3Accid Emerg Med 1997;14:387-391

Post-traumatic stress disorder

Suzanne Mason, Alison Rowlands

Post-traumatic stress disorder is a psychologi-cal disorder which develops following exposureto trauma, and is diagnosed only after fourweeks have elapsed. The characteristic featuresare persistent re-experiencing of the event,avoidance of stimuli associated with thetrauma, and symptoms of increased arousal. Itsprevalence in the general population is around1%, and risk factors for developing it includepre-existing psychological morbidity.

Very little is known about the psychologicalconsequences of accidents. Following injury,there is a whole range of physical, social, andpsychological sequelae that interrelate andhave a direct bearing on recovery. Somepsychological sequelae are well recognised,such as anxiety, depression, panic disorders,specific phobias, and post-traumatic stress dis-order. In this article we shall for the most partfocus on post-traumatic stress disorder and itsrelevance to accident and emergency (A&E)medicine. However, some of the other disor-ders mentioned (for example, travel anxiety)may be more common to our practice and needto be acknowledged.The rates of post-traumatic stress disorder

following incidents related to sport, occupa-tions, and road accidents varies according todifferent studies from 1% to 46%, highlightingthe need for further research. There may besome predisposing factors that influence thedevelopment ofpsychological problems follow-ing injuries.

Department ofAccident andEmergency Medicine,Northern GeneralHospital, HerriesRoad, Sheffield S57AU, UKS MasonA Rowlands

Correspondence to:Ms Suzanne Mason,Research Fellow in Accidentand Emergency Medicine.

Accepted for publication1 April 1997

Development ofpost-traumatic stressdisorder as a diagnostic entityIt was not until the late 19th and early 20thcenturies that health care professionals beganstudying the human reactions to extremestress. In 1882 John Erichsen described thesymptoms of post-traumatic stress disorder inpatients involved in rail accidents, labellingthese as "railway spine". This condition waslater used by litigants to claim compensationfor symptoms which appeared to have no

physical basis.Large numbers of casualties occurring dur-

ing the two world wars and from industrialaccidents prompted professionals to investigateand explain the extreme emotional distur-bances sometimes seen. Traumatic stress reac-

tions were described in various forms, mostcommonly "shell shock" during the first worldwar, when some men were observed to suffercatastrophic reactions following exposure toexplosions.More recently there has been a surge of

interest in post-traumatic stress, particularlyfollowing studies on Vietnam war veterans inthe USA and Holocaust survivors after the sec-

ond world war, leading to its formal descrip-

tion in 1980 by the American PsychiatricAssociation, and publication in the thirdedition of the Diagnostic and statistical manual ofmental disorders (DSM III).' Since then the cri-teria have been revised in the updated versionsof the manual as DSM III-R in 1987 and DSMIV in 1994 and is also now classified in theInternational classification of mental and behav-ioural diseases, 10th edition.2

Definition ofpost-traumatic stressdisorderPost-traumatic stress disorder is defined as thedevelopment of a characteristic symptom com-plex in certain individuals consisting of persist-ent re-experiencing of an event, avoidance ofstimuli associated with the event, and symp-toms of increased arousal. The criteria fordiagnosis are shown in table 1 as defined byDSM IV.

Importance ofthe stressorAn event which may precipitate post-traumaticstress disorder is described as a traumatic stres-sor in which the individual experiences intensefear, helplessness, or horror. In post-traumaticstress disorder, the stressor may involve actualor threatened death or serious injury, thewitnessing of death, injury, or threat to thephysical integrity of another person, or learn-ing about the death, serious harm, or threat ofdeath or injury to a family member or closeassociate. Many different types of stressor maylead to the development of symptoms, forexample rape, road traffic accidents, torture,natural disaster, or physical assault.

Other criteria involved in diagnosisTo make a diagnosis of post-traumatic stressdisorder, all the criteria in table 1 must be ful-filled, which means that the diagnosis is madeon an all or nothing basis, rather than on a slid-ing scale of symptoms. However, individualsmay present with some of the symptoms insmaller clusters, thereby not fulfilling all thecriteria for a diagnosis of post-traumatic stressdisorder, but still experiencing some psycho-logical distress.

