the millennium scientiﬁc meeting, manchester, 9–11...

ABSTRACTS

The Millennium Scientific Meeting, Manchester, 9–11 October 2000(Hosted by the Faculty of Accident and Emergency Medicine)

EMRS/SSL International plcPrize Session

Use of plasma DNA analysis to deriveearly prediction rules for post-traumaticorgan failureT H RAINER, Y M D LO, L Y S CHAN, P K W LAM, E K

WONG, N M HJELM, R A COCKS

Accident and Emergency Medicine AcademicUnit and Department of Chemical Pathology,TheChinese University of Hong Kong and Prince ofWales Hospital, Hong Kong

Background—The discovery of eVective treat-ments preventing or mimising post-traumaticcomplications may be hampered by ourinability to target high risk patients early afterinjury. The purpose of this study was to derivea guideline for the prediction of post-traumatic organ failure using cell free(plasma) DNA and other predictors ofpost-traumatic complications.Methods—The research ethics committee ap-proved a prospective, observational studyinvestigating molecular responses of patientsto injury. Plasma DNA was measured using areal time, quantitative, polymerase chain reac-tion assay for the â globin gene in 84 patients(mean age 38 (SD 16) years; 83% male)triaged to an emergency department resusci-tation room within a median time of 60minutes (interquartile range 50, 90; range30–240) from injury. Other variables wereinjury severity scores, white cell count (WCC)and shock index (SI). Organ failure (OF) andmultiple organ dysfunction syndrome(MODS) occurred in 21 of 84 (25%) and 9 of84 (11%) cases respectively. After univariateand receiver operator curve analysis, data werefurther analysed using a classification andregression tree.Results—Within four hours of injury, OFcould be correctly predicted (predictive valuepositive, PV+) in 85.7% cases (n = 18 of 21;95%CI 62.6% to 96.2%) by: (1) cell freeDNA > 140 000 genome equivalents/ml, (2)injury severity score (ISS) > 27 and (3) WCC> 13.4 (×109/l). OF could be correctlyexcluded (predictive value negative, PV−) in95.2% cases (n = 60 of 63; 95%CI 85.8% to98.8%) by (1) cell free DNA < 141 000genome equivalents/ml, (2) ISS < 27 and (3)WCC < 13.4. Sensitivity and specificity were85.7% (95%CI, 62.6% to 96.2%) and 95.2%(95% CI, 85.8% to 98.8%) respectively.MODS could be correctly predicted (PV+) in87.5% cases (n = 7 of 8; 95%CI 46.7% to99.3%) by: (1) cell free DNA > 108 000copies/ml, (2) maximal abbreviated injuryscore (MAIS) > 3, and (3) SI > 0.72. MODScould be correctly excluded (PV−) in 97.3%cases (n = 71 of 73; 95%CI 89.6% to 99.5%)by (1) cell free DNA < 108 000 copies/ml, (2)MAIS < 3.0 and (3) SI < 0.72. Sensitivityand specificity were 77.8% (95%CI, 40.2% to96.1%) and 98.6% (95% CI, 91.5% to99.9%) respectively.Conclusion—Plasma DNA analysis allows thedevelopment of early accurate guidelines for

the prediction of post-traumatic OF andMODS. These guidelines now require pro-spective validation.

The eVects of in vivo haemodilution with0.9% sodium chloride and Gelofusine onwhole blood coagulation and plateletfunctionE V BRAZIL*, U SHAH**, M MACEY**, T J COATS**Academic Unit, Accident and Emergency De-partment, and **Department of Haematology,Royal London Hospital, Whitechapel, London,E1 1BB

Introduction—In vitro haemodilution withGelofusine reduces clot quality in contrastwith saline, which has a procoagulant eVect.These coagulation changes are poorly under-stood but may involve alterations in plateletfunction. The Sonoclot analyser allowsmeasurement of a number of parameters ofwhole blood coagulation. After platelet activa-tion CD62P (P selectin) expression is in-creased on the surface membrane of the plate-let. CD42a (GPIX) is also expressed on theplatelet surface and CD45 on leucocytes butnot platelets. Platelet-leucocyte aggregates(CD45-CD42a positive events) may be in-creased in thrombotic states. Monoclonalantibodies to these antigens can be conjugatedto fluorescent molecules and analysed by flowcytometry.Aim—The purpose of this study was to inves-tigate the eVect of in vivo haemodilution with0/9% sodium chloride and Gelofusine onwhole blood coagulation, platelet activationand platelet-leucocyte interaction.Methods—The study was performed as arandomised, controlled, crossover study. Eightadult volunteers received 1000 ml of eachsolution over 30 minutes on separate occa-sions with a one week washout period betweentests. Atraumatic blood sampling was per-formed from a free flowing upper limb veinbefore and immediately after fluid infusion.Fresh blood was used for Sonoclot analysis.Blood for platelet analysis was collected intosodium citrate containing vacutainers andanalysed within 10 minutes of collection.Results—Mean Sonoclot values and plateletmolecule expression pre/post solution are pre-sented in table 1.Conclusions—In vivo haemodilution withGelofusine significantly prolongs the time toreach maximum clot strength. Other wholeblood coagulation parameters are unaltered.Both platelet activation and platelet-leucocyte

interaction are impaired by Gelofusine in con-trast with saline, which promotes thesechanges.

A prospective randomised controlledtrial to investigate the clinical and costeVectiveness of emergency physiotherapySUSAN J BUTTRESS, KEVIN MACKWAY-JONES, JACKIE

OLDHAM

Accident and Emergency Department, Manches-ter Royal Infirmary, Oxford Road, Manchester,M13 9WLBackground—Musculoskeletal problems ac-count for a high proportion of patients attend-ing emergency departments. As many as onethird of patients have musculoskeletal prob-lems. The most common pathway is forpatients to be placed on an outpatient waitinglist with considerable delays. The result is thatacute physiotherapy is rarely provided. Analternative pathway is for emergency depart-ment physiotherapy delivering treatment ear-lier than would otherwise be possible. Earlierintervention may not only reduce the durationof acute symptoms and length of treatmentbut also aVect the long term outcome. Thereis evidence to suggest that this is the case.Objective—To investigate the clinical and costeVectiveness of an emergency physiotherapyservice.Design—Prospective randomised controlledtrial.Setting—An inner city teaching hospital.Subjects—Adult patients attending the emer-gency department with soft tissue injuries ofthe knee, ankle and neck that are assessed assuitable for physiotherapy by the emergencyphysician. Patients were randomised intoeither an emergency physiotherapy group or atraditional late physiotherapy group.Main outcome measures—Clinical outcome wasassessed using the short form-36, and theEuroQol questionnaires. The cost of eachpatient episode was determined from patientdata.Results—Each patient subgroup was ran-domised producing three separate trials.Analysis was performed using the diVerentdimensions of the SF-36. There were signifi-cant diVerences between the emergency groupand the traditional treatment group (p< 0.05)in the ankle and knee subgroups. There wasno significant diVerence in the neck subgroup.Conclusions—This study has demonstratedthat the early treatment of soft tissue injuriesof the knee and ankle is more eVective than

Table 1

Before saline After saline ChangeBeforecolloid After colloid Change

Peak (s) 1142 1281 139 1168 2200** 1032Rate 19.7 19.8 0.1 19.7 18 1.7ACT 120 97* −23 118 115 −3Maximum amplitude 75 76 1 73 75 2CD 62P 1.42 2.19 0.766 1.31 0.95 −0.363CD45-CD42a 4.55 7.12 2.57 5.37 3.46 −1.91

**p=0.009, *p=0.016, paired t test.

J Accid Emerg Med 2000;17:434–450434

www.jnlaem.com

on 14 July 2018 by guest. Protected by copyright.

http://emj.bm

j.com/

J Accid E

merg M

ed: first published as 10.1136/emj.17.6.434 on 1 N

ovember 2000. D

ownloaded from

http://emj.bmj.com/

treatment by the traditional method of adelayed service.

Identifying and managing risks in emer-gency medicineM THOMAS, K MACKWAY-JONES, N BOREHAM

EMERGE OYce, Department of EmergencyMedicine, Manchester Royal Infirmary, Man-chester M13 9WL

Objectives—To investigate the underlyingcauses of clinical risks in emergency depart-ments (EDs). To suggest appropriate counter-measures to these underlying causes.Design—Prospective observational study in-volving identification of critical incidents inEDs by staV incident reporting, by directobservation, and by daily review of patients’records. Derivation of causal trees for eachincident identified with classification of rootcauses, obtained from the causal trees, using arisk management tool known as MECCA(Medical Errors Complications and CausalAnalysis). Comparison of root causes betweendiVerent EDs by Kruskall-Wallis and Mann-Whitney tests. Production of countermeasureprofiles for each ED using a matrix derived forthis purpose.Setting—Four UK EDs, studied for sixmatched one week periods.Main outcome measures—Root cause profiles ofcritical incidents in each ED.Results—349 critical incidents were studied,giving 852 root causes. Some 3.6% of rootcauses were technical in nature, indicatingthat equipment inadequacies are not a com-mon cause of risks in EDs. The majority ofroot causes were human (42.1%) or organisa-tional (34.6%) in nature. The remainder ofroot causes involved patient related factors(14.8%) or were unclassifiable (4.8%). Sig-nificant diVerences (p<0.001) were detectedbetween EDs for three subcategories of rootcause. These were root causes relating to lackor inadequacy of protocols or guidelines in usewithin the EDs; those relating to collectivebehaviour; and those relating to organisationalfactors outside the departments (mainly at-tributable to bed shortages). The counter-measure profiles for each ED diVered, reflect-ing the diVerent underlying causal factors ofrisks in the EDs.Conclusion—Most critical incidents in EDs arerelated to human or organisational factors.MECCA analysis can detect significant diVer-ences in the causal factors of risk betweenEDs, and thus focus countermeasures appro-priately.

Can serum S-100B level predict anadverse outcome following head injury?W J K TOWNEND, M A PANI, J M GUY, D W YATES

Department of Emergency Medicine,Hope Hospi-tal, Salford

Objective—Patients with head injuries of anyapparent severity at presentation can experi-ence significant, long term morbidity. A serummarker of brain injury that provided anindication of potential complications would bea valuable addition to our diagnostic armoury.S-100B is a protein expressed almost exclu-sively in astroglial cells of brain tissue thatcrosses the blood brain barrier in measurablequantities following brain injury. The aim ofthis study is to evaluate serum S-100B level asa predictive test for adverse outcome followinghead injury of any apparent severity at presen-tation.Methods—Patients with head injury of anyseverity (GCS 3–15) attending the emergency

departments of two hospitals in Manchesterwere included in this prospective study. SerumS-100B levels were measured within six hoursof injury, and patient outcome assessed at onemonth using the extended Glasgow OutcomeScore (GOSE).Results—94 patients have been recruited todate, and 24 of these have now been followedup. A further 300 will be recruited in the nextfour months. Initial data analysis shows a sig-nificant correlation between serum S100Blevel and GOSE (Spearman’s ñ= −0.564; p=0.004). Analysis of data for the subgroup ofpatients attending with initial GCS 15 (n= 20)revealed a significant correlation also (Spear-man’s ñ= −0.450; p= 0.047).Conclusion—In a heterogeneous populationwith predominantly mild head injury wefound a significant correlation betweenS-100B level measured within six hours andneurological outcome at one month. S-100Blevel may have a role as a predictor of adverseoutcome in those patients who present with aninitial GCS of 15.

The eVect of the right and left lateralrecovery positions on vena cavaldiameterKRISTIAN BAILEY, LESLEY BETHUNE, KEN UZOKA,J STUART

No address suppliedBackground—The lateral recovery position isrecommended for patients with reducedairway protective reflexes. Anaesthetic prac-tice favours the left lateral recovery positionwhile standard resus texts do not state a pref-erence. Previous unpublished work (by JS)demonstrated a significant fall in strokedistance in the left lateral recovery positionbut not the right lateral recovery position. Thephysiological mechanism for this observationwas unclear.Objectives—To determine the eVects of the leftand right lateral recovery positions on thediameter of the inferior vena cava.Method—Thirty healthy volunteers had theirvena caval diameter measured (and cross sec-tional area calculated) by ultrasound imagingin three positions; supine, left lateral recoveryand right lateral recovery. The changes wereexpressed as percentage change from supine.Results—There was a mean fall in IVC crosssectional area in the left lateral position of48.9% (p<0.05) while the IVC cross sectionalarea increased by 62% (p<0.05) in the rightlateral recovery position. This eVect was dueto direct compression by the liver.Conclusion—This observation would explainthe previously observed fall in stroke distancein the left lateral recovery position in terms ofcaval compression by the liver. From a cardio-vascular perspective, the right lateral recoveryposition is superior to the left lateral position.

Oral presentations: Paediatric emergen-cies and others

Nasal diamorphine for acute paediatricpain: a multi-centre randomised control-led trialJASON KENDALL, BARNABY REEVES, VICTORIA

LATTER

Frenchay Hospital, Frenchay Park Road, Bristol,BS16 1LE

Objective—To evaluate the eYcacy and safetyof nasal diamorphine (ND) as an analgesicagent for use in children in the emergencydepartment (ED).

