Journal of the Royal Society of Medicine Volume 84 January 1991 15

10 Pynoos RS, Nader K. Psychological first aid andtreatment approach for children exposed to communityviolence: research implications. J Traumatic Stress1988;1:243-67

11 Nader K, Pynoos RS, Fairbanks L, Frederick C.Childhood PTSD Reactions one year after a sniperattack. 1990 in press

12 Horowitz MJ, Wilner N, Alvarez W. Impact of eventscale: a measure of subjective stress. Psychosom Med1979;41:209-18

13 Yule W, Williams R. Post traumatic stress reactions inchildren. J Traumatic Stress 1990;3:279-95

14 Birleson P. The validity of depressive disorder in child-hood and the development of a self-rating scale: a researchreport. J Child Psychol Psychiatry 1981;22:73-88

15 Birleson P, Hudson I, Buchanan DG, Wolff S. Clinicalevaluation of a self-rating scale for depressive disorderin childhood (Depression Self-Rating Scale). J ChildPsychol Psychiatry 1987;28:43-60

16 Reynolds CR, Richmond BO. What I think and feel:A revized measure of children's manifest anxiety.J Abnorm Child Psychol 1978;6:271-80

17 Yule W, Udwin 0. Screening child survivors forpost-traumatic stress disorders: experiences fromthe "Jupiter" sinking. Br J Clin Psychol 1991; inpress

18 Yule W, Udwin 0, Murdoch K. The "Jupiter" sinking:effects on children's fears, depression and anxiety.J Child Psychol Psychiatry 1990 in press

19 Ollendick TH, Yule W, Ollier K. Fears in Britishchildren and their relationship to manifest anxiety anddepression. J Child Psychol Psychiatry 1990 in press

20 Tsui E. MSc Dissertation, University of London, 199021 Joseph SA, Brewin CR, Yule W, Williams RM. Causal

attributions and post-traumatic stress in children. 1990in press

22 Earls F, Smith E, Reich W, Jung KG. Investigatingpsychopathological consequences of a disaster inchildren: A pilot study incorporating a structureddiagnostic approach. J Am Acad Child AdolescPsychiatry 1988;27:90-5

23 American Psychiatric Association Diagnostic andstatistical manual ofmental disorders, 3rd edn, revised.Washington, DC: APA, 1987

Psychiatric response to theClapham rail crash

Paper read toSection ofPsychiatry,13 March 1990

T P Burns MD MRCPsych S C Hollins MBs FRCPsychDepartment of Psychiatry, St George's HospitalMedical School, Jenner Wing, Cranmer Terrace,Tooting, London SWl 7 ORE

Keywords: post-traumatic stress disorder; disaster plan; psychologicaldebriefing

SummaryThe psychiatric response to the Clapham rail crashis described. The psychiatric input was short term,dealing with the 38 inpatients and over 200 hospitalstaff involved in the response. The need to evolve acompact, responsive team structure is noted. Thevalue of a proactive approach and provision ofpsychological debriefing is defended. Incorporation ofcomponents of the psychiatric response into theHospital's major incident plan is reported.

IntroductionThree trains were in collision just south of ClaphamJunction immediately after 08.00 h on the morningof 12 December 1988. Two were packed commutertrains carrying over 1800 passengers, the majoritytravelling from Hampshire and Dorset to work inLondon. The third was an empty outward bound trainfrom Waterloo. Thirty-five people were killed (33 deadon site, one dead on arrival at St George's Hospitaland one died later of severe injuries) and 118 wereinjured and taken to the hospital Accident &Emergency Department of whom 38 were admitted(10 to the ICU).St George's Hospital, Tooting was designated as the

receiving hospital and its major incident plan initiated.The hospital is sited four miles from the accident andthe relevant medical and paramedical staff weredispatched to the accident site. Rescue operations

were prompt and access good, if cramped. A steepbank up from the cutting posed some difficulties inevacuation. Transfer to the hospital was rapid withthe 41 admissions and the 80 patients assessed anddischarged directly from A & E all registered by10.00 h with the peak admissions between 09.30 h and09.50 h. The A & E Department was only closed tolocal services for 2 hours.Psychiatry was not included in the hospital's major

incident plan so no response was mobilized directly.The first author (TB) holds the appointment of liaisonpsychiatrist, but was on leave and contacted at16.30 h. The psychiatric response was initiated, incollaboration with the hospital chaplaincy (alreadyextensively engaged in counselling and support work)that evening. The second author (SH) became involvedthe next morning when the two authors tookresponsibility for directing and coordinating thepsychiatric response. This paper will describe thecomponents of that response and critically appraisea number of its aspects. It will outline the psychiatricresponse which has since been incorporated into themajor accident plan.Because of the nature of this particular disaster