The controversyThere is controversy among some profession-als about whether post-traumatic stress disor-der exists at all. It is a relatively new diagnosisand some of the criteria have been challenged,for instance Feinstein and Dolan' have ques-tioned the importance of the stressor causingsymptoms of intense fear, helplessness, or hor-ror in the development of symptoms. This is anarea where more investigation is needed toconfirm or refute such a question.

387

on March 21, 2020 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.6.387 on 1 Novem

ber 1997. Dow

nloaded from

Mason, Rowlands

Table 1 Diagnostic criteria for post-traumatic stress disorder

The person has been exposed to a traumatic event in which both of the following were present:The person experienced, witnessed, or was confronted with an event or events that involved actual or threatened death or serious injury, or a threat to thephysical integrity of self or others,The person's response involved intense fear, helplessness, or horror.The traumatic event is persistently re-experienced in one (or more) of the following ways:

Recurrent and intrusive distressing recollections of the event, including images, thoughts, or perceptions.Recurrent distressing dreams of the event.Acting or feeling as if the traumatic event were recurring (includes a sense of reliving the experience, illusions, hallucinations, and dissociative flashbackepisodes, including those that occur on waking or when intoxicated.Intense psychological distress at exposure to internal or external cues that symbolise or resemble an aspect of the traumatic event.Physiological reactivity on exposure to internal or external cues that symbolise or resemble an aspect of the traumatic event.

Persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness (not present before the trauma), as indicated by three (ormore) of the following:Efforts to avoid thoughts, feelings, or conversations associated with the trauma.Efforts to avoid activities, places, or people that arouse recollections of the trauma.Inability to recall an important aspect of the trauma.Markedly diminished interest or participation in significant activities.Feeling of detachment or estrangement from others.Restricted range of affect, for example unable to have loving feelings.Sense of foreshortened future, for example does not expect to have a career, marriage, children, or a normal life span.

Persistent symptoms of increased arousal (not present before), as indicated by two (or more) of the following:Difficulty falling or staying asleep.Irritability or outbursts of anger.

Difficulty concentrating.Hypervigilance.Exaggerated startle response.

Duration of the disturbance is more than one month.The disturbance causes clinically significant distress or impairment in social, occupational, or other important areas of functioning.

The diagnosis has been embraced by thelegal profession in their pursuit of compensa-

tion claims, where malingering for financialgain has long been assumed to be a significantproblem and the condition "compensationneurosis" has been described. This involvesexaggeration of symptoms, a worse outcomethan in those not claiming compensation, andthe improvement of symptoms on settlement.Miller45 found that patients who were lessskilled or from more poorly educated back-grounds were more strongly representedamong compensation neurosis cases, and thatthose with less work satisfaction were more

likely to seek gain from compensation. Mayou,in a recent paper6following up 188 consecutiveroad accident victims, has suggested that use ofa term such as "compensation neurosis" israrely appropriate as the condition is far lesscommon than is often claimed in compensa-tion proceedings. In addition, Cornes,7 study-ing 609 accident victims after moderate or

severe injury, all making claims, found that71% returned to work before the claim was

settled, and that of those who did not, only 8%were thought to have compensation neurosis.Tarsh and Royston,5 following up 35 individu-als claiming compensation after accidents thatcaused severe disability, found a lack ofimprovement in the symptoms after theirclaims were settled, reinforcing the argumentthat victims rarely simulate symptoms forfinancial gain.

Predisposing factors for post-traumaticstress disorder in the general populationEpidemiological studies of the prevalence ofpost-traumatic stress disorder are rare anddepend on local understanding of the reactionsto trauma in clinical practice. Those performedto date9'-4 provide an insight into the disorderand reveal gaps in our knowledge, suggestingthat further studies are required to identifythose individuals most at risk. All the studies

found strong associations of post-traumaticstress disorder with other psychiatric disorderssuch as depression, anxiety disorders, and sub-stance abuse. They also found a slightly higherincidence among women9'-3 and among indi-viduals with family histories of psychiatricillness or early parental separation.9'-" Eventsmost likely to precipitate post-traumatic stressdisorder in the general population includedseeing someone seriously hurt or killed, physi-cal attack, and combat situations.