Methods—Prospective, randomised, multi-centre controlled trial (following Good Clini-cal Practice guidelines) comparing ND (0.1mg/kg) with intramuscular morphine (IMM)(0.2 mg/kg). Eight centres recruited children(aged 3 to 16 years) with clinically suspectedlimb fractures. Outcome measures includedtolerance to, and acceptability of, administra-tion of analgesic, pain scoring (at baseline, 5,10, 20 and 30 minutes after analgesic admin-istration), physiological measures (pulse, res-piratory rate, GCS and oxygen saturation atbaseline, 5, 10, 20 and 30 minutes), andadverse events.Results—410 children were recruited (IMM =203, ND = 207). Demographic and potentialconfounding factors were equally distributedbetween the two groups. Tolerance andacceptability measures were significantly bet-ter in the ND group (p < 0.001). Baseline painscores were the same in both groups. Painscores as reported by the child were signifi-cantly better at 5, 10 and 20 minutes in theND group (p = 0.04, p = 0.006, and p = 0.003respectively). There was no significant diVer-ence in pain scores at 30 minutes. Parentalreporting of pain scores showed a similar dis-tribution with similar statistical significance.There were no diVerences between the groupsin physiological variables measured, with theexception of oxygen saturation, where a statis-tically significant reduction in oxygen satura-tion was observed at 5, 10 and 20 minutes inthe ND group; this was, however, notclinically significant. There were no seriousadverse events; there was one moderateadverse event in the IMM group; there wereseveral minor adverse events in both groups.Conclusions—ND is an eVective, safe and welltolerated method of analgesia for children inacute pain in the ED.

Magnetic resonance imaging in the man-agement of paediatric scaphoid injuriesE SYMONDS, F HAIGH, K JOHNSON

The Birmingham Children’s Hospital, SteelhouseLane, Birmingham B4 6NH

Background—The scaphoid is the most com-monly fractured carpal bone. Plain radiographsare poor at detecting scaphoid fractures atinitial presentation. The consequences of amissed scaphoid fracture can be serious andincreasingly other imaging modalities are beingused to detect scaphoid fractures. We presentour work on the routine use of magneticresonance imaging (MRI) in the paediatricpopulation to detect scaphoid fractures early.Methods—All children presenting to the Bir-mingham Children’s Hospital emergency de-partment with a suspected scaphoid fracturehad an initial “scaphoid radiography” seriesand were immobilised appropriately. At earlyclinical review those patients in whom therewas clinical suspicion of a scaphoid injurywere referred for MRI. The MRI wasperformed within 10 days of the initial injury.The MRI was regarded as the definitive inves-tigation for the scaphoid injury and thosepatients who had a negative MRI weredischarged from care. Those patients with apositive MRI were treated appropriately.Results—114 children had scaphoid MRI, theyaged from 4.5 to 16 years (73 boys). All chil-dren tolerated the MRI scanner. Forty two ofthe examinations were positive for a scaphoidfracture (39%). Additionally, other carpalbone fractures, soft tissue injuries and gan-glion cysts were identified on the MRI andwere treated appropriately. There were 51negative examinations (48%). All these chil-

Abstracts 435

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

dren were discharged from care. None of thisnegative group has represented.Conclusions—MRI can be routinely used in abusy paediatric emergency department toexclude a scaphoid injury. The routine use ofMRI avoids the children having repeat plainfilm radiographs and confirms or refutes thediagnosis of a scaphoid fracture early. Thosechildren having a negative examination can bedischarged from care avoiding unnecessaryimmobilisation and treatment. The use ofMRI for scaphoid injuries reduces patients’attendance at hospital, improves diagnosticaccuracy and improves patient care.

Whiplash associated disorder inchildren—an emergency departmentbased studyR BOYD*, L DUANE**, R MASSEY***, D YATES****Department of Emergency Medicine, Hope Hos-pital, **Department of Accident and EmergencyMedicine,Booth Hall Childrens Hospital, ***De-partment of Accident and Emergency Medicine,Royal Liverpool Childrens Hospital, ****Aca-demic Department of Emergency Medicine, HopeHospital

Aims—To determine the incidence, severityand course of whiplash associated disorder(WAD) in children aged 4–16 years afterinvolvement as a passenger in a road traYcaccident (RTA).Methods—Prospective surveillance of all paedi-atric attendances to three UK urban emergencydepartments following RTAs for a six monthperiod. An initial structured telephone inter-view at day 2 after the RTA was followed byselective clinical review utilising the QuebecTask Force criteria for outcome assessment.Results—One hundred and one children wereidentified as having been involved in a RTA asa passenger. Forty three experienced symp-toms of a WAD (42.5%). Thirty six experi-enced a WAD of grade 1 with sevenexperiencing grade 2 severity. There was nosignificant age diVerence between the childrenwho did and did not develop WAD symptoms.Twenty five of the children developing WADwere female (58%). The mean duration ofsymptoms was 8.5 days (range 3–70, SD10.7). Sixty one children developed symptomswithin 24 hours (60%) with the remainderdeveloping symptoms by 48 hours.Conclusions—Despite the paucity of publishedevidence, WAD does occur in children. Theincidence of WAD in children is in fact similarto that found in adult studies. This studysuggests that the clinical course however ismore favourable in children than that inadults.Limitations of this study—This pilot work lackssuYcient numbers to enable meaningfulstatistical analysis of subgroups. Inherent biasexists relating to the selection bias of childrenpresenting to the emergency department.What does the paper oVer that is new to thefield?—This is the first prospective study of theincidence and clinical course of WAD in apaediatric population.

What is a normal systolic blood pressureand pulse rate in an injured child?PAUL DARK, MARALYN WOODFORD, ANDY VAIL,KEVIN MACKWAY-JONES, DAVID YATES

MRC TRAUMA GROUP (NWIRC), StopfordBuilding, Oxford Road, University of Manches-ter, Manchester M13 9PT

Objective—To determine the eVect of injuryon “normal” age related systolic blood pres-sures and pulse rates in children.

Design—Nine year prospective cohort study.Setting—Participating emergency depart-ments in the Trauma Audit and ResearchNetwork (TARN)—formerly UK MajorTrauma Outcome Study.Subjects—All injured children below 16 yearsof age and who were eligible for inclusion inthe TARN database.Main outcome measures—Age specific systolicblood pressure (mm Hg) and pulse rates (perminute) in injured children were comparedwith standard quoted values for restingchildren and with Advanced Paediatric LifeSupport “norms”. The Injury Severity Scorewas used to categorise by anatomical injuryseverity.Results—Advanced Paediatric Life Supportage specific “norms” of systolic blood pressureand pulse rate resembled standard quotedvalues for children at rest. Our cohort of over10 000 injured children had higher systolicblood pressures than would be expected fromthe standard quoted “norms”. These normsfor children at rest always appeared below the50th centile for injured children (range 7th to27th centile). This relative systolic hyper-tension in injured children appeared unrelatedto injury severity and age. Pulse rate norms forchildren at rest were always within theinterquartile range for injured children (range41st to 54th centile) and did not appear influ-enced by age or injury severity.Conclusion—Injured children have a relativesystolic hypertension on arrival in the emer-gency department, compared with children atrest. However, pulse rates in these two groupsare comparable. Age related systolic bloodpressures and pulse rates may be poor indica-tors of haemodynamic status of children afterinjury.

Occupational stress in consultants inaccident and emergency medicine: anational survey of levels of stress at workSUSAN ROBINSON*, RACHEL BURBECK**, SALLY

COOMBER**, CHRIS TODD****Emergency Medicine, Addenbrookes NHSTrust, **Occupational Medicine and ***Instituteof Public Health, University of Cambridge,Addenbrooke’s NHS Trust, Hills Road, Cam-bridge CB2 2QQ

Background—Occupational stress is a recog-nised problem in healthcare workers, anddoctors are considered to be at special risk.1 Anumber of surveys suggest that about a quarterof senior doctors are distressed or depressed.2 3

Among emergency physicians, available data,primarily from the US, suggest that theproblem of burnout and stress may exist to aconsiderable degree.4 However, to date littlework has been carried out in the UK.Objective—To measure the prevalence ofoccupational stress as measured by psycho-logical morbidity and depression among acci-dent and emergency (A&E) consultants in theUK.Method—Postal survey of practising BAEM/FAEM members incorporating the GeneralHealth Questionnaire-12 (GHQ-12) andSymptom Checklist-Depression Scale(SCL-D)Results—From a 78% response rate (371 of479), 45.6% of respondents had GHQ-12scores indicative of psychiatric disorder(GHQ-12 > 3), and 18.1% had an SCL-Dscore indicative of depression. Women hadsignificantly higher SCL-D scores than men(U = 6604, p < 0.009). Scores on both meas-ures rose with the numbers of hours worked,and there was a significant association be-

tween GHQ-12 score and hours worked forthose working full time (ñ = 0.126, p < 0.028,n= 318). Respondents were highly satisfiedwith their choice of A&E as a specialty. Theirmost highly rated stressors were: “lack of bedswithin the main hospital”, “being over-stretched at times” and “the eVect of hours ofwork on personal/family life”.Conclusions—Despite being highly satisfiedwith A&E as a specialty, large numbers ofA&E consultants seem to be suVering fromsymptoms indicative of psychological distress.Chief stressors seem to be factors outside ofthe specialty itself. The most highly rankedsuggestions for relieving stress were: the avail-ability of experienced middle grade doctors 24hours a day, having managers who understandthe issues within A&E, and resolving delays inpatient transfers out of A&E.

1 Stress at work—a guide for employers. HS(G) 116.London: HSE Books, 1995

2 Kapur N, Borrill C, Stride C. Psychological mor-bidity and job satisfaction in hospital consultantsand junior house oYcers: multicentre, crosssectional survey. BMJ 1998;317:511–12.

3 Blenkin H, Deary I, Sadler A, et al. Stress in NHSconsultants. BMJ 1995;310:534.

4 Doan-Wiggins L, Zun L, Cooper MA, et al. Prac-tice satisfaction, occupational stress, and attritionof emergency physicians. Acad Emerg Med 1995;6:556-63.

The scientific basis of the refined casemixmeasure for the specialty of accident andemergency medicine (HRG version 2.2)N F BRAYLEY, MEMBERS OF THE CASEMIX SUBCOM-MITTEE OF THE CLINICAL SERVICES COMMITTEE OF

BAEM AND THE CASEMIX OFFICE OF THE NHS

INFORMATION AUTHORITY

Emergency Department, Colchester GeneralHospital

The development of a robust casemix measurefor the specialty of accident and emergency(A&E) medicine is desirable to secure appro-priate funding for the resources necessary tomeet the varying demands for patient careprovided in diVerent departments from oneyear to the next.

This presentation explains the developmentof Healthcare Resource Group (HRG) ver-sion 2.2, the role of classification andregression trees (CART) analysis to producethe casemix measure nodes that give the newversion a reduction in variance (RIV) of 64%.Investigations and disposal are the key nodesidentified from a refined UK database of over100 000 episodes prepared in 1999 from fivediVerent types of A&E departments.

The role of the national Triage Scale as acasemix measure was considered but foundnot to be as robust as disposal.

The Australian casemix measure based onTriage provides a similar RIV but found not tobe reproduced with the English dataset.

From June 2000 HRG casemix returns willbe mandatory for each English A&E depart-ment. The possibilities for further refinementwork and practical applications for HRGs willbe discussed.

Trauma

Improving prediction of outcome frommajor traumaHIU NAM TONG, SALLY HOLLIS, ALAN WROTHFORD

Trauma Audit & Research Network, ClinicalSciences Building, Hope Hospital, Stott Lane,Salford M6 8HD

Objective—To determine the impact of addingpre-existing disease status to the TRISSformula.

436 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

Method—The study was based on TraumaAudit and Research Network data from1994–98 inclusive. Revised Trauma Score(RTS), Injury Severity Score (ISS), age andpre-existing disease (PMC) status were re-corded in 53 954 cases. Logistic regressionmodels were produced for four configurationsusing a split sample validation technique. A=Standard TRISS components (age 0–54,55+); B=A plus PMC status (present, absent,not recorded); C=A with age profile 0–54,55–64, 65–74, 75–84, 85+; D=C plus PMCstatus (present, absent, not recorded). PMCwas defined as respiratory, renal, liver orischaemic heart disease, diabetes mellitus orcongenital coagulopathy.Results—The results are shown in table 2.Hosmer-Lemeshow (H-L) score is a goodnessof fit statistic. A lower score indicates a betterfit of data. All results were highly significant(p<0.0001) but this is to be expected becauseof the very large sample size. ROC is thereceiver operating characteristic curve thatcompares sensitivity and specificity in graphi-cal form. The area under the curve shouldapproach unity.Discussion—The addition of PMC status(model B) to the standard TRISS model Aappears to give a better prediction accordingto both the H-L score and the area under theROC curve. There is a greater gain, especiallywith the ROC curve, from refining the agegroups into deciles above age 55 (model C) towhich the inclusion of PMC status has littleto add (model D). This may be because therefinement in age grouping will account formost of the PMC eVect, as prevalence of pre-existing disease rises with increasing age.Conclusion—Analysis of almost 54 000 patientdata files reveals that better prediction of out-come can be achieved by adding eitherpre-existing disease status or a better profile ofage to the existing TRISS formula. The latteris more reliable and is therefore recom-mended.

Diaphragmatic rupture and the associ-ation with occupant position in righthand drive vehiclesSHOBHAN THAKORE, JENNY HENRY, ALISTAIR W

TODD, SHOBHAN THAKORE

Ninewells Hospital, Dundee DD1 9SY

Ruptured diaphragm following blunt traumaoccurs with an incidence of 5% in thoserequiring laparotomy. It was initially reportedas almost entirely a left sided phenomenon,however right sided rupture is now recognisedwith increasing frequency. The vast majorityof case series have come from countries wherevehicles are left hand drive. This study aimedto investigate the influence of occupantposition in right hand drive (RHD) vehicleson the side of diaphragmatic injury.

An analysis of the Scottish Trauma AuditGroup database was performed to gather dataon blunt diaphragmatic lacerations reportedbetween 17 April 1992 to 16 May 1999. Driv-ers and front seat passengers (FSPs) wereincluded in the analysis. Table 3 shows theadmission characteristics of patients and table

4 shows their associated injuries. In total, 35patients were studied, 25 drivers and 10 FSPs.The incidence of right sided rupture was 40%in drivers and 20% in FSPs. The incidence ofassociated pulmonary contusion, rib fractureand liver injury was also higher in drivers.Given the small sample size, these diVerenceswere not statistically significant but show aninteresting trend.