(those involved living in stable communities with goodlocal services distant from the accident) our servicedid not become extensively involved in counsellingPTSD sufferers or bereaved relatives. These needs arebeing met locally and this paper will confine itself tothe hospital response in the first 2 weeks.

The immediate responseIn the early evening (9 hours after the accident)A & Ewere still processing the consequences ofthe incidentwith many passengers and survivors still unaccountedfor. The final death toll was uncertain. Having metwith the A & E consultant and senior chaplain, it wasagreed that there should be a meeting the nextmorning to coordinate the psychiatric response.Discussions took place with the police at this stageabout arrangements for support of relatives whowould be coming to identify the dead. The chaplaincy

0141-0768/91/010015-05/$02.00/0© 1991The RoyalSociety ofMedicine

16 Journal of the Royal Society of Medicine Volume 84 January 1991

undertook to negotiate directly with social servicesto make suitable arrangements.The senior chaplain and liaison psychiatrist (TB)

undertook a joint round of the receiving and back-upwards. Although involvement of the medical socialworkers with the wards was already established,it was decided to offer daily proactive contactfrom psychiatry. The round enabled an accurateassessment of the extent of the problems. Of the38 patients admitted, 10 were in the intensive careunit (ICU) and very seriously ill. Over half ofthe remainder, although seriously shaken up withbruising and some fractures, were physically comfort-able. They anticipated early discharge with noscarring or long-term problems. The remainder hadmore serious injuries which they knew would requiremore prolonged treatment but were not in any seriousdanger. Only one inpatient on the ICU was outwardlyvery distraught at this stage, as were her family. Onthe whole, people appeared calm and relieved.The round also established the availability of

psychiatric and pastoral care in the minds of wardstaff and was welcomed by them. Our concerns thatthe visit might be seen as intrusive or disruptive forbusy ward staff were clearly unfounded. Contacttelephone numbers and names were left with staffandthey were asked to make sure that patients were notdischarged without a psychiatric assessment.Later that evening telephone calls were made to

colleagues who had had experience of previousdisasters (King's Cross, Piper Alpha) and advicesought. All three contacted gave essentially the sameadvice:Collect a team (establish control)Keep it smallCentralize information (clerical help).On the morning of the second day a steering group

was formed (see Figure 1). This consisted of the twoauthors, senior chaplain, A & E consultant andDepartment of Psychiatry biostatistician (both ofwhom had been collaborating on a study of post-traumatic stress disorder (PTSD) in A & E attenders).The meeting was concerned with identifying prioritiesand allocating tasks. A decision was made to keep thesteering group membership to those already presentat that meeting. Offers of help were flooding in, aswere differing suggestions for the composition of andtasks for, the steering group. It was felt that attempts

Core GroupPsychiatric senior registrars+ consultants (TB & SH)

Medical social workers

IInpatients

to assess and compare these alternative suggestionswere simply not practicable, given the need for urgentaction. The existing steering group had the advantageof all knowing each other and already being involved.We agreed to a daily meeting until it was decided to'stand down'.The Unit General Manager had already agreed

secretarial resources and the biostatistician organizedfor a part-time secretary in the Department to workfull-time for up to 3 months. He undertook to organizethe database on a computer which was providedpromptly by British Rail. In the course ofthe disasterresponse over 400 individuals were registered on thisdatabase including:

38 inpatients80 survivors assessed in A & E

200 hospital staff involved40 ambulance staff involved7 telephone requests for counselling

29 volunteer counsellors3 police and fire brigade contacts.