Post-traumatic stress disorder andaccidentsVery little is known about the psychologicalreactions of accident victims who mightpresent to the A&E department followinginjury. There have been some studies in recentyears attempting to assess the range and extentof psychological problems after accidents.However, these studies have been very limitedin some respects, and there is a great need forfurther research.

PREVALENCEData on prevalence of post-traumatic stressdisorder among accident victims are given intable 2.During 1994 50 181 people were killed or

seriously injured as a result of road traffic acci-dents.15 Many more sustained less significantphysical injuries, but their experiences can

cause sufficient psychological damage to have a

severely debilitating effect on their lifestyle.The sequelae of injury are often not only expe-rienced by the individual and their immediatefamily and friends, but by society as a whole inthe form of disruption to family structure,occupation, and increased dependence on statesupport. It is not only road accident victimswho suffer; individuals injured at work or playare equally susceptible. During the period1994 to 1995 a total of 109 738 men sustainedoccupational or industrial injuries, of which

A.(1)

(2)B.(1)(2)(3)

(4)(5)C.

(1)(2)(3)(4)(5)(6)(7)D(1)(2)(3)(4)(5)E.

F.

388

on March 21, 2020 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.6.387 on 1 Novem

ber 1997. Dow

nloaded from

Post-traumatic stress disorder

Table 2 Prevalence ofpost-traumatic stress disorder(PTSD) among accident victims in studies to date

Study Subjects Rates ofPTSD

Blanchard et al, 1994'7 50 RTA victims 46%Mayou et al, 1993" 188 RTAvictims 11%Feinstein and Dolan,

19913 45 Fractured leg 14.5%Malt, 198820 107 accidents 1%

RTA, road traffic accident.

176 were fatal.'6 Although with the advent ofimproved health and safety regulations therehas been a general decrease in mortality andmorbidity, these numbers still represent asignificant proportion of the working popula-tion.

In 1994 Blanchard et al'7 published theresults of work they had done with 50 roadtraffic accident victims who had sought medi-cal attention, and compared these individualsto 50 non-accident controls. He found that46% of the victims met the criteria forpost-traumatic stress disorder and a further20% had what he described as a "subsyndro-mal form." In contrast, Mayou et al'8 in 1992studied the psychological sequelae of roadaccidents among three representative groups(n = 188), car occupants, motorcycle riders,and whiplash injury victims. They followedthese patients up over a year and foundpost-traumatic stress disorder in around 11%of victims, with a further 18% experiencing anacute distress syndrome characterised by anxi-ety or depression.

Feinstein and Dolan followed up 48 accidentvictims admitted after lower limb fracture overa six month period, interviewing them on threeoccasions, and found that within days of theaccident over 60% of patients could be classedas "psychiatric cases," but this number haddecreased to 25% by six weeks, and to 21% bysix months. In addition, post-traumatic stressdisorder was noted in 14.5% of patients at sixmonths. The study also suggested that earlyreturn to work was beneficial, since thosepatients who had done so by six weeks aftertheir accident had less psychiatric morbidity;however, this distinction was not present at sixmonths."9Malt interviewed a group of 107 trauma

patients on admission and followed them upsix to nine months later. He found that 16.8%of the group had a non-organic psychiatric dis-order during the follow up period, but foundpost-traumatic stress disorder in only 1% ofpatients.'oRoca and Spence interviewed 43 burns

patients and followed them up over a fourmonth period for signs of psychologicaldistress. They found that 7.1 % met thepost-traumatic stress disorder criteria ataround one month, with 22.6% meeting themat four months. They commented that indi-vidual symptoms were more common thanpost-traumatic stress disorder itself.2'

COMORBIDITY IN ACCIDENT VICTIMSMayou et al in 1992, following his prospectivestudy of 188 road accident victims, observedthat there appeared to be an overlap between

emotional distress, post-traumatic stress disor-der, and phobic anxiety.'8 Blanchard and Hick-ling, in a study of 20 road accident victims,found an increased comorbidity of otherpsychiatric disorders among patients withpost-traumatic stress disorder. These includedmajor depressive illness, simple phobias, anddysthymic disorders (a form of chronic depres-sion).22