Ruptured diaphragm following blunttrauma was initially thought to involve the leftside almost exclusively. However, a higherindex of suspicion and accidents occurring atgreater speeds has seen right sided rupturesreported with increasing frequency. Themechanism of rupture is thought to be a sud-den rise in intra-abdominal pressure followinga compressive force. The role of lateral shear-ing forces has also been advocated as a causeof rupture. Our results may reflect the factthat the right side of the driver’s body is moreexposed to injury in RHD vehicles, a featureof particular importance in the UK. As rightsided injury is more diYcult to detect, it isimportant that a high index of suspicion ismaintained especially when managing injureddrivers from RHD vehicles.

Mathematical modelling in the develop-ment of a trauma systemT J COATS, S ROBBINS, C LOWDELL, M DAMIANI

Accident and Emergency Department, RoyalLondon Hospital, Whitechapel, London E1 1BB

Current systems of emergency care are oftenbased on historical accident rather than design.The London Severe Injuries Working Grouphas reviewed trauma care in Greater London.To test the eVects of diVerent configurations oftrauma system a novel approach was adopted—mathematical modelling.

The inputs to the model were: (1) the geo-graphical distribution of injury (from ambu-lance service data); (2) the critical and defini-tive interventions for each type of injury; (3)the optimum timing of interventions andspecialists/facilities required (from a clinicalconsensus conference); (4) the relative pro-portions of diVerent combinations of injuries(from UK national trauma data); (5) the loca-tions of hospitals and the specialist servicesavailable in each; (6) the time taken to move apatient to various hospitals (from mappingand drivetime software); (7) flow charts forpossible treatment routes for the 19 common-est types of injury.

Key outputs from the model were (a) thetime taken to achieve critical interventionsand (b) the time to definitive intervention (forexample, surgery).

The diVerent configurations of trauma sys-tem that were then tested included: (1)decreasing the length of time spent pre-hospital by the ambulance service; (2) de-creasing the time taken in hospitals beforereaching specialist care; (3) pre-hospital triageto bypass local hospitals if appropriate special-ist care was within 20 minutes travel time fromthe accident scene; (4) inclusion of theLondon Helicopter Emergency Medical Serv-ice (HEMS). Maps were then drawn toillustrate numbers and geography of patientsmeeting clinical targets. For example, the pro-portion of patients with intracranial haemato-mas that reached a neurosurgeon within fourhours was 20% if all patients were transportedto the nearest hospital, 52% if both ambulanceand hospital times decreased and HEMS wasincluded, and 90% if all patients within a 20minute radius of a specialist centre were takenthere rather than the nearest hospital.

This is a novel approach to the developmentof trauma services. The method is timeconsuming, but once programmed the math-ematical model allows the clinical eVect of dif-ferent options to be tested, and quantifies thetrade oVs between time to critical interventionand time to definitive intervention.

Role of neutrophil L-selectin in post-traumatic organ failureT H RAINER, N Y L LAM, T Y F CHAN, R A COCKS

Accident and Emergency Medicine AcademicUnit, The Chinese University of Hong Kong andPrince of Wales Hospital, Hong Kong

Objective—To determine whether early nu-merical and functional changes in circulatingneutrophils, the expression of neutrophil andsoluble L-selectin predict the development ofpost-traumatic complications.Methods—Ethical approval was obtained for aprospective, observational study to investigatethe immune response of patients to injury.Neutrophil counts, expression of L-selectin(mean channel fluorescence, mcf) and solubleL-selectin (sL-selectin) were measured usingcell counters, flow cytometry and ELISA in

Table 2

Model H-L score Sensitivity (%) Specificity (%) Accuracy (%)Area under ROC (95%confidence intervals)

A 70.3 99.32 35.46 95.97 0.925 (0.920, 0.930)B 42.5 99.26 36.20 95.96 0.929 (0.925, 0.934)C 36.2 99.33 38.64 96.15 0.940 (0.936, 0.944)D 34.2 99.34 38.75 96.17 0.941 (0.937, 0.945)

Table 3 Admission data of drivers and FSPswith ruptured diaphragms

Drivers FSPs

Total number 25 10Male 22 4Female 3 6Mean age 32.8 32.6Median ISS 34 31.5Median SBP 115 122.5Median GCS 14 15Left sided rupture 13 (60%) 8 (80%)Right sided rupture 10 (40%) 2 (20%)Bilateral rupture 2 0MV v MV 17 3MV v other 7 4

Table 4 Associated injuries in drivers and FSPswith ruptured diaphragms

Driversnumber*

FSPsnumber*

Chest injuriesRibs (n) 14 2Haemothorax 10 4Pneumothorax 8 3Pulmonary contusion 17 3Flail segment 4 0Myocardial contusion 2 2Mediastinal haemorrhage 2 0Aortic injury 2 1Abdominal injuriesLiver injury 15 3Splenic injury 11 6Mesenteric injury 3 4Large bowel injury 3 0Small bowel injury 2 0Renal injury 5 0Pelvic (n) 9 9Vertebral (n) 5 2Severe head injury 7 1

*Number of patients.

Abstracts 437

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

164 trauma patients (mean (SD) age 39 (15)years; 81% male) triaged to an emergencydepartment resuscitation room within fourhours of injury (median 60 minutes; inter-quartile range 50, 90; range 30–240). The pri-mary outcome measures were organ failure(OF; n=51), multiple organ dyfunction syn-drome (MODS; n=20), acute lung injury(ALI; n=10) and mortality (n=21). All values(means (SD)) were analysed with the Mann-Whitney U test.Results—Neutrophil counts were higher inthose who developed OF (9.5 (5.1) × 109/ml v12.1 (7.1) × 109/ml; p<0.05) and ALI (10.1(5.6) × 109/ml v 15.9 (7.9) × 109/ml; p<0.02).Neutrophil L-selectin mcf was higher in thosewho developed OF (7.9 (3.4) v 9.2 (3.2);p<0.05). Soluble L-selectin levels were lowerin those who developed OF (763 (231) ng/mlv 670 (243) ng/ml; p=0.06), MODS (764(228) ng/ml v 630 (239) ng/ml; p=0.08), ALI(764 (226) ng/ml v 563 (228) ng/ml; p<0.05)and in those who died (757 (230) ng/ml v 523(209) ng/ml; p<0.01). The ratio of totalneutrophil L-selectin (neutrophil count ×neutrophil mcf × % neutrophils expressingL-selectin) to sL-selectin was higher in thosewho developed OF (0.2 (0.1) v 0.1 (0.08);p<0.0001), MODS (0.11 (0.09) v 0.21(0.13); p<0.005), ALI (0.3 (0.1) v 0.1 (0.09);p<0.005) and in those who died (0.21 (0.16)v 0.12 (0.09); p=0.09).Conclusion—Changes in neutrophil L-selectinexpression may have prognostic value forpost-traumatic complications. An early imbal-ance in the relation of neutrophil to plasmaL-selectin favours the development of post-traumatic organ failure.

Systemic arterial pressure wave reflec-tions during haemorrhageP M DARK, J PURDY, J ATHERTON, D DODDS, R A

LITTLE

MRC Trauma Group (NWIRC), StopfordBuilding, University of Manchester, OxfordRoad, Manchester M13 9PT

Objective—To determine the contributionfrom pressure wave reflection to systemicarterial blood pressure shape during a surviv-able haemorrhage (25%) and during re-infusion of shed blood.Design—Prospective randomised control labo-ratory experiment.Subjects—SaVan anaesthetised and mechani-cally ventilated immature large white pigs.Main outcome measures—Aortic reflectionwave ratio (Augmentation Index) was deter-mined at the input to the systemic circulation(AIaa) and the input to the trunk circulation(AIda) using high fidelity digital recordings ofintravascular pressure contours. Thermodilu-tion derived global haemodynamics, oxygentransport and oxygen metabolism parameterswere recorded simultaneously.Results—AIaa fell from control to post-haemorrhage values (0.42 ± 0.14, (n=25), to0.25 ± 0.18, (n=18), p<0.05) and recoveredafter re-infusion of shed blood compared withthe control group (0.45 ± 0.14, (n=9), v 0.44± 0.13, (n=7), p>0.05). AIda followed a similarpattern after haemorrhage, although 8 of 18developed a negative value (−0.35 ± 0.22,(n=8), v 0.20 ± 0.20, (n=10), p<0.05) withassociated conservation of a positive AIaa (0.17± 0.08, (n=8), v 0.30 ± 0.22, (n=10), p>0.05).A negative AIda following haemorrhage wasassociated with the worst thermodilutionderived global haemodynamics.

Conclusions—Pressure wave reflectionswithin the systemic arterial circulation areknown to be important in determiningaorto-ventricular coupling in health and inchronic systemic hypertension. In this study,pressure wave reflection measured at theinputs to the systemic circulation and thetrunk circulation were reduced during sur-vivable haemorrhage and recovered afterre-infusion of shed blood. After haemor-rhage, a negative AIda with conservation of apositive AIaa was associated with the worsthaemodynamic profiles. This pattern ofchange indicates that pressure wave reflec-tion towards the left ventricle from the lowerlimbs is reduced during haemorrhage andgradually replaced by a predominant reflec-tion site originating from within the trunkarterial system. Such changes may notfavour optimal aorto-ventricular couplingduring the evolution of haemorrhagicshock.

From “ambulance volantes” to accidentand emergency squads—is deploymentstill appropriate?J J MCINERNEY, M NORWOOD, G LLOYD, D QUINTON

The Leicester Royal Infirmary, Infirmary Close,Leicester LE1 5WW

Introduction—Similar to Baron Larrey’s “ambu-lance volantes” deployed during the Napo-leonic wars, accident and emergency (A&E)flying squads today provide assistance “in thefield”, despite a paucity of evidence to supporttheir continuing use alongside modern para-medics. The aim of this study was to evaluatethe appropriate deployment of a flying squad.Methods—A prospective survey of all serioustrauma patients attending A&E over one year.Ambulance control, which activates thesquad, was kept blind to the study. Traumateam leaders completed a structured pro-forma in every case, assessing appropriatesquad development, listing interventions, anddocumenting time out of A&E. Revisedtrauma scores (RTS), injury severity scores(ISS), inpatient stay and death were alsodetermined.Results—221 patients were included of which15 incidences involved squad deployment.Demographic data for both the squad (S) andnon-squad (NS) groups were comparable.Appropriate deployment occurred in 10 of 15call outs, while in 2 of 206 instances thesquad was under-utilised. Opioid administra-tion was the most frequent squad interven-tion. Mean squad time was 73 minutes (range40–185). A RTS <8 existed in 3 of 15 ingroup S (20%; 95%CI=0.11, 0.61) comparedwith 33 of 206 in group NS (16%;95%CI=0.11, 0.21). Mean (SD) inpatientstay diVered between group (S=11.5 (11.9)days; NS=9.8 (18.5) days), but there was nodiVerence in mortality (S=1 of 15;95%CI=0.001, 031: NS=14 of 206;95%C1=0.03, 0.11).Conclusions—Although appropriate deploy-ment of the flying squad was achieved, therewas occasional delays in activation, as well asduplication of eVort. Furthermore, despiteobvious training benefits, routine deploymentof a flying squad has to be weighted againstthe detrimental loss of senior staV from A&E.A larger study is required to assess the costeVectiveness of continued flying squad provi-sion.

Patients admitted from accident andemergency departments to intensive careunits: a descriptive analysis using theICNARC database and quantification ofthe eVect of admission to a ward prior tointensive care unit, compared with directadmission from accident and emergencyHOWARD SIMPSON, CAROLINE GOLDFRAD, KATHY

ROWAN, MIKE CLANCY

Accident and Emergency Department,Southamp-ton General Hospital, Tremona Road, Southamp-ton SO16 6YD

Introduction—This is the first UK study todescribe accident and emergency/intensivecare unit (A&E/ICU) activity. The hypothesisis tested that admission to a ward prior to ICUleads to poorer outcome compared with directadmission from A&E to ICU.Method—A retrospective analysis of the Inten-sive Care National Audit and Audit andResearch Centre (ICNARC) database wasperformed. Ninety two ICUs have submitteddata between 1996 and 1999. Admissionswere categorised according to source: directadmissions from A&E to ICU; indirect admis-sions who were admitted to a ward betweenA&E and ICU; and non-A&E admissions toICU. The case load, case mix and outcome ofthe three admission groups are described.Multivariate analysis is used to calculate theadditional risk of death associated with admis-sion to a ward.Results—Of 46 587 ICU admissions, 9389were direct admissions from A&E to ICU,2789 were indirect admissions from A&E to award before admission to ICU; 34 319 werenon-A&E admission to ICU. Direct admis-sions were younger, more likely to be male,less likely to have a serious past medicalhistory, more likely to present with trauma,more likely to be admitted out of hours and atweekends, and had the lowest predicted mor-tality and the shortest length of stay comparedwith the other two groups. Indirect admissionswere more likely to present with respiratoryand cardiac conditions, had the highest meanAPACHE II scores, the highest predictedmortality, the greatest length of stay for survi-vors, and the highest mortality compared withthe other two groups. Multivariate analysisdemonstrated an excess mortality of 34%(95% confidence intervals: 19%, 52%) associ-ated with admission to a ward before ICUafter adjustment for case mix (type and sensi-tivity of illness).Conclusion—The case mix and case load of thethree admission groups are as anticipated.However, the process of care among indirectadmissions from A&E to ICU must be investi-gated further to explain this apparentanomaly.