One psychiatrist (TB) organized for the local socialservices, who had set up a 24-h helpline, to liaise withthe psychiatric emergency clinic nurse who would actas a coordinator for both requests for, and offers ofhelp with, counselling. For volunteer counsellors werequired a commitment of at least one hour per weekfor a minimum of six weeks. The senior chaplain andthe A & E consultant accepted responsibility fordealing with any staff issues and proposed thatpsychological debriefing be offered to all involvedhospital staff. It was agreed that this should beorganized in groups and arranged for the 5th dayafter the incident. The A & E consultant ensuredcommunication with the outpatient group and theirGPs. The two psychiatrists undertook to organize a'core group' of psychiatrists to provide cover andtreatment for the inpatients. At this first meetingit was decided (in order to avoid intrusion) thatpsychiatric assessment would follow a specific request,either from ward staff or the medical social workerallocated to the patient. It was decided that the 'coregroup' should be pitched at senior registrar level andthat a minimum commitment to daily involvementwas required.On the 3rd day we found that a couple of patients

had already been discharged or transferred to local

Social Services helpline

Counsellors

Steering Group tc n

Liaison psychiatrist (TB)~E Eergency clinic nurseLConsultant psychiatrist (SH) Hsfa tf

Senior chaplain _ Hospital staffAccident & Emergency Consultant ] Outpatients

Biostatistician DatataentaDatabase (Secretary)

Figure 1. Team structure

Journal of the Royal Society of Medicine Volume 84 January 1991 17

hospitals without psychiatric assessment. The coregroup, therefore, made direct contact with theirallocated inpatients and a duty rota was establishedwith a hospital switchboard bleep.The pros and cons of using questionnaires to help

structure the psychiatric interview were discussed indepth. Some team members had used the question-naires, already part of an A & E study of traumasurvivors and found them helpful. It was eventuallydecided that they should be used with all patients.The questionnaires were the Horowitz Impact ofEvent Schedule', the Beck Depression Inventory2and the Spielberger State and Trait Anxiety Scales3.The steering group stood down on the 5th day,

although regular meetings continued between itsmembers and core group members over the ensuingweeks.

SurvivorsAll patients discharged directly from A & E took ahandwritten letter with them to their GPs. It rapidlybecame clear from contact with the inpatients thatwritten material, outlining the possible emotionalconsequences of a major disaster, was much appre-ciated. Most patients still felt 'numb' and recognizedthat an emotional reaction was likely later. They wereaware that their concentration was poor and thatverbal advice might be easily forgotten. A leaflet'Coping with a Major Personal Crisis . . .', evolvedafter the Herald of Free Enterprise disaster wasrapidly adapted for our circumstances and printed.This leaflet outlined the main features ofPTSD withadvice on what to expect and when and who to contactif help was needed. A & E arranged for all dischargedpatients and their GPs to receive these leaflets withan explanatory letter.Twenty-seven of the 38 inpatients (of whom two

died) were assessed by a psychiatrist. As well as aroutine psychiatric assessment, patients were askedto fill in the questionnaires which often promptedfurther discussion of their experiences. A smallnumber ofpatients neededregular contact for individualcounselling and some intensive family work wasprovided by the chaplain for one patient. For mostpatients exploration and reassurance seemed appro-priate and adequate. The opportunity to discuss whatcould be expected after such an incident wasparticularly valued. Very few patients needed ongoingcounselling from the hospital team after discharge.They were encouraged to contact local services forfurther help if need be. There were no cases of majorpsychiatric disorder precipitated in the inpatientgroup during their brief stay. We informed patientsthat we would keep some minimal contact with themover the ensuing year to monitor their progress.

Helpline and counsellingWandsworth Social Services manned a helpline 24-ha day for the first week. It soon became clear that,despite adequate publicity in local papers, the helplinewas not used and it was closed down. A total of eightreferrals were passed on from it to the psychiatricservice for counselling. Three of these were personnelinvolved in the rescue (two ambulance staff and onecouncil workman). Ofthe others, two used the helpline(effectively, it transpired) for personal crises unrelatedto the accident. The helpline was often used by thirdparties - family, employers etc. to ask advice abouthelping survivors.