Interestingly, Blanchard and Hickling alsofound that 50% met the criteria for post-traumatic stress disorder and 60% met the cri-teria for a travel anxiety, indicating that notonly are post-traumatic stress disorder symp-toms often linked to other disorders, but thatproblems such as travel anxiety can be moreprevalent and possibly more disruptive thanpost-traumatic stress disorder symptoms.22Changes in driving behaviour as a result ofroad accidents are well documented. Travelanxiety constitutes an avoidance or reductionin driving or travel in a vehicle, associated withmarked subjective discomfort when such travelhas to be endured. In 1994, reporting on thesame consecutive group of patients as in the1992 study,'8 Mayou and Bryant commentedthat a return to driving by injured people canbe delayed for medical, financial, or legalreasons. They found that 63% of motorcyclistswere either not riding at all or had changed todriving a car. Several patients expressed a wishto drive a safer car and stated that they weremore cautious and alert to dangers aroundwhen driving. These driving concerns werepresent in 65% of drivers one year after theaccident. They did not correlate with severityof injury or premorbid psychopathology, butseemed to be concerned with initial "horrificmemories" of the accident. 3

PREDISPOSING FACTORSUp to now only a few studies have suggestedfactors that may predict those individuals whowill go on to develop problems with post-traumatic stress disorder after an accident.These factors can broadly be described asbeing either related to the event (for example,the nature of the accident, the type and rangeof injuries), or to characteristics specific to theindividual.

Event* The nature of the stressor and perceived

stressfulness of the event were not found tobe important as discriminating variables.'

* Those victims suffering more extensiveinjury were found to be at increased riskfrom developing post-traumatic stress disor-der after road traffic accidents.2425

* The development of post-traumatic stressdisorder did not correlate with the severityof the injury following a burn.26

* Patients who were unconscious after anaccident did not develop post-traumaticstress disorder,"8 though a single case studyreported post-traumatic stress disorder de-veloping in a patient who was unconsciousfollowing injury.27

389

on March 21, 2020 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.6.387 on 1 Novem

ber 1997. Dow

nloaded from

390 Mason, Rowlands

Individual characteristics* Symptoms of post-traumatic stress disorder

in the early stages after injury are likely topredict problems at a later stage.3 19

* Higher than average weekly alcohol intakebefore injury may predict problems at a laterstage.3 19

* A previous history of problem drinking doesnot correlate with an increase in psychiatricmorbidity following an accident.28

* Horrific intrusive memories for the eventimmediately following injury strongly pre-dict the development of post-traumaticstress disorder. 18

* High levels of emotional distress immedi-ately after injury correlate with the develop-ment of post-traumatic stress disorder in thelong term.26

* Low levels of perceived emotional supportcorrelate with the development of post-traumatic stress disorder in the long term.26

* Persistent anxiety and depression followingtrauma are much more likely to occur inpatients with high neuroticism scores (asmeasured by the Eysenck personality inven-tory) or a history of previous consultationfor psychological problems.'8From these findings, it can be seen that con-

flicting observations have arisen from the vari-ous studies done so far. This can only serve tohighlight the need for further investigation. Itdoes seem that the initial level of distress andmemory of the event has a significant bearingon the reaction of the individual in the longterm. There also appears to be confusion as tothe importance of alcohol intake in thedevelopment of psychological problems. Inter-estingly, Noyes found increased numbers ofalcoholics involved in road accidents, thenumber involved increasing with the severity ofinjury, and being the highest for fatalities.29Patients convicted of drink-driving tended tohave a history of social maladjustment andantisocial behaviour. It is felt that alcohol mayenhance personality traits that are responsiblefor the cause of the accident, rather than physi-ological impairment being the sole cause.30Less appears to be known about the effects ofdrug abuse on driving behaviour, although thisis receiving increased attention.

ConclusionPost-traumatic stress disorder became a diag-nostic entity in 1981, but since then very littlehas been observed about the range and extentof individual reactions to everyday trauma andhow the physical injury interrelates with socialand psychological factors. In this article wehave summarised some of the research done sofar. However, this seems to have raised as manyquestions as it has answers about the incidenceof post-traumatic stress disorder and why someindividuals fall victim to the disorder whileothers do not. More work is needed to identifythe predisposing factors for post-traumaticstress disorder and to target those individualswho appear to be at risk.