Respiratory emergencies and others

EVects of mask type, and method of ven-tilator triggering on tolerability of non-invasive ventilation: in chronic obstruc-tive pulmonary disease after exerciseI M STELL*, S G ELLUM**, J MOXHAM***Accident and Emergency Department, BromleyHospital, Cromwell Avenue, Bromley, KentBR2 9AJ and **Department of RespiratoryMedicine,King’s College Hospital,Denmark Hill,London SE5 9RS

Success with NIPPV in acutely ill patients withchronic obstructive pulmonary disease(COPD) may be influenced by ventilator/patient synchronisation and interaction. In thislaboratory study nine patients with severe

438 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

stable COPD (3F 6M), mean age 68, FEV1

0.85 litres completed a questionnaire of 12visual analogue scales (VAS), after using eachof eight modalities of NIPPV (randomised,single blind) for five minutes after successiveexhaustive treadmill walks. The NIPPV mo-dalities included two comparisons: (a) betweensix diVerent triggering/flow delivery systems,and (b) between diVerent masks/mouthpiece.The NIPPV modalities included: a conven-tional flow triggered ventilator (Nippy 2), Pro-portional Assist Ventilation (PAV), whichadjusts support from one breath to the next, aventilator with software improved triggering(Respironics Vision), continuous positive air-ways pressure (CPAP) and a prototype ventila-tor in which the clinician controls the trigger-ing with a button (as observation of the patientmay allow accurate synchronisation). Thefollowing physiological parameters were alsorecorded each minute during ventilator use: O2

saturation, pulse, respiratory rate and breath-lessness (Borg scale). The 12 VAS scales wereaggregated into their underlying three themes:(a) the help given by the ventilator, (b) panicarising from using the ventilator, and (c) senseof loss of control of breathing.Results—The results are shown in table 5. Allof the physiological parameters recoveredmost rapidly when using PAV. In conclusionthere was a strong trend for PAV to be the besttolerated form of ventilation, and the one withthe most rapid recovery of the physiologicalmeasures recorded. This project was sup-ported by the National Health Service R&D(South Thames).

Endotracheal intubation in the accidentand emergency departmentJ P BEALE, C A GRAHAM, S B THAKORE, E

DOCHERTY, J BRITTLIFF, A OGLESBY, D BEARD, M A

JOHNSTON, T R J PARKE

Department of Accident and Emergency Medi-cine, Royal Infirmary of Edinburgh, LauristonPlace, Edinburgh EH3 9YW

Introduction—Definitive airway control is fun-damental to eVective resuscitation. Endotra-cheal intubation represents the gold standardof airway management. It may be performedby both anaesthetists and emergency physi-cians with or without drugs.Objective—The aim of this study is to charac-terise endotracheal intubation in the accidentand emergency (A&E) departments in sixScottish teaching hospitals.Methods—A prospective multicentre observa-tional study was set up under the auspices ofthe Scottish Trauma Audit Group. A pro-forma was completed at the time of intubationand subsequently checked by a local investiga-tor in each site. This analysis reports on thefirst complete year of data collection (1999).Results—There were 773 intubations regis-tered during the first year of this study. A&Edoctors performed 536 intubations (69%). A

total of 277 patients were in non-traumaticcardiac arrest on arrival and they wereintubated without the administration of an-aesthetic drugs. Intubation without anaes-thetic drugs was undertaken in a further 123patients, 21 of whom had maintenancesedation given. These two groups are ex-cluded from further analysis. Rapid sequenceintubation (RSI) was performed on a total of373 patients. Trauma patients constituted45% (167) of the RSI group. RSI wasperformed on 107 patients (29%) within 15minutes of arrival. A&E doctors intubated 194patients (52%). The overall complication ratewas 7.5%. There was no statistically signifi-cant diVerence in complication rates betweenA&E and anaesthetics when undertaking thisprocedure.Conclusions—Endotracheal intubation andrapid sequence induction are performed byA&E doctors at least as frequently as byanaesthetists in the emergency department.The complication rates for both specialties arelow and are comparable.

Facial continuous positive airway pres-sure therapy for cardiogenic pulmonaryoedema: a study to assess its eYcacy in anaccident and emergency department set-ting within the UKC READ, J J MCINERNEY, N O’CONNOR, M NORWOOD,P A EVANS

Accident and Emergency, Leicester Royal Infir-mary, Infirmary Close, Leicester LE1 5WW

Introduction—Recent studies outside the UKsuggest that patients with acute cardiogenicpulmonary oedema (CPO) may benefit fromthe application of facial continuous positiveairway pressure (CPAP) support in emergencydepartments. The aim of this pilot study wasto assess the impact of facial CPAP on patientswith CPO within a UK A&E department.Methods—A prospective powered study com-paring CPAP with supplemental oxygen atambient pressure, using historical controlsmatched for CPO severity. Forty patients withinternationally accepted criteria for CPO wereincluded. Twenty patients received 20 min-utes of facial CPAP using a dedicated DragerCPAP system with a 5 cm H2O positive endexpiratory pressure valve and maximal in-spired oxygen (Group CPAP). Twenty con-trols received identical therapy but receivedmaximal inspired oxygen via a non-rebreathing mask (Group C). Outcome meas-ures compared were arterial partial pressuresof oxygen and carbon dioxide, objective clini-cal signs, intubation rate and death. Statisticalanalysis was by a t test and 95% confidenceintervals (95% CI).Results—Patients receiving CPAP had a betterimprovement in arterial blood gas indicescompared with controls, with higher meanoxygenation (CPAP=8.4 kPa and C=3.5 kPa,p=0.017, (95% CI=0.92 to 8.77), and better

mean excretion of CO2 (CPAP=−0.9 kPa andC= +0.9 kPa, p=0.011, (95% CI= −3.27 to−0.45). CPAP produced a higher meanreduction in respiratory rate, and lower meanreductions in blood pressure/pulse rate, al-though these were not statistically significant.The median length of inpatient stay(CPAP=4, C=5 days) and intubation rates(CPAP=0, C=3) diVered between groups, butwere not statistically significant. There was nodiVerence in hospital mortality betweengroups (CPAP=3, C=3).Conclusions—CPAP utilisation in the A&Edepartment impacts favourably on the physi-ological manifestations of CPO. A larger pow-ered study is underway to assess the impact ofCPAP on intubation rate, length of inpatientstay and hospital mortality.

Comparative studies on platelet functionand haemostasis in patients undergoingsurgical trauma: eVects of four intra-venous fluids in patients undergoing hipreplacement. A prospective double blindrandomised studyE C CROWHURST, S J DAVIDSON, J GLENN, S HEPTIN-STALL, W MADIRA, P A EVANS

Accident and Emergency Department, LeicesterRoyal Infirmary NHS Trust, LeicesterLE1 5WW

Using an elective orthopaedic model oftrauma we investigated the eVects of four dif-ferent intravenous fluids on platelet aggrega-tion and measures of coagulation as well asbleeding time and postoperative blood loss.This provided an eVective and standardiseddegree of trauma and enabled pre-traumameasurements to be made without the con-founding eVect of factors such as sepsis, hypo-thermia and acidosis.

We studied 55 patients undergoing primarytotal hip arthroplasty. Two litres (as ATLSstandard) of 4.5% albumin, or Gelofusine orHaemaccel or saline were given during theoperation. Blood was taken and bleeding timeperformed immediately before start of surgery(PRE), at the end of surgery (POST) and twohours after the end of surgery (LAST). Plate-let aggregation was determined in response toa variety of agonists.

Gelofusine and Haemaccel completelyabolished aggregation in response to ristocetinat the POST (p<0.001 and p=0.001) andLAST time points (p=0.001 and 0.006).Albumin inhibited aggregation in response tocollagen) at POST(p=0.004) and LAST timepoints (p=0.006).

There were persistent increases in pro-thrombin F1+2 complex and thrombin/antithrombin III complexes (p<0.0017 andp<0.0013 for all fluids) showing that theresponse to injury in our model was signifi-cant. Persistent reduction of fibrinogen andincrease of INR was seen for all fluids(p<0.0146 and<0.0125 respectively), andpersistent reduction of factor XIII for the col-loid fluids (p<0.0077). APTT and factor VIIIactivity only fell at the POST time point whencolloids were used (p<0.0105 and p<0.0166).

Bleeding time increased for all the colloidfluids (p<0.0324) at the POST time point,returning back to baseline by the LAST timepoint. There were no significant eVects onpostoperative blood loss. Plasma viscosity waspersistently depressed only by saline and albu-min (p<0.0001).

Overall colloid fluids inhibit certain aspectsof platelet function and the coagulation system.This may be clinically useful as trauma is asso-ciated with a pro-thrombotic state.

Table 5

NIPPV modality

Aggregate VAS scores for theme (SD) (n=9)

Help (max 40) Panic (max 40)Loss of control(max 40)

Control (CPAP 2 cm H2O) 31.1 (2.4) 6.3 (1.7) 8.0 (2.1)CPAP 5 cm H2O 32.4 (4.4) 6.3 (3.7) 5.7 (4.3)RV PAV 34.6 (4.0) 6.5 (3.9) 4.6 (2.9)Clinician trigger 30.2 (6.5) 9.3 (5.1) 7.9 (4.5)RV 32.3 (5.9) 9.0 (8.8) 6.5 (5.0)Nippy 2: face mask 28.8 (9.3) 11.3 (6.4) 7.1 (4.8)Nippy 2: nasal mask 31.2 (7.3) 5.7 (3.5) 6.7 (5.2)Nippy 2: mouth piece 27.1 (7.3) 9.9 (6.3) 9.9 (6.0)

RV = Respironics Vision.

Abstracts 439

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

Whole blood coagulation analysis of 40%haemodilution with resuscitation solu-tionsE V BRAZIL, D A PEACOCK, M HERON, T J COATS

Academic Unit, Accident and Emergency Depart-ment, Royal London Hospital, Whitechapel,London E1 1BB

Introduction—Severely injured patients havealtered haemostasis that may be compoundedby the type of resuscitation fluid used. In vivohaemodilution in excess of 40% can occurduring fluid resuscitation. The Sonoclot ana-lyser measures the quality of a developingblood clot, including the rate of fibrin forma-tion (RATE) and the time to reach maximumclot strength (PEAK).Aim—To examine in vitro the eVect of 40%haemodilution with various resuscitation flu-ids on whole blood coagulation by Sonoclotanalysis.Methods—Each of 10 volunteers had ninefresh blood samples taken from a free flowingupper limb vein and added to solution tomake a 40% dilution. The solutions studiedwere 0.9% sodium chloride, Hartmann’s, 5%dextrose, gelofusine, haemaccel, dextran,hydroxy-ethyl starch (HES) and albumin. Oneundiluted sample was taken as control.Results—Rate and PEAK measurements forcontrol and the various solutions are pre-sented in table 6.

Conclusions—The colloids gelofusine, HES,albumin and dextran all impair whole bloodcoagulation. In contrast haemaccel promotescoagulation. Unlike blood diluted with 0.9%sodium chloride, clot quality is reduced byhaemodilution with 5% dextrose and Hart-mann’s.

What is the eVect of reporting all emer-gency department radiographs?J R BENGER, I D LYBURN

Emergency Department, Frenchay Hospital,Bristol BS16 1LE

Objective—To evaluate the eVect of formalradiological reporting of emergency depart-ment (ED) radiographs on clinical practiceand patient outcome.Methods—All radiographs taken in a single EDover a six month period were prospectivelystudied simultaneously in both the emergencyand radiology departments to detect discrep-ancies between the ED interpretation andfinal radiologist’s report. Where a discrepancywas detected the patient was followed up todetermine the source of the discrepancy, thecorrect interpretation and the clinical impactof the reporting process.Results—During the study period, 19 468 newpatient attendances generated 11 749 radio-graphic examinations. Discrepancies weredetected in 175 patients (0.9% of new attend-ances and 1.5% of radiographic examina-

tions). Of these, 136 were subsequently shownto have been incorrectly interpreted in theED, with 40 patients undergoing a change inmanagement as a result. In the remaining 39the ED interpretation was judged to becorrect, with 16 patients undergoing furtherinvestigations or visits to the ED to confirmthis. Review of the discrepant cases suggeststhat some groups of ED radiographs (such asthose interpreted by an ED consultant andcertain peripheral limb films) may not requireformal radiological reporting.Conclusion—Radiological reporting of EDradiographs detects a small number of clini-cally important radiological abnormalitiesthat would otherwise be overlooked, but alsogenerates an appreciable proportion of falsepositive interpretations requiring further in-vestigation. The adoption of a selectivereporting policy may reduce the false positiverate and radiology workload without compro-mising patient care.

Emergency nurse practitioners in theUK—eVectiveness and costI SAMMY, O GOODALL, J WINDLE

Accident and Emergency Department, Hope Hos-pital, Salford M6 8HD

Objectives—To assess the eVectiveness, compe-tence and costs of emergency nurse practi-tioners (ENPs) compared with SHOs in acci-dent and emergency (A&E).Methods—A retrospective study of A&E casenotes, covering a six week period. A total of643 study subjects and 1295 controls wereincluded. Costs were calculated using hourlycosts for both groups of clinicians as obtainedfrom the Trust’s finance department. Thesewere related to rate of work of doctors andnurse practitioners. Costs of training andwork generated per consultation are also dis-cussed.Results—Patients seen by ENPs and SHOswere comparable in terms of age and sex, butwere significantly diVerent with respect totriage category, time of attendance andanatomical site of injury/illness. Patients seenby nurse practitioners were seen quicker thanthose seen by SHOs. This diVerence was sig-nificant. There was no significant diVerencebetween the number of radiographs orderedby nurse practitioners and SHOs. Thenumber of radiographs “misdiagnosed” inboth groups was similar. The number ofpatients with misdiagnosed radiographs whorequired recall was also similar. SHOs weremore likely to refer patients to inpatientteams. But the rate of admission for referredpatients was similar for SHOs and nursepractitioners. Nurse practitioners were morelikely to request senior advice thanSHOs.The number of patients re-presentingfor unbooked reviews was similar in bothgroups. Patients seen by SHOs were morelikely to require admission than those seen bynurse practitioners. The cost of service ofnurse practitioners was similar to that ofSHOs. The costs of training of each groupwas also identified for discussion, but notquantified.Conclusion—Nurse practitioners are capableof providing a safe and eVective service forpatients with less serious injuries. The cost ofsuch a service is comparable with that ofSHOs.