The service was flooded with offers of help fromcounsellors in the first days. A list of 29 voluntarycounsellors was eventually compiled, but in the endonly five individuals were taken on for counsellingby them. Most of these individuals complainedof features of PTSD4 (such as tension, intrusivethoughts about the accident, poor sleep, concentrationetc). It is clear that most of the survivors of thisdisaster who received counselling did so elsewhere.

Staff debriefingOver 200 hospital staff were involved in the disasterresponse. Of these, 120 took part in nine debriefinggroups which were held on Friday 16 December,5 days after the accident. It was decided thatpsychological debriefing should be offered to all staffinvolved. We emphasized that this was an integralpart of the whole disaster proceedings and not to beconfused with counselling. Procedures for staffcounselling are well established to ensure confi-dentiality and are organized through the Staff/Student Health Department.The debriefing groups were scheduled for one hour

intervals throughout the day and they were thoroughlymixed both in terms of professional background andseniority. They varied in size from eight to 16members. Psychiatric staff with group therapyexperience were recruited to lead them. Only veryminimal preparation of the group conductors waspossible. Lacking any past experience ofleading suchgroups, we saw our task as encouraging open and frankdiscussion of the experience of the disaster and, inparticular, emotional responses. The leaders saw itas their task to ensure that vulnerable individualsdid not over-expose themselves and that the groupsended with some sense of positive closure.The groups proved to be moving and vivid experi-

ences for all involved. They often had distinct andvarying characters according to their composition, butthere were a number of regularly occurring themes(Table 1). There was general agreement that thedisaster experience had been emotionally draining,and qualitatively different from normal work. Manycommented that this stress was not related toexposure to particularly tragic sights (most of thefatalities occurred on site and were not brought to thehospital), but more to a sense of being unable toencapsulate and control the task. Although theemotions felt 'overwhelming', nobody reported fearthat the service itself would be overwhelmed or failto cope. It was a relief to share this experiencein the group. Dealing with requests for informationfrom distressed relatives was identified as probablythe most stressful part of the whole experience,

Table 1. Staff debriefing - themes

1 Surprise that the experience so powerful2 Telephone enquiries and information requests most

stressful3 Worry about unsupported juniors4 Anger at the media (intrusion; patchy cover)5 Abreaction that evening (alcohol)6 Response to VIPs7 Envy/deprivation8 Guilt over 'buzz': professional pride9 Difficulty settling back10 Poor concentration; accident proneness

18 Journal of the Royal Society of Medicine Volume 84 January 1991

particularly as the information requested was rarelyavailable. Clerks and receptionists, not normally seenas having high-stress roles, were exposed to intensepressure without obvious available support.Many experienced staff were concerned that they

may have inadequately supported their juniors andtrainees. There was anger about the media - both attheir intrusiveness (complained of by patients) andalso of their failure to acknowledge the contributionof staff other than doctors and nurses. A number ofstaff talked of how they found much of their pent-upemotions spilling out when they were off-duty athome, having taken a drink to wind down.There was also a great need to talk about the

experience of suddenly being in the limelight. VIPvisits had infuriated some, but made others feelappreciated. Staff removed from the ward to deal withthe sudden demand in A & E, felt guilty on returningto their colleagues who had had to stay behind tocover for them, and thereby miss the 'glamorous'experience. There was also some guilt over theexhilaration and professional satisfaction felt bymany - often referred to as a 'buzz'. Many foundit difficult settling back into routine ward workafter the incident. Interestingly, a number of staffspontaneously reported poor concentration andaccident-proneness in the ensuing days.We asked staff to indicate anonymously ifthey felt

the groups helped and overwhelmingly they did with14 staff requesting a follow-up group. Only twoindicated a wish for one-to-one counselling. The groupleaders considered in review that the groups were tooshort, and would have benefited by being morehomogeneous in their membership and structured intheir approach. For debriefing groups subsequentlyrun for the London Ambulance Service, the durationwas extended to 90 minutes and the approach outlinedby Raphael5 was adopted. This work will be reportedelsewhere.

