of their letters forreference as personal communications. Thanks are also due toMr P I Narayanan, associate professor of biostatistics, MedicalCollege, Calicut, India, for the kind help offered for statisticalanalysis.

I Simpson JF. Meningeal signs and symptoms. In: Vinken PJ,Bruyn GW, eds. Handbook of clinical neurology. Vol 1.Amsterdam: North-Holland Publishing Company, 1%9:536-49.

2 DeJong RN. The neurologic examination. 4th Ed. Philadelphia:Harper and Row, 1979.

3 Wartenberg R. The signs of Brudzinski and Kernig. J Pediatr1950;37:679-84.

4 O'Connell JEA. The clinical signs of meningeal irritation.Brain 1946;69:9-21.

5 Juel-Jensen BE, Phuapradit P, Warrell DA. Bacterialmeningitis. In: Weatherall DJ, Ledingham JGG, WarrellDA, eds. Oxford textbook of medicine. Vol 2. 2nd Ed.Oxford: Oxford University Press, 1987;21:129-41.

6 Bannister R. Brain's clinical neurology. 6th Ed. Oxford:Oxford University Press, 1985.

7 Munro J, Edwards C. Macleod's clinical examination. 8th Ed.Edinburgh: Churchill Livingstone, 1990.

8 Forfar JO. Acute bacterial meningitis. In: Forfar JO,Arneil GC, eds. Textbook of paediatrics. Vol 2. 3rd Ed.Edinburgh: Churchill Livingstone, 1984:1381-6.

9 Swash M. Hutchison's clinical methods. 19th Ed. London:Bailliere Tindall, 1989.

10 Wehrle PF, Mathies AW Jr. Meningitis. In: Wehrle PF,Top FH, eds. Communicable and infectious diseases. 9th Ed.St Louis: C V Mosby Company, 1981:416-32.

11 Illingworth RS. Common symptoms of disease in children.8th Ed. Oxford: Blackwell Scientific Publications, 1984.

218

on July 3, 2020 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.68.2.215 on 1 February 1993. D

ownloaded from