Posters: Trauma (1)

Views and needs of specialist registrarson research in accident and emergencymedicineA WALKER, T B HASSAN, A GRAY

General Infirmary at Leeds, Great George Street,Leeds, West Yorkshire LS1 3EX

Channelling the enthusiasm and needs of spe-cialist registrars (SpRs) in accident and emer-gency (A&E) medicine to do research.TheFaculty of A&E Medicine encourages partici-pation in research as part of the A&E special-ist registrar training programme. SpRs arerequired to document completed researchprojects, presentations, and publications. Aninformal review of SpR research in our regionundertaken in 1999 revealed that 50% ofSpRs had started research projects that theyhad failed to complete.Objectives—A structured questionnaire wasdevised to identify A&E SpRs’ opinions andmotivation for undertaking research.Method—The questionnaires were sent to all27 A&E SpRs in the region, and consisted ofinitial questions to assess the numbers oforiginal publications and formal researchtraining. The second part was a Likert-stylequestionnaire to identify SpR attitudes toundertaking research. Agreement with thestatements made was defined as 5–7, disagree-ment as 1–3, and 4 was neutral.Results—25 (93%) questionnaires were com-pleted satisfactorily. SpRs’ reasons forcarrying our research included; for personaldevelopment, to improve job prospects and tomake them better A&E consultants oVeringbetter advice to their trainees (median 6, range4–7). The major limitations to research werethought to be insuYcient time (median 6,range 2–7) and insuYcient guidance (median5, range 2–6). Overall there was support for aregional research forum to provide expertisein research methodology, help coordinateprojects, and advise on unrealistic projects(median 6, range 5–7). Trainees were keen tobring research ideas to a research forum(median 6, range 4–7). The trainees suggestedthat publication of at least one researchproject in a peer reviewed journal should bemandatory for an A&E trainee (median 6,range 1–7). There was support for either aweekly research commitment (median 5,range 2–7) or training with a 3–6 monthsecondment for a specific project (median 5,range 1–7).Conclusions—SpRs in A&E medicine are moti-vated to undertake research. A regionalresearch forum could help to meet theirresearch activity needs and channel theirenthusiasm.

Study comparing use of plain abdominalfilms among accident and emergencydoctors of varying experienceRUSSELL MCLAUGHLIN

Newtonabbey

Objectives—Determine whether appropriate-ness of request and documentation of findingsvaried with accident and emergency (A&E)experience of referring doctors.Methods—Over six weeks, a list of plainabdominal films (PAFs) requested by theA&E department of Belfast City Hospital wasobtained from the radiology department.Charts were reviewed by an A&E SpRwhereby radiological findings of the A&Edoctor were recorded and the appropriatenessof the indication was determined using RCR

Table 6

PEAK (seconds) RATE

Control 1172 20.7NaCl 966 19Hartmann’s 1745 15.2*Haemaccel 897† 17Gelofusine 1979* 11.5*HES 2353* 8.5*Albumin 1703 10.4*Dextose >2500‡ 3.4*Dextran >2500‡ 3.5*

*p<0.008, †p=0.0195 Wilcoxon signed rank test,solution v control; ‡no PEAK measurement possible.

440 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

guidelines. The doctors were grouped: group1 (< 6 months full time A&E departmentwork), group 2 (6 months–1 year full timeA&E department work), group 3 (> 1 year fulltime A&E department work), group 4 (1 yearAED work part time equivalent—that is, GP).Results—Over six weeks 100 PAFs wereordered, representing 2% total attendance(5274). Fifty eight per cent were not indi-cated. Group 1 ordered 42 PAFs—69% werenot indicated and 48% had no documentedinterpretation. Group 2 ordered 35 PAFs—49% were not indicated and 43% had nodocumented interpretation. Group 3 ordered14 PAFs— 43% were not indicated and 29%had no documented interpretation. Group 4ordered nine PAFs—67% were not indicatedand 33% had no documented interpretation.÷2 Testing using 10% probability showedsignificant diVerence between group 1 andgroups 2, 3 and 4 regarding appropriateradiographic ordering. Significant diVerencewas not shown between groups 2, 3 and 4. Nosignificant diVerence was demonstrated be-tween any groups regarding documentation ofradiological findings.Conclusions—There is statistically significantimprovement in appropriate ordering of PAFsfollowing six months full time A&E depart-ment work. There is no such diVerencefollowing six months but there is a perceivedtrend toward increasing appropriateness withexperience. No significant improvement indocumentation is shown but there is also per-ceived improved documentation with experi-ence. These results strengthen the argumentfor staYng A&E with doctors of six monthsA&E experience and greater, that is, increasedsenior shop floor work.

How well do accident and emergencydepartments resuscitate patients? Apostal questionnaire to assess the views ofburns and intensive therapy unitsE C CROWHURST, N O’CONNOR, P A EVANS


Background—Patients requiring transfer toburns units or intensive therapy units (ITUs)from the accident and emergency (A&E)department should be resuscitated and ad-equately analgesed before transfer to thespecialist unit.Methods—All burns unit consultants and ITUconsultants in the UK and Republic of Irelandwere sent questionnaires on their assessmentof A&E management prior to transfer. If inad-equate, they were asked to specify all deficien-cies that applied. Consultants were identifiedas those listed in the 1997 Directory of Emer-gency and Special Care Units.Results—Response rates: burns survey. Firstcapture January 1999 41%. Second captureJune 1999 14%; overall response rate 55%.ITU survey. First capture June 1999 39%,second capture September 1999 16%; overallresponse rate 55%. Thirty per cent of burnsconsultants thought resuscitation was ad-equate (45% inadequate, 25% no answer),whereas 46% of ITU consultants thought itwas adequate (37% inadequate, 17% noanswer or not applicable). The order ofcommonest inadequacies are as follows. Burnsunit: burn area assessment (89%), too little ivfluid (82%), burn depth assessment (80%),analgesia not appropriate (62%), other injurynot identified (60%), temperature of patient(57.8%), too much iv fluid (53%), inappropri-ate iv fluid (49%), assessment of patient

weight (38%), other (18%). ITU: too little ivfluid (31%), temperature of patient (25%),inadequate venous access (22%), analgesianot appropriate (19%), inappropriate fluid(15%), poor documentation (13%), assess-ment of patient weight (10%), too much ivfluid (9%), inadequate airway (8%), otherpathology not identified (8%), next of kingiven little or wrong information (8%).Conclusions—Our survey suggests that indi-vidual specialties have diVerent interpreta-tions as to the adequacy of treatment. Closerliaison between these specialties is required toimprove and standardise patient care.

Referral of paediatric accidental headinjuries for intensive careS S AHMAD*, M KENNY*, D THOMPSON**, M J

PETERS**Paediatric Intensive Care Unit and **Depart-ment of Neurosurgery, Great Ormond Street Hos-pital for Children, London WC1N 3 JH

Introduction—Head injury is the most com-mon cause of death at age 1–15. The RCSWorking Party (1999) highlights the role ofthe emergency department (ED) and the needfor neuroscience unit referral. Increasingnumbers of children with acute head injury(AHI) are admitted to our paediatric intensivecare unit (PICU) and therefore we undertooka retrospective study to determine the origin,timing and frequency of referrals.

Methods—Retrospective review of retrievaldatabase and admission books 1995 to May2000. Inclusion criteria were, a clear history ofaccidental trauma, same day referral and GCSof <8.Results—Of 173 patients admitted to ourPICU, 13 were excluded as lacking a clear his-tory of accidental trauma. This populationwas similar to other series with (median age 7years range: 4 weeks–15 years), a malepredominance (66%), and pedestrian roadtraYc accidents being most common mech-anism of injury (50%, 87 of 173), whichaccounted for the majority (87%, 13 of 15) ofthe deaths. Craniotomy was rarely required(3%, 5 of 173). Cervical spine and other inju-ries were rarely seen. Cases referred for PICUrose each year. Forty eight diVerent EDsreferred an average of three patients every fiveyears. Eight one per cent of cases werereferred in the eight hour period 1600–2400.Referrals rate in May to September is threetimes that in October to April, with peaksattached to bank holiday weekends. Themedian time between referral and PICUadmission was 3.5 hours. The average dis-tance from ED to ICU was 18.9 miles (tables7 and 8).Conclusion—AHI are clustered in summer.EDs see these cases infrequently. StaV skilledin paediatric advanced airway and head injurymanagement should be targeted for the earlyevening hours. The increased referral rate mayreflect altered local PICU provision but is yet

to plateau. Closer inspection of the mecha-nisms and nature of these injuries in additionto late follow up is planned.

Developing guidelines for paediatric headinjuryP A YOUNGE

Emergency Department, Frenchay Hospital,Bristol BS16 1LE

Objective—To develop pragmatic evidencebased paediatric head injury guidelines.Method—A Medline search using OVID inter-face, evidence based web sites, references ofreferences and regional experts was per-formed. Draft guidelines were drawn up and aprospective three month audit of currentpractice was performed. Guidelines and auditresults were reviewed by a multi-disciplinarypanel and further refinements made. Continu-ous three monthly audit cycles are ongoing.The principles upon which the guidelineswere based are as follows:1 Children with a reduced GCS should have

urgent computed tomography2 The presence of skull fracture is strongly

predictive of intracranial lesions, howevermost children with skull fracture and GCS15 do not have neurosurgical lesions.

3 Children over 2 with intracranial lesions,with or without a skull fracture manifestsymptoms.

4 It is not practical to undertake computedtomography for all children with GCS 15and skull fracture or symptoms. We suggestcombining these predictors by using skullradiographs in children with persistentsymptoms. If a fracture is discoveredcomputed tomography should be arranged,otherwise children are admitted as neces-sary

5 Children under 2 are diYcult to assess, mayharbour occult intracranial injury and havea higher incidence of NAI. We recommendthat they all undergo skull radiography andobservation unless injury is trivial.

6 Other factors associated with intracraniallesions are focal signs, seizures, base of skullfracture, loss of consciousness > 5 minutesand abnormal clotting. These features war-rant computed tomography.Audit demonstrated that our current prac-

tice already followed these principles with aslight overuse of skull radiography in thosewith trivial injury. Observation of children wasoccurring with most recovering within fourhours.Conclusion—By using the evidence, a multi-disciplinary approach and repeating auditcycles we have developed a successful prag-matic set of paediatric head injury guidelines.

Haemorrhage into an arachnoid cyst—aserious complication of minor headtraumaKATIE DE, KATHLEEN BERRY

Birmingham

Background—Arachnoid cysts are intra-arachnoidal cerebrospinal fluid collectionsthat are usually asymptomatic, however theymay become acutely symptomatic due tohaemorrhage and cyst enlargement, whichmay result from minor head trauma. The

Table 8 AHI admission by time of day (% of total admission)

00.01–40.00 04.01–08.00 08.01–12.00 12.01–16.00 16.01–20.00 20.01–24.00

ICU referral 2 0 3 15 46 35ICU admission 39 1 0.6 8 16 33

Table 7

1995 1996 1997 1998 1999 2000

10 20 28 33 46 Est 65

Abstracts 441

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

range of symptoms is wide and many are“soft” signs. Diagnosis is important as thosecausing mass eVect require surgery.Methods—We report a case of a child present-ing with localised headaches following minorhead trauma. Computed tomography demon-strated an arachnoid cyst with evidence ofhaemorrhage and mass eVect, which requiredsurgical intervention. Other cases presentingto our hospital or reported in the literature arereviewed with respect to presenting symptomsand signs.Results—Localised headaches, behavioural orcognitive changes and ataxia are more com-monly associated with this pathology thannausea, vomiting, visual disturbances or sei-zures in the cases we reviewed.Conclusion—This range of symptomatologyfollowing minor head trauma may warrantcomputed tomography when other criteria forthis investigation are not met.

Traumatic spine fractures in Englandand WalesR S SIEBER, M CLANCY, F LECKY, M WOODFORD, D W

YATES

The Trauma Audit & Research Network(TARN), Clinical Sciences Building, Hope Hos-pital, Stott Lane, Salford M6 8HD

Objectives—The Trauma Audit & ResearchNetwork (TARN) database was analysed toascertain spine fracture prevalence, distribu-tion and mechanism of injury.Method—Retrospective analyses were done forpatients with traumatic spine fracture from adataset of 110 129 patients collected prospec-tively by 94 hospital between 1989 and 1999.A spine fracture was defined as a damage to asingle vertebral bone, including burst, anterioror lateral compression, facet, pedicle, laminaror dens fractures with exclusion of spinousprocess and transverse process fractures.Results—9374 patients or 8.5% of databasepatients were registered with at least one spinefracture. The database prevalence for thethree spine levels is 1.8% for cervical fractures(1979), 2.9% for thoracic (3.202) and 3.8%for lumbar (4.193), leading to a three quarterdominance of thoracic and lumbar fractures.The median ISS was 10 in the cervicalsection, 8 in the thoracic and 6 in the lumbarspine. The median database ISS was 10. Themean age was 45 years. On the database anequal proportion of RTA victims (3.372/36.824) and fall victim (5.104/52.745) had aspine fracture. However, falls were the mostfrequent injury mechanism (54.5%/5104),divided in 22%<2 m and 32%>2 m. RTA vic-tims accounted for 36% (3.372). Altogether709 patients had multilevel fractures (7.6% of

all fractures patients). These had mainly frac-tures within thoracic and lumbar segments(473/66.7%). Second are cervico-thoracicfractures (153/21.6%).Conclusions—Unlike other spine injury re-ports, we identified falls as the main cause ofspine fractures even if <2 m. We documenteda large burden of thoraco-lumbar fractures,which is frequently underestimated. Theseresults are gathered from a broad mix of hos-pital and patients and are therefore represen-tative. Prospective data collection of traumapatients is essential for the ongoing analysis oftrauma pattern and the subsequent evolutionof trauma care.