Major Incident PlanThe involvement of the psychiatric team in theresponse to the Clapham train crash has led to therecognition that it has a major role to play in disasterplanning and the delivery of care in the event of afuture disaster. The psychiatric response is nowincorporated in the hospital's major incident plan withsix of a total of 52 action cards devoted to it.In the event ofa major incident the general hospital

switchboard will notify the psychiatric hospitalswitchboard to initiate the psychiatric disaster team.The duty or liaison consultant psychiatrist and bothon-call psychiatric senior registrars will then becontacted via the psychiatric hospital switchboard.One senior registrar (SR) will proceed to Accident &Emergency and the other to psychiatric outpatientswhere the Emergency Clinic (EC) nurse will set upthe specially designated direct line. The consultantwill take direct charge of both SRs. Both SRs,consultant and EC nurse will be automaticallyrelieved of their normal duties and meet daily fromthe initiation of the disaster response until the teamstands down.The SR and EC nurse will be responsible for

manning a helpline, organizing volunteer counsellorsand outpatient counselling. They will also functionas a psychiatric control point after day 1. The controlpoint for day 1 will be in A & E manned by theconsultant and SR, who will deal directly with A & E

and the wards. This SR will be responsible forensuring that the explanatory leaflets 'Coping witha Major Disaster' are made available.In addition to the six action cards designating

the team's immediate response, copies of a fullerdocument outlining the mechanisms of running asteering group, core team, selecting and organizingvolunteer counsellors, setting up a database andorganizing staff debriefing are kept at the psychiatrichospital switchboard and the emergency clinic,as are copies of the information leaflet and thequestionnaires used after the Clapham disaster.

ConclusionOur involvement in the Clapham train crash responsehas indicated clearly the need for, and appreciationof, a substantial organized psychiatric input. Nodisaster is typical and the nature of the response toany future one cannot be precisely predicted. Certainmajor aspects of other reported disasters (such as long-term counselling of survivors, support for bereavedrelatives, etc) were not issues for us. Indeed someaspects of this disaster, which required our inter-vention (eg the response to the school overlooking theaccident site and to the London Ambulance Service)are very specific and will be reported elsewhere. Wedid, however, feel able to draw some conclusions abouteffective disaster response from this experience.Keeping the team small and manageable was

crucial. Directly after a major disaster, despite theobjective efficiency of the response, there was apervasive sense of imminent chaos. Maintaining theteam's clear boundaries is a vital defence againstdisorganization and exhaustion. Our steering groupof five worked well together, but we all knew eachother previously. How important this latter factor wasis impossible to judge, but if the team were not veryfamiliar with each other, then the group might benefitby being even smaller.We would strongly endorse the advice we received

to set up a database and organize secretarial resource.We would also recommend that those responsible forsupervising it (the EC nurse and biostatistician in ourplan) also take responsibility for communications.Much time was occupied in the first 2 days receivingoffers of help and keeping colleagues informed. Suchactivities are best dealt with by a team member whois not responsible for front-line clinical activity. Thismember can communicate directly with the team ifthe message is sufficiently urgent, otherwise offerscan be reviewed at the steering group meeting eachmorning.Whilst a formal team structure proved most useful

in simplifying boundary issues it is also necessary forinternal matters. Decisions often had to be madequickly with imperfect knowledge and no one havingobviously relevant prior experience. Psychiatric teamshave long evolved practices based on working towardsconsensus. Where consensus did not immediatelyarise (eg whether to use questionnaires or not inassessments) it is necessary to have a simplehierarchy to identify quickly who can make thedecision. We also felt it was important that theteam leader (or two leaders in our case) safeguardthe primacy of their supervisory role. They mustavoid the temptation to take on too much clinicalwork too early. The action guidelines laid downfor the team are very general and the leader mustcontinuously ensure that resource is flexibly