SummaryPost-traumatic stress disorder is an anxietydisorder that can develop following exposureto a traumatic event. The symptoms are of per-sistent re-experiencing of the event, avoidanceof situations associated with the event, andsymptoms of increased arousal. It was firstdescribed following studies on Viemnam warveterans and Holocaust survivors in 1981. Inthis article we have examined the impact of thisdisorder on accident victims attending A&Edepartments by reviewing the studies done sofar. The results suggest that the rates ofpost-traumatic stress disorder after injury varyenormously, from 1% to 46%, indicating verylittle is certain about how the physical effects ofinjury interrelate with the social and psycho-logical effects. Factors predisposing topost-traumatic stress disorder may be theinitial distressing memories of the event,previous high alcohol intake, and coexistingpsychiatric problems. However, some ofthe observations made in these studiesseem to conflict, suggesting further studies areneeded.

1 American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders, 4th ed. Washington DC: Ameri-can Psychiatric Association, 1994.

2 World Health Organisation. International classification ofmental and behavioural diseases: clinical description anddiagnostic guidelines, 10th ed. Geneva: WHO, 1992.

3 Feinstein A, Dolan R. Predictors of post-traumatic stressdisorder following physical trauma: an examination of thestressor criterion. Psychol Med 1991;21:85-91.

4 Miller H. Accident neurosis. BMJ 196 1;i:992-8.5 Miller H. Accident neurosis. BMJ 1961;i:919-25.6 Mayou R. Medico-legal aspects of road traffic accidents. J

Psychosom Res 1995;39:789-98.7 Cornes P. Return to work of road accident victims claiming

compensation for personal injury. Injury 1992;23:256-60.8 Tarsh MJ, Royston C. A follow-up study of accident neuro-

sis. Br J Psychiatry 1985; 146:18-25.9 Breslau N, Davis GC. Post traumatic stress disorder in an

urban population ofyoung adults. Risk factors for chronic-ity. Am J Psychiatry 1992; 149:671-5.

10 Breslau N, Davis GC, Andreski P. Traumatic events andpost traumatic stress disorder in an urban population ofyoung adults. Arch Gen Psychiatry 1991;48:216-22.

11 Helzer JE, Robins LN, McEvoy L. Post-traumatic stress dis-order in the general population. N Engl J Med 1987;317:1630-3.

12 Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB.Posttraumatic stress disorder in the national comorbiditysurvey. Arch Gen Psychiatry 1995;52:1048-60.

13 Davidson JRT, Hughes D, Blazer DG, George LK.Post-traumatic stress disorder in the community: an epide-miological survey. Psychol Med 1991 ;21:713-21.

14 Wolfe J, Keane TM. New perspectives in the assessment anddiagnosis of combat-related post traumatic stress disorder.In: Wilson JP, Raphael B, eds. International handbook oftraumatic stress syndromes, 1st ed. New York: Plenum Press,1993:157-64.

15 Department of Transport. Male casualties by road user typeand severity 1987-1994. In: Road accidents Great Britain.London: Department of Transport, 1996.

16 Health and Safety Executive Commission. Injuries to maleemployees 1989-1995. In: Health and safety statistics, 1994/1995. London: Health and Safety Executive Commission,1995.

17 Blanchard EB, Hickling EJ, Taylor AE, Loos WR, GerardiRJ. Psychological morbidity associated with motor vehicleaccidents. Behav Res Ther 1994;32:283-90.

18 Mayou R, Bryant B, Duthie R. Psychiatric consequences ofroad traffic accidents. BMJ 1993;307:647-51.

19 Feinstein A. A prospective study of victims of physicaltrauma. In: Wilson JP, Raphael B, eds. International hand-book of traumatic stress syndromes, 1st ed. New York: Ple-num Press, 1993:157-64.

20 Malt U. The long-term psychiatric consequences ofaccidental injury. A longitudinal study of 107 adults. Br JPsychiatry 1988;153:810-8.

21 Roca RP, Spence RJ, Munster M. Posttraumatic adaptationand distress among adult burn survivors. Am J Psychiatry1992; 149: 1234-8.