Spinal boards in accident andemergency—their use and abuseK MURALI

Accident and Emergency Department, City Hos-pital, Birmingham B187QH

Introduction—The spinal board is a transportdevice used predominantly in the pre-hospitalsetting. Prolonged retention of an injuredpatient on a spinal board in the accident andemergency (A&E) department can potentiallylead to complications.Aim—To determine the pattern of use of spi-nal boards in a busy inner city A&E depart-ment, draw up a protocol and reassess theusage pattern after implementation of the pro-tocol.Methods—This study is a prospective longitu-dinal study done in two phases. Data werecollected using a proforma in both the phases.Phase two was after implementation of anagreed protocol.Results—The results are shown in table 9.Conclusion—Significant reduction in patienttime on the spinal board was achieved afterthe introduction of a protocol. More impor-tantly, a perceptible change in the attitude tothe use of spinal board has been achieved. Butcontinuing education and monitoring is em-phasised to maintain the positive change.

The use of ultrasound in the initialassessment of blunt diaphragm ruptureSHOBHAN THAKORE*, ALISTAIR W TODD***Ninewells Hospital, Dundee DD1 9SY,**Raigmore Hospital, Inverness

Rupture of the diaphragm occurs in approxi-mately 5% of cases of blunt abdominal traumarequiring laparotomy, and is considered amarker of severity. Late diagnosis is associatedwith herniation and possibly strangulation ofabdominal contents. It is therefore important todetect the injury with early investigations.Chest radiographs and and computed tomog-raphy are commonly used but there are fewreports describing the use of ultrasound. Wehave collected six cases showing clinical situa-tions in which ultrasound may be useful andone that demonstrates a potentially significantpitfall.

Cases were collected at Raigmore Hospitalin Inverness (table 10). Ultrasound precededcomputed tomography in all cases where bothinvestigations were performed. Four cases didnot undergo computed tomography, with thedecision for operative intervention based onthe ultrasound report, chest radiographs andclinical characteristics. The only false negativeultrasound and CT scans were performed onthe same patient. The ultrasound showed theleft hemidiaphragm to be moving with respira-tion and it was therefore felt to be intact. Wesuggest that this sign depends on spontaneousrespiration, where diaphragm movement isproduced by its own contraction. Positivepressure ventilation causes movement of thediaphragm and inferior structures by expand-ing the overlying lung. This can lead to thefalse impression of an intact hemidiaphragm.

In diaphragm rupture, chest radiographscan be helpful, but are diagnostic in only27–64% of left and 17% of right sidedinjuries. Computed tomography is frequentlyused, however reports show wide ranging sen-sitivities. Ultrasound is used in the initialassessment of blunt abdominal trauma withsensitivities similar to computed tomography.It is portable and relatively rapid and hencecan be used in the resuscitation room while

Table 9

Phase one Phase two

Mean waiting time to be seen by the A&E doctor 17 minutes 13 minutes(1–57 minutes) (1–48 minutes)

Mean time to remove spinal board 31 minutes 8 minutes(3–235 minutes) (3–22 minutes)(median 10 minutes)

Mean time on the spinal board when removed after the primarysurvey only

29 minutes 8 minutes(3–45 minutes) (3–22 minutes)

Mean time on the spinal board when removed after the primarysurvey and radiographs

147 minutes —(60–235 minutes)

Table 10 Results of initial investigations

Case Chest radiograph Ultrasound scan CT Surgery

A Raised RHD High liver with fluid visible above and below Not done Ruptured RHD with liver herniationB Indistinct LHD LHD completely disrupted Not done Ruptured LHD with herniation of

stomach, colon and omentumC Extensive right haemothorax,

multiple rib #s, raised RHDRHD and liver elevated. Rupture “notexcluded”

RHD rupture with liverherniation

RHD rupture with herniation of liver andliver laceration

D Right pneumothorax andindistinct RHD

Poor views of RHD. Dome of RHDunsighted with considerable free fluid inMorrison’s pouch

Liver laceration, elevatedRHD suggesting rupture

RHD rupture, liver and spleen laceration

E Indistinct LHD Poor views but LHD seen to move withrespiration

Poor views but no evidenceof rupture

LHD rupture at operation 7 days afteradmission

F Right haemo-pneumothorax,# ribs

RHD rupture Not done RHD rupture with liver and greateromentum herniation and liver laceration

G Loss of definition of RHD andright cardiac border, raisedRHD

Liver elevated, no diaphragm seenintervening liver and lung, liver failed tomove with respiration

Not done RHD rupture with liver herniation

LHD, left hemidiaphragm; RHD, right hemidiaphragm.

442 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

potentially life saving interventions are ongoing. This series supports the role ofultrasound in the diagnosis of blunt diaphrag-matic rupture when it is suspected on clinicalgrounds or after the initial chest radiograph.Potential pitfalls should be borne in mind andthe investigation performed by experiencedoperators to maximise sensitivity. An ultra-sound scan supporting the diagnosis ofdiaphragmatic rupture can negate the need forcomputed tomography and hence reduce anypotential delay to operation.

Traumatic haemarthrosis of the kneeM A ANSARI

City Hospital NHS Trust, Dudley Road, Bir-mingham

Objective—To evaluate the injuries sustainedfollowing haemarthrosis of the knee in a rela-tively younger population and to assess theneed of doing culture of aspirated blood.Pateint and methods—Notes of 510 patientswere retrospectively studied to record age, sex,mechanism of injury final diagnosis and man-agement. Reports of culture of aspirated bloodwere collected from the department of pathol-ogy.Results—Most cases were of 20–30 years ofage group and the male to female ratio was4:1. Slipping on the floor, pavement or stairswas the major cause of haemarthrosis. Sportsinjuries accounted for 38% haemarthrosis. Onfinal discharge, 40% patients had idopathichaemarthrosis, followed by anterior cruciateligament damage in 15% and meniscal injuryin 12% of patients and posterior cruciateinjury in 3%. All cases were attributable totrauma. There were only eight isolates from510 aspirates, five á haemolytic streptococci,two coryne bacterium and one mixed skinflora were cultured. All these were thought tobe non-pathogenic.Conclusions—Sports injuries are a major causeof haemarthrosis in a young population.Culture of aspirated blood showed no growthand thus seemed to be a waste of resourcesand were totally unnecessary.

Poisoning

“Herbal highs”—natural and safe, ornasty and scary?CHARLOTTE DOUGHTY, ALISON WALKER

Hull Royal Infirmary, Anlaby Road, Hull, EastYorkshire HU3 2JZ

Background—Demand for herbal based alter-natives to recreational drugs is on the increase.These are widely available both via theinternet and at many music festivals. They aregenerally assumed by those using them, to bea safe alternative to taking illegal recreationaldrugs. Several studies have highlighted prob-lems with these drugs.Methods—Several possible adverse reactionsto these substances occurred in patients whoattended V ’98 and Leeds ’99 music festivals.Forensics laboratory analysis was carried outon seven herbal drugs sold on site. Mass spec-trometry was undertaken to ascertain whetherthey contained illicit drugs, caVeine or ephe-drine. The analysis was carried out withadvice from the regional drug squad and withthe help of the Forensic Science Service. Theliterature was searched to determine whetherthe herbal constituents were as safe as alleged.Results—The literature search identified dataon eight diVerent herbal constituents. Theseincluded an anxiolytic (Kava) more eVectivethan benzodiazepines. In addition Sweetflag,

was also identified as a potentially carcino-genic substance. The results of the forensicanalysis revealed that two of the sevencapsules contained significant levels of caf-feine, and a further two contained caVeine andephedrine. No significant levels of illicit drugswere identified.Conclusions—Accident and emergency staVshould include the use of herbal medicines aspart of a patient’s drug history. A patient’sclinical condition may be significantly influ-enced by the use of these substances.

ã Hydroxybutyric acid: a new case of drugmisuseS H BOYCE, K PADGHAM, L D MILLER, J STEVENSON

Accident and Emergency Department, CrosshouseHospital, Kilmarnock.

The use of recreational drugs in society isbecoming a widespread problem increasing theworkload of all the emergency services. ãHydroxybutyric acid is one of these, a drugused primarily for its euphoric eVect. ToxiceVects of ingestion include bradycardia, slowrespiration or apnoea, coma and death. Wepresent seven cases, all of which had consumedã hydroxybutyric acid either alone or inconjunction with other drugs and alcohol. Thepresentation, clinical features and managementof these cases is described. All health care per-sonnel involved in the emergency setting needto know of its existence, toxic eVects and initialmanagement with particular reference to air-way control and possible assisted ventilation.

Under-recording of deaths due to drugmisuse?LUCINDA B EBDON, CAROLYN A MEREDITH, JOHN A

HENRY

Academic Department of Accident and Emer-gency Medicine, St Mary’s Hospital, SouthWharf Road, London W2 1NY

Drug misuse is an increasing cause of morbid-ity and mortality in Britain. The currentoYcial figure of 1800 deaths per year due todrug misuse is based on death certificates andcoroners’ reports. We suspected that these fig-ures are an underestimate, and carried out ahospital-based survey of all the deaths occur-ring in an inner city hospital over a one yearperiod (1 January–31 December 1999). Therecords of all patients under 70 years of agewho died in the hospital were examined todetermine whether the patient had been anillicit drug user, and the course of illness andcause of death as recorded on the deathcertificate were noted. Three independentdoctors reviewed each case to determinewhether the cause of death could be attributedto drug misuse.

There was a total of 955 deaths. Of theseseven were recorded as due to drug misuseand reported to HM Coroner. Of the remain-ing 683 available sets of notes, 17 had ahistory of drug misuse. In 12 cases these werereported to the coroner, but in none of thesecases was drug misuse recorded as implicatedin the death. Of the remaining five, drug mis-use was not mentioned in any of the deathcertificates. The three independent doctorsattributed death to drug misuse in 6 of the 17cases with a high concordance.

Deaths due to drug misuse are probablyunder-reported. Further work is needed toestablish its extent. Drug misuse should forma part of history taking and be recorded in theinterests of the patient to ensure a correctdiagnosis. This will also lead to better record-

ing and reporting of the consequences of illicitdrug use.

Detection of alcohol misuse: attitudes ofjunior doctorsJ S HUNTLEY, C BLAIN, R TOUQUET

Department of Accident and Emergency Medi-cine, St Mary’s Hospital NHS Trust, PraedStreet, London W2 1NY

Introduction—Alcohol misuse is a major bur-den on the emergency system. Despite thedeveloping the one minute Paddington Alco-hol Test (PAT), the eVectiveness of interven-tions and the possibility of medicolegal conse-quences, the detection rate of misuse remainslow. We sought to assess SHO attitudes (onepossible problem) to PAT usage and thedetection of alcohol misuse.Methods—A questionnaire was given to StMary’s accident and emergency (A&E)SHOs, in their last month of employment.Questions pertained to the overall epidemiol-ogy of PAT possibility/positivity, attitudes todetecting/referring alcohol misuse, and selfevaluation of PAT positivity/misuse. Cur-rently, data are available for three serial teams(n = 13, 12, 12; total = 37).Results—SHOs varied widely in their preva-lence estimations of PAT possible complaints,but were unanimous on the importance ofearly detection, that detection was their remit,and that treatment could be successful. Themajority thought A&E to be appropriate forPATs (35 of 37), that drinking alcohol waspart of our culture (32 of 37), but did notthink that it should be negligent to miss a PATpossible condition (30 of 36). SHOs (12 of13) found audit increased awareness ofalcohol as a root cause; 11 of 13 thought theyidentified more PAT positive patients. Lack oftime was identified as a limiting factor. Thefact that 16 of 37 SHOs thought themselvesPAT positive, with 25 of 37 admitting alcoholmisuse >1/month, is an indication of theextent to which alcohol (mis)use is anaccepted part of our culture.Conclusions—SHOs are committed to theprinciples underlying the PAT and theimportance of early detection of alcoholmisuse.

Transatlantic diVerences in the manage-ment of alcohol intoxicationSENTHI SALLATURAY, JOHN A HENRY

Department of Accident and Emergency, StMary’s Hospital, South Wharf Road, LondonW2 1NY

The large cultural diVerences between Britainand the Unites States of America led us tocompare policies towards violent or intoxi-cated patients in four hospitals in New Yorkand four hospitals in London. Each hospitalwas visited and answers to a 15-pointquestionnaire were obtained from a seniorstaV member. The four New York hospitalshad similar policies. In each, intoxicatedpatients were restrained to prevent them leav-ing the department, if necessary by applyingrestraint bandages until they were deemed tohave zero alcohol levels as judged either bybreath alcohol levels or by calculating thedecline from an initial blood alcohol level. Thefour London hospitals were similar in thateach attempted to exclude potentially seriousconditions before allowing the patient to leavewhen clinically judged to be in no danger ofimmediate harm. Blood or breath alcohol lev-els are not routinely measured. In contrastwith New York, violent or disruptive patients

Abstracts 443

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

were usually ejected from the department.This tale of two cities demonstrates howwidely diVering approaches may be usedtowards the same problem. The New Yorkapproach appears to be a defensive policybased on minimising the risk of subsequentlegislation against the potential consequencesof an alcohol intoxicated patient being allowedto leave. In Britain there seems to be a policyof balancing the duty of care towards anintoxicated patient with the duty of caretowards other patients in the department, sothat a seriously disruptive patient is ejectedeven though intoxicated.