Journal of the Royal Society of Medicine Volume 84 January 1991 19

redistributed from areas of overprovision to areasof need.The team needs to be visible and proactive. Being

on the spot both ensures that colleagues know whoto call on and also allows you to assess needs directlyand act on your own judgement. It also has theimmeasurable benefit of raising the team's facevalidity with your colleagues so that important,potentially threatening activities such as psychologicaldebriefing are more readily accepted. Staff debriefingneeds to be clearly understood as an integral part ofthe incident procedure for all staff and distinguishedclearly from counselling or psychotherapy. There istime for staff to make some preparation for thesegroups, and we have found Raphael's book WhenDisaster Strikes helpful in this area5. Attending tothe team's own needs can easily be forgotten. Thework is exciting, but exhausting. It can be difficultto switch off. The difficulty of assessing yourperformance can make it a very stressful time. Theteam needs to monitor its members to ensure thatthey are taking adequate time off and not losing asense of perspective. We have some simple rules, suchas insisting on taking a lunch break, reading thepaper daily and going home at a reasonable hour.These, allied with attending the daily steering/core

group meeting, should ensure that staff are protected.There may also be a role for establishing a supportgroup to explore and discharge tensions within theteam.Staff also need training. It is essential that all staff

who may become involved in the major incidentprocedure familiarize themselves with the plan andthe relevant locales. We have decided that thepsychiatric response should be reviewed and rehearsedeach year.

References1 Horowitz MJ. Stress response syndromes, 2nd edn.

Northvale NJ: Aaronson, 19862 Beck AT. Beck depression inventory. Philadelphia:

University of Pennsylvania, 19763 Spielberger CD, Gorusch RL, Lushene RE. The state-trait

anxiety inventory. Palo Alto, California: ConsultingPsychologists Press, 1970

4 American Psychiatric Association. Diagnostic andstatistical manual ofmental disorders, 3rd edn, revised.Washington, DC: American Psychiatric Association, 1987

5 Raphael B. When disaster strikes: how individuals andcommunities cope with catastrophe. New York: BasicBooks, 1986

(Accepted 5 September 1990)

Theoretical issues inresponses to disaster

Paper read toSection ofPsychiatry,13 March 1990

J Thompson CPsychol FBPsS Department ofPsychiatry, University College and Middlesex Schoolof Medicine, Mortimer Street, London WIN 8AA

Keywords: post-traumatic stress disorder; disasters; stress theories;threat and loss assessment

IntroductionThere are now numerous descriptions of humanbehaviour before, during, and after disasters, and inparticular of the problems subsequently reported bydisaster survivors'-4. Summaries of these reportedproblems have been collected in an attempt todelineate diagnostic categories, of which the mostused are DSM-IIIR and ICD-10. The question whichmust be answered is whether it is possible to gobeyond a detailed description of post-disaster reactions,and to provide an explanatory framework andthe beginnings of a model of human reactions toadversity.From a research point of view, disasters can be

considered as natural experiments, in which un-selected groups ofpeople are subjected to events whichare outside the range of ordinary human experience.As such they have implications for the understandingof the causes and development of stress relateddisorders. They may help us to understand more aboutthe nature of anxiety and depression, and why it isthat so many people are able to cope with adversitywithout major psychological ill affects. However, these

objectives can only be achieved if observations arelinked to explanations, and these explanations arethen tested against new data.The explanations which tend to be given about

stress reactions are often simple and non-theoretical.Arguments are drawn by analogy from the engineeringof solid structures, where stress is the force per unitarea exerted between contiguous bodies in parts of abody. This is commonly depicted as a large weightresting on a fragile and unstable structure. As a term,stress is often used in situations in which strainwould be more appropriate. 'Strain' describes the casewhere objects are stretched tightly and made tautand exercised to their greatest possible extent.Compressive stress is what a large weight applies toanything beneath it. Strain is the extent to which thesupporting structure is deformed and compressed bythe weight above. By plotting stress against strainit is possible to derive what is called the 'elasticmodulus' for any material, and this measure allowsarchitects and engineers to choose appropriatematerials to bear particular loads. Common to allthese views of psychological stress is the notion thatwhen a soft human body comes into contact with hardobjects, damage is likely to result. By analogy, eventsare seen as hard objects which weight a person down,and which require an elastic and stable structure towithstand.The term 'stress' is used so loosely that distinctions

are often not drawn between cause and effect. Thishas led to considerable confusion both in publicperceptions and in the stress literature. Properlyspeaking, stress is not an object in the world. It is thereaction of an organism to events in the world. Inorder to make this distinction clear it is usual todistinguish between stressors, which are objects and

0141-0768/91/010019-04/$02.00/0© 1991The RoyalSociety ofMedicine