22 Hickling EJ, Blanchard EB. Post traumatic stress disorderand motor vehicle accidents. J Anxiety Disord 1992;6:285-91.

on March 21, 2020 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.6.387 on 1 Novem

ber 1997. Dow

nloaded from

Post-traumatic stress disorder 391

23 Mayou R, Bryant BM. Effect of road traffic accidents ontravel. Injury 1994;25:457-60.

24 Blanchard EB, Hickling EJ, Mitnick N, Taylor AE, LoosWR, Buckley TC. The impact of severity of physical injuryand perception of life threat in the development of posttraumatic stress disorder in motor vehicle accident victims.Behav Res Ther 1995;33:529-34.

25 Blanchard EB, Hickling EJ, Vollmer AJ, Loos WR, BuckleyTC, Jaccard J. Short term follow up of post traumatic stresssymptoms in motor vehicle accident victims. Behav ResTher 1995;33:369-77.

26 Perry S, Difede JA, Musgni G, Frances AJ, Jacobsberg L.Predictors of post traumatic stress disorder after burninjury. Am J Psychiatry 1992;149:931-5.

27 McMillan TM. Post-traumatic stress disorder and severehead injury. Br J Psychiatry 1991;159:431-3.

28 Mayou R, Bryant B. Alcohol and road traffic accidents.Alcohol Alcoholism 1995;30:1-3.

29 Noyes R. Motor vehicle accidents related to psychiatricimpairment. Psychosomatics 1985;26:569-82.

30 Waller JA. Impaired driving and alcoholism: personality orpharmacological effects? J Safety Res 1969; 1: 174-7.

EMERGENCY CASEBOOKBilateral traumatic anterior dislocation of the hip joint

Bilateral anterior dislocation of the hip is exceedingly rare, six cases having been reported. 1-5Only one of these cases resulted from a road traffic accident3: the authors considered thatthe dislocation could have been prevented if a seat belt had been worn. We report a case ofa patient who sustained bilateral hipdislocation in a road traffic accident _despite wearing a seat belt. A 1.83 m, i85 kg boy of 15 years was the frontseat car passenger involved in a headon collision with a tree at 40 mph. Atthe time of impact the patient wasslouching in his seat; he slid under- _neath his seatbelt, and his knees hit 3the dashboard producing extremeabduction of his hips. On admissionboth hips were flexed to 800 exter-nally rotated, and abducted to 45'........Radiographs showed bilateral ante-rior dislocation of the hips (fig 1).Reduction was achieved under gen-eral anaesthesia, by hip flexion, trac- A

tion, and adduction with an assistant Fg 1 AP f hepels showngbilateralanterstabilising the pelvis. Skin traction dislocation of the hips.was applied for 12 days. Gentle mobi-lisation was allowed after 16 days. It is believed that the slouched sitting position meant thatthe patient slid underneath the seatbelt and that, because of his height and size, the lowerlimbs came into contact with the dashboard more easily. Sitting upright with the lower strapof the seat belt running around the anterior superior iliac spines could have prevented thisdislocation.

1 Aggarwal ND, Hardas S. Unreduced anterior dislocation of the hip. J Bone Joint Surg Br 1967;49-B/2:288-92.2 Gibbs A. Bilateral obturator dislocation of the hip joint. Injury 1980; 12:250-1.3 Zamon MH, Saltzman DI. Bilateral traumatic anterior dislocation of the hip. Clin Orthop 1981; 161:203-6.4 Tezcan R, Erginer R, Babacan M. Bilateral traumatic anterior dislocation of the hip: brief report. J Bone Joint Surg Br

1988;70B/1:148-9.5 Endo S, Hoshi S, Takayama H, Kan E. Traumatic bilateral obturator dislocation of the hip joint. Injury 199 i;22/3:232-3.

ROBERT J S SNEATH, N P H MORGAN, M PATTERSON Department of Orthopaedics, Princess RoyalHospital, Haywards Heath. Correspondence to: R J S Sneath, Bone Tumour Unit, The RoyalOrthopaedic Hospital, The Woodlands, Bristol Road South, Northfield, Birmingham B312AP.

on March 21, 2020 by guest. P

rotected by copyright.http://em

j.bmj.com

/J A

ccid Em

erg Med: first published as 10.1136/em

j.14.6.387 on 1 Novem

ber 1997. Dow

nloaded from