An in vitro assessment of the mechanismof adverse reactions to pharmaceuticalacetylcysteineRAJ BANERJEE, PHILIP DARCY, JOHN A HENRY

Department of Accident and Emergency, StMary’s Hospital, South Wharf Road, LondonW2 1NY

Background—Paracetamol overdose is com-mon, leading to 30 000 hospital admissionsand 345 deaths each year. Acetylcysteine (AC)will prevent toxicity if given early enough afteroverdose. However, its administration may becomplicated by anaphylactoid reactions. Parv-olex is a commercial preparation of AC forintravenous administration that contains AC200 mg/ml together with Na-EDTA 33 mg/mlas a stabiliser. As part of a research pro-gramme into the mechanism of toxicity of ACwe have examined its eVects in rat mesenterymast cells. Our objective was to determinewhether Na-EDTA contributed to the toxicityof the intravenous preparation.Methods—Rat mesentery was isolated andtreated with various concentrations of AC,Na-EDTA and Parvolex. The mesentery wasthen placed on a slide, stained, dehydratedand fixed. Cells were examined under blindedconditions from a video record to determinewhether they were intact or degranulated—that is, whether they showed disruption of his-tamine granules. The extent of disruption wascalculated as a percentage of that observed atthe highest dose of AC administered.Results—EDTA showed minimal eVect onmast cell degranulation, with a mean degranu-lation of 7% (range 6–9%) at doses 0–1290mM. AC showed a dose dependent increase inmast cell degranulation ranging from 20% at10 mM to 100% degranulation at 80 mM.Parvolex also showed a dose dependent eVecton mast cell degranulation ranging from 15%at 10 mM to 100% at 80 mM. Our findingsshowed that AC had a dose dependent, andtherefore, anaphylactoid eVect on mast cells,while Na-EDTA had no such eVect, eitheralone or when formulated with AC. A changein formulation is unlikely to aVect the severityof reactions, but our results and clinicalexperience suggest that a lower initial dosemight reduce the severity of the reactions,which are dose dependent.

Compliance of activated charcoalS K SHUBBER, F DUNN, L A MCKINNEY, J STEELE

Altnagelvin Area Hospital, LondonderryBT47 1SB

Objective—To assess the compliance of acti-vated charcoal in adults and children.Design—A questionnaire was sent out to oneconsultant in each of the 260 accident andemergency (A&E) departments nationwide inthe United Kingdom and the Republic of Ire-land as per BAEM directory 1999/2000. Thequestionnaire asked firstly whether compli-

ance of activated charcoal was a majorproblem in adults and/or children and sec-ondly what proportion of the prescribed char-coal was ingested on average by the patientsboth in adults and children respectively. It alsolooked at the commercial types of activatedcharcoal dispensed in the departments.Results—131 (50%) departments respondedto the questionnaire. Only 40% felt that com-pliance of activated charcoal was a majorproblem in adults compared with 76.2% whofelt compliance was a major problem inchildren. It was perceived that 68.82% of theprescribed activated charcoal would be swal-lowed in adults compared with 39.7% in chil-dren. Vomiting was perceived to be next mostimportant problem after compliance.Conclusion—Compliance of activated charcoalis perceived to be greater problem in childrenthan in adults.

Poison information in the accident andemergency departmentS K SHUBBER, F DUNN, L A MCKINNEY, J STEELE

Accident and Emergency Department, Alt-nagelvin Hospital, Londerry BT47 1SB

Objective—To assess how accident and emer-gency (A&E) departments access poisoninformation.Method—A questionnaire was sent out to eachconsultant led A&E department in UK andIreland as seen in the BAEM directory 1999/2000. The questionnaire asked how poisoninformation was accessed giving three mainoptions, phone, Internet Toxibase and previ-ous printed sheets.Results—260 consultants were sent a question-naire, 130 replied, of these 11 (8.4%) weredepartments seeing less 25 000 new patients ayear, 55 (42%) were seeing between 25 000–50 000, 48 (36.9%) 50 000–75 000 and 16(12.3%) over 75 000. Fifty five departmentsused only the phone to access the informationthough eight (14.5%) had plans to install theInternet Toxibase. Printed sheets were used in14 (10.7%) but only one (0.7%) departmentused them on their own. The InternetToxibase was used in 67 (51.5%) of thedepartments in which 31 (46%) departmentstill used a phone.Conclusion—Approximately 63 (48.4%) of thedepartments still do not have Internet Toxi-base installed.

Can a CD-ROM teaching programmeimprove the confidence of accident andemergency SHOs in assessing deliberateself harm patients?ALISON WALKER, JANE BRENCHLEY

General Infirmary at Leeds, Great George Street,Leeds, West Yorkshire LS1 3EX

Background—Studies have shown that theassessment of deliberate self harm patients inthe accident and emergency (A&E) depart-ment is often inadequate. Many A&E SHOshave little training in the assessment of thesepatients. Such patients can present diYcultiesin both assessment and management and fre-quently present “out of hours” when moresenior advice may not be available within thedepartment.Methods—A CD-ROM was produced cover-ing topics on deliberate self harm assessmentincluding basic history taking, mental stateexamination, assessment of risk, the MentalHealth Act, and management of deliberate selfharm patients. Multiple video segments sum-marise the assessment of a patient who hadallegedly taken an overdose and presented to

an A&E department, and the trainee is guidedthrough the management of the patient. Base-line knowledge and confidence in assessmentwere recorded using Likert-style question-naires. Confidence was defined as a score of5–7. After completing the CD-ROM thetrainees repeated the questionnaires and a fur-ther assessment of the value of the CD-ROM.Results—14 SHOs completed the study. Thepre-CD-ROM assessment showed reasonablelevels of confidence in the assessment ofdeliberate self harm cases (median 5, range3–6) but low confidence levels in the applica-tion of the Mental Health Act (median 3,range 1–4). Following the CD-ROM 79% ofSHOs felt more confident in the assessment ofthese patients (median 5, range 4–7) and allfelt more confident applying the MentalHealth Act (median 5, range 2–6) an eVectthat was sustained over the following threemonths. The SHOs also assessed the useful-ness of the CD-ROM as a learning tool andrated it highly compared with other teachingmethods. They found it easy to use and ahelpful educational tool, and recommendedits use in the teaching of deliberate self harmassessment for future trainees.Conclusions—The majority of trainees foundthe CD-ROM format to be preferable to othereducational methods for this subject.

Does having the algorithm in front of youmake a diVerence?MARK G JENKINS

Ulster Hospital, Belfast

Aim—To assess the eVect on simulated paedi-atric resuscitation, if the algorithm is in frontof the operator.Subjects—10 senior house oYcers (SHO), sixwith six months accident and emergency(A&E) experience and all with no paediatricexperience or courses.Methods—Three scenarios involving acutepaediatric emergencies were simulated. Eachscenario had key points. Each doctor wastaken through the scenario twice. On the sec-ond occasion the algorithm was placed infront of the SHO. The algorithms were similarto APLS guidelines and had been collated intoa flip chart.Results—All SHOs improved markedly withthe copy of the algorithm in front of them. Inparticular choice of drug and drug dose werecorrect.Conclusion—Algorithms simple and readilyaccessible improve the quality of simulatedpaediatric resuscitation. For departments withpaediatric attendances such algorithms couldresult in better outcomes until definitive helpand care arrive.

Audit of antibiotic usage in a districtaccident and emergency department:before and after introduction of guide-linesO O JIBUIKE*, ROGERS SAGE**, BARBARA BAIRD**,MICHAEL IMANA***Accident and Emergency Medicine UniversityHospital of Wales, CardiV, **Basildon Hospital,Basildon, Essex

Objective—To evaluate the eVectiveness ofantibiotic guidelines in an accident and emer-gency (A&E) department, three months aftertheir introduction.Methods—Patients receiving antibiotics weretraced on computer, October 1998 prior tointroduction and January 1999. Antibioticregimens used were located in the A&Erecords—choice of drug, dose, frequency, and

444 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

duration of course for various conditions werenoted. Changed prescribing habits were evalu-ated in the re-audit by the number of patientsreceiving entirely appropriate regimens asstipulated in the guideline.Results—The results are shown in table 11.For discussion purposes we subdivided theconditions into two groups: (1) skin and sub-cutaneous conditions. Correct prescribingimproved for lacerations and bites, but not intreating cellulitis. (2) Throat, ear, chest condi-tions. Correct prescribing improved in realtonsillitis. There was little change in chestinfection.Discussion—3.6% to 4% of our A&E attend-ances received antibiotics—1 per 26 attend-ances on average. Individually, the guidelinechanged the prescribing habits of A&Edoctors to soft tissue conditions. Overall, therewas statistically significant improvement in theappropriateness of prescriptions (p<0.01) (÷2

test of significance).Conclusion—Three months after introduction,overall, our antibiotic guidelines eVectivelyimproved practice significantly. The problem,though, is with antibiotic course duration,where our prescribers seem uncomfortablewith five days

Acute medicine

Pulse oximeter waveforms in healthyhumansP M MIDDLETON, A J RETTER, J A HENRY

Academic Department of Accident and Emer-gency Medicine, St Mary’s Hospital, SouthWharf Road, London W2 1NY

Variations in pulse oximeter waveforms(POW) may be used to provide a non-invasivesurrogate of the intra-arterial waveform. Thispotentially represents a readily accessiblemeasure of cardiovascular status. Before theinvestigation of patients with diVerent diseasestates, we examined the eVects of age, bloodpressure and smoking on waveform patterns.These factors are known to decrease arterialcompliance, increasing peripheral wave reflec-tion and altering the contour of the pressurewaveform.

Supine recordings were taken from 185healthy volunteers, recorded on Micromedicalpulse software and analysed on a custom Mat-lab program. Pulse oximeter probes wereplaced on index and middle fingers of the lefthand. Factors known to aVect peripheral arte-rial compliance were recorded. Blood pressurewas measured using a Propaq monitor. Thewaveform patterns were categorised into fourdistinct classes, similar to those described byMurgo and Nichols for intra-arterial pressurewaveforms. Wave classification was comparedwith age and systolic blood pressure usingANOVA, and with smoking status using ÷2

analysis.

ê Analysis gave a measure of agreementbetween index and middle fingers of 0.894(p<0.001). Post-hoc ANOVA revealed thatthere was a significant relation between wave-form class and age (p<0.001), blood pressure(p<0.001), and smoking status (p=0.021).

These results show that POW from indexand middle fingers can be considered as iden-tical. Waveform classification varies with age,blood pressure and smoking. These eVectsmay represent changes in the peripheral arte-rial system resulting in altered wave reflection.We have shown that POW produces reproduc-ible patterns. Further research in this area maylead to the use of this technique as a measureof peripheral vascular responsiveness, in thediagnostic evaluation of acute disease.

Accidental hypothermia and the systemicinflammatory response to rewarmingJ J MCINERNEY, W MADIRA, T DAVIES, A B BREAKELL,P A EVANS

Accident and Emergency, The Leicester RoyalInfirmary, Infirmary Close, Leicester LE1 5WW

Introduction—The mortality rate from moder-ate accidental hypothermia (core temperature< 32°C) is in excess of 50%. Previous studyhas revealed significant pH related alterationsin divalent cations and parathyroid hormone(PTH) during rewarming. The primary aim ofthis study was to determine the eVect ofrewarming on tumour necrosis factor á(TNFá) and interleukin 6 (IL6). The second-ary aim was to determine the relation of IL6 tothe PTH/ionised calcium axis.Methods—Eight patients, four men and fourwomen, age 45 to 85 years (mean age 74.3),admitted with accidental hypothermia wereincluded in the study. Patients were rewarmedwith dry warm blankets and fluid replaced bycrystalloid at 40°C. Blood for serum TNFá,serum IL6, ionised calcium (Ca2+) and serumPTH was collected at presentation, duringrewarming, and at 24 hours.Results—Four patients were admitted withmild (32–35°C) and four with moderate (28–32°C) hypothermia. All patients responded torewarming, with the initial mean rate rise incore temperature 0.78°C/h. There was a nega-tive correlation between serum TNFá andincreasing temperature in all patients duringrewarming (r=0.63, p=0.0008), while IL6 andPTH showed a similar though non-significanttrend (r=0.18, p=0.18, and r=0.22, p=0.14respectively). Comparison of IL6 to PTHshowed a positive correlation (r=0.68,p=0.0001). IL6 and PTH correlated posi-tively (r=0.775, p=0.002 and r=0.72,p=0.005) and negatively (r=0.7, p = 0.004and r=0.71, p=0.004) with Ca2+ in moderateand mild hypothermia respectively.Conclusion—Rewarming is associated withdecreasing serum levels of the cytokinesTNFá and IL6. Poor outcome—that is,

survival < 7 days, was associated with age(>84 years), presentation temperature(<32°C), and non-physiological correlationbetween IL6, PTH, and Ca2+.

Osteoporosis risk factors associated withfractured necks of femurC TOVEY, T LECKIE


Introduction—A recent report from the RoyalCollege of Physicians (RCP) Osteoporosis:Clinical Guidelines for Prevention and Treatmentrepresents the first definitive, evidence-basedguide to managing the disease. The manage-ment of osteoporosis can be complex but thePrimary Care Rheumatology Society recom-mend a minimum of activity that all practicesshould be considering. The guidelines statethat as a minimum those patients receivingosteoporosis prophylaxis should include: (a)previous osteoporotic fracture; (b) earlymenopause (age 45 years or less); (c) patientstaking prednisolone greater than 7.5 mg dailyfor three months over age 50 years; (d)nursing home elderly should be on calciumand vitamin D tablets. The aim of this study isto look at 100 consecutive patients attendingaccident and emergency (A&E) with fracturednecks of femur and to determine whether theyhave risk factors for osteoporosis and whetherthey are taking osteoporosis prophylaxis.Results—Up until 1 April we have the resultsof 51 patients and it is clear that few patientstake osteoporosis prophylaxis despite risk fac-tors. Twenty one patients had had a previousosteoporotic fracture and only three were tak-ing prophylaxis. Eleven patients had earlymenopause (or hysteretomy) before the age of45 (four of these patients also had a previousosteoporosis fracture, none of whom weretaking osteoporosis prophylaxis). A numberof patients were also residents of nursinghomes.Conclusion—There is very little point in theDepartment of Health commissioning theosteoporosis report by the RCP if their adviceis not implemented. Orthopaedic staV shouldbe aware of the guidelines and patients withosteoporotic fractures assessed for osteoporo-sis prophylaxis before discharge from hospital.A&E should perhaps notify GPs that theirpatients have sustained an osteoporotic frac-ture and should be considered for osteoporo-sis prophylaxis. The osteoporosis risk indaughters of elderly patients should also beconsidered. The RCP report recommendsthat primary care take a case-finding approachto osteoporosis using bone density measure-ment as a cost eVective tool for identifyingindividuals at high risk of fracture.

Table 11

Clinical condition

October 1998 (n=224) January 1999 (n=221)

Total number of antibioticprescriptions

Number with appropriateantibiotics, dose, frequency andduration (%)

Total number of antibioticprescriptions

Number with appropriateantibiotics, dose, frequency andduration (%)

Laceration 14 3 (21) 25 9 (36)Bites 18 1 (6) 43 18 (42)Cellulites 30 2 (7) 14 2 (14)Tonsillitis 10 1 (10) 18 2 (11)Chest infection/pneumonia 38 6 (16) 37 6 (16)Otitis media 13 4 (31) 27 8 (30)

0.001<p<0.01.p Value is very significantly less than 0.05 at 95% confidence limits.

Abstracts 445

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

Meningococcal disease in accident andemergency: characteristics and manage-ment. Can diagnosis and care be im-proved?D R H BECKER, T A CHANDLER, P A EVANS

Leicester Royal Infirmary, Infirmary Square,Leicester, LE1 5WW

Objectives—(1) To identify the sensitivity ofthe characteristics of meningococcal disease atthe time of presentation in accident and emer-gency (A&E) to assess their diagnostic values.(2) To evaluate the quality of current manage-ment in A&E.Methods—A retrospective analysis of the notesof patients identified as having had meningo-coccal disease who came through A&E over a21 month period.Results—60 casenotes were available (88%);results refer to where information was re-corded. Twenty one (35%) were adults and 39(65%) were children under 16. Sixteen (27%)were GP referrals of whom 12 (75%) hadreceived antibiotics. Forty nine (89%) saw adoctor within 30 minutes, (mean 16 minutes)and mean time to antibiotics after presentingwas 24.8 minutes with 25 (71%) within 30minutes and 34 (97%) within one hour.Headache, fever, vomiting and rash were themost sensitive features, but occurred in nomore than 79% of patients (lower 95% confi-dence interval parameters, see charts). Thediagnosis was made/suspected in 45 (75%)patients. Thirty six (60%) were referred to award team, 23 (38%) to PICU/ITU, and nonedischarged with one (2%) not recorded. Eight(29%) patients sent to a ward ended up onITU/PICU. Fifty six (94%) patients wereeventually discharged, two (3%) were trans-ferred and two (3%) died.Conclusions—At initial A&E presentation thereare no highly sensitive features to this disease.Formal A&E protocols based on symptoms/signs are unfeasible and are no substitute forclinical judgement based on sound teaching.Despite the inconsistency in presentationdefinitive treatment of identified patients isachievable within 60 minutes of presentation.All such patients need definitive resuscitationand treatment in A&E. ITU/HDU referralshould be considered for all suspected cases.More research from the A&E perspective isneeded.

Test lung comparison of the triggering ofnon-invasive ventilation ventilators underconditions simulating acute exacerbationsof chronic obstructive pulmonary diseaseI M STELL*, G PAUL**, K LEE**, J PONTE***, J

MOXHAM*****Accident and Emergency Department, BromleyHospital, Cromwell Avenue, Bromley, KentBR2 9AJ, **Departments of Mechanical Engi-neering, ****Anaesthetics and ***RespiratoryMedicine, King’s College, University of London

There is increasing evidence that the ability ofnon-invasive ventilation (NIPPV) ventilators tosynchronise with the patient’s breathing is animportant determinant of success with NIPPV;

and that the available ventilators vary consider-ably in this respect. Comparisons between ven-tilators are conventionally made using “testlungs” to ensure stable test conditions and toavoid the need for invasive monitoring ifpatients were used. In this collaborativemedical/engineering project, a new electroni-cally controlled test lung was developed, able tomimic the tidal flows of acute exacerbations ofchronic obstructive airways disease, and inwhich the physiological factors likely to aVecttriggering could be independently varied. Thir-teen ventilators were tested under a range ofconditions selected to be increasingly challeng-ing to ventilator triggering. Condition A: “low”airway resistance (Raw) (3.5 cm H20/l/s at 60l/min), functional residual capacity (FRC) 3.3litres, 2 mm leak, tidal volume (Vt) 782 ml,condition B: “high” Raw (20.0 cm H20/l/s at 60l/min), FRC 3.3 litres, Vt 629 ml, with threeleak sizes: 0, 2 and 4 mm diameter leaks; andcondition C: “high” Raw, FRC 6.3 litres, and 4mm leak, Vt 422 ml. For each ventilator, undereach condition, the test lung was run for oneminute, and the mean (ci) (ms) inspiratory andexpiratory trigger delay times over 20 breathswere calculated; these being the delays from theonset and end of the test lung inspiration to theappropriate ventilator responses.Results—The results are shown in table 12.Conclusion—(a) NIPPV triggering variesmarkedly between ventilators, and (b) formost ventilators is poorer under the morechallenging conditions that may be encoun-tered in acute exacerbations of chronicobstructive pulmonary disease to an extentthat is likely to be clinically important formany of the ventilators tested.

The diagnosis and management of deepvenous thrombosisC TOVEY, J B HOUGHTON, P A DRISCOLL, I A SAMMY,D W YATES


Introduction—The outpatient treatment of deepvenous thrombosis (DVT) was initiated in theUK in 1996 by the Haematology and Accidentand Emergency (A&E) Departments in StPeter’s Hospital, Chertsey. The outpatienttreatment of DVT is safe and leads to substan-tial savings in bed costs but many hospitals stilldo not treat the majority of patients in the com-munity. The aim of this study was to developevidence-based algorithms for the diagnosisand management of DVT and to evaluate anA&E consultant based service.Method—Since November 1999 all patientswith possible DVT are seen by an A&Econsultant or registrar. Diagnostic algorithmshave been evaluated using selective screeningwith SimpliRED D-dimer tests and “nearpatient” screening with digital photoplephys-mography before imaging with ultrasonogra-phy or venography. Patients with a confirmedDVT will see a consultant haematologist withresponsibility for screening for malignancyand thrombophilia.

Results—Audit has shown that 75% of suchpatients will arrive in A&E between 8 am and5 pm (85% from 8 am to 6 pm) on weekdaysmaking a consultant based service easy toachieve. Nearly 50% of patients arriving inA&E will be seen within half an hour and 75%of patients will see an A&E consultant on theinitial visit. More than 40% of patients under-going venography or ultrasonography willprove to have a DVT (it is accepted inter-nationally that 25–30% positive rate is reason-able). In six months there has been only one“missed” DVT. This patient was not suitablefor venography and was discharged after anegative ultrasound investigation. She re-turned two days later with pleuritic chest painwith an intermediate probability of a pulmo-nary embolus on a VQ scan.Conclusion—The combination of experiencedA&E staV, SimpliRED D-dimer tests and dig-ital photoplethysmography has resulted in asafe and eYcient service with an estimatedsaving of over 1000 bed days a year, areduction in the number of investigations per-formed by radiology and a reduction in theuse of low molecular weight heparin. A&Econsultants who have considered taking aninterest in thromboembolism can gain en-couragement from this study.

The use of a SimpliRED D-dimer test inthe accident and emergency departmentC TOVEY, G YOUSSEF, T UNDERHILL


Introduction—In recent years there has been aninterest in the use of D-dimers to exclude thediagnosis of deep venous thrombosis (DVT).The introduction of a bedside (SimpliRED)D-dimer test allows “near patient” testing inaccident and emergency (A&E). The aim ofthis study was to evaluate the use of this testwhen performed by A&E staV.Method—63 patients attending the A&E de-partment with suspected DVT were enteredinto the study. All patients had blood D-dimerlevels measured by A&E staV using the Simp-liRED test before duplex ultrasound.Results—Three patients with a negativeD-dimer test initially had a positive ultrasoundscan. In two of these patients a venogram wasrequested because of the disparity between apositive ultrasound and a negative D-dimertest. Both of these venograms confirmed theabsence of a DVT. In the third patient a repeatultrasound by a consultant radiologist con-firmed a thrombosis in the superficial shortsaphenous vein rather than a deep vein. Thesensitivity, specificity, negative predictivevalue and positive predictive value of the Sim-pliRED D-dimer test were 100%, 79%, 100%and 69% respectively.Conclusion—Duplex ultrasound is operatordependent and is known to produce falsepositive results (6–8% of scans). The D-dimertests for this study cost £169 (63 tests at £3each) and detected three false positive resultswith duplex ultrasound. If it had not been for

Table 12

Ventilator

Condition A Condition B (4 mm leak) Condition C

Insp delay Exp delay Insp delay Exp delay Insp delay Exp delay

BiPAP 20 86 (20) 168 (25) 107 (14) 128 (35) 236 (37) 416 (46)PB 335 146 (11) 114 (7) 169 (16) 128 (11) 228 (11) 182 (15)VPAP 2 240 (13) 141 (13) 247 (4) 161 (11) 332 (11) 326 (11)Nippy 2 128 (11) n/a (timed) 91 (14) n/a (timed) 175 (11) n/a (timed)Rspncs Vision 86 (13) 88 (13) 90 (13) 99 (10) 86 (13) 88 (15)Rspncs PAV 73 (10) 80 (8) 88 (14) 85 (12) 152 (16) 76 (6)

446 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

this test three patients would have beentreated with warfarin for three months withsubstantial cost to the NHS, inconvenienceand anticoagulation risks to the patients.From a clinical and risk management point ofview duplex ultrasound should perhaps becombined with a SimpliRED D-dimer test tohelp prevent false positive results. In this study34 of the 63 patients with a negative D-dimertest could have been discharged from A&Ewithout an ultrasound investigation and noDVT would have been missed.

Eliminating barriers to rapid thromboly-sis in accident and emergency. Are thetargets appropriate and achievable?A A KHAN, A TAYLOR, C KIRKE, H MCCLELLAND, I

BARLOW, W HAMER, T HASSAN

Accident and Emergency Department, The Gen-eral Infirmary at Leeds, Leeds LS1 3EX

Background—The National Service Frameworkfor Coronary Artery Disease (NSF-CAD) hasset tough targets for patients requiring throm-bolysis in the accident and emergency (A&E)department. At present there is little infor-mation to confirm the feasibility or reproduc-ibility of using or being able to achieve thesestandards in a busy A&E setting.Objectives—To describe the quality assuranceprogramme for thrombolysis of acute myocar-dial information (AMI) in our unit, targetingareas for improvement in the process. Sec-ondly, to suggest a more sophisticated, robust,and reproducible model for quality improve-ment (QI) of patients considered eligible forthrombolysis as compared with that proposedby the NSF-CAD.Methods—Observational study with targetedfeedback of information to clinical staV atregular intervals. Relevant time points fromonset of symptoms to interventions in theA&E department including reported ECGfindings were recorded. Data from January1999 to March 2000 are submitted.Results—A total of 280 patients were includedin the study. Seventeen were excluded becauseof inadequate data collection. Of the remain-ing 263, 114 (43.1%) were thrombolysedwithin 30 minutes. Patients who were notthrombolysed within 30 minutes weregrouped into separate QI cohorts. Seventy five(49%) were deemed to have insuYcient initialECG changes to justify treatment. Seventyfour (51%) had delayed treatment due to anumber of factors including, delay in obtain-ing an ECG, delay in being seen by a doctorand failure to recognise changes suggestive ofAMI. Other groups considered in the QI pro-gramme included those that were throm-bolysed but showed no evidence of AMI andpatients presenting with bundle branch blockon their initial ECG.Conclusions—Significant optimisation of re-sources will be required to meet the NSF-CAD targets for A&E thrombolysis. A moresophisticated and robust QI programme issuggested which will make comparisons be-tween entries more meaningful.

Children

Why do children attend accident andemergency with acute asthma?CLIVE TOVEY, K BAILEY, H CLARKE


Introduction—Children often attend accidentand emergency (A&E) suVering with acuteasthma and after treatment with a nebulised

bronchodilator the symptoms frequently sub-side. As bronchodilators should be as eVectivewhether administered via a spacer, a breathactuated inhaler (BAI), a dry powder systemor a nebuliser, it may suggest that manychildren may be using inappropriate medica-tions, no medications at all, an inappropriatedrug delivery system or have a poor inhalertechnique. The aim of this study was to seewhat children use to treat acute asthma in thecommunity. As children should keep a supplyof medications in school an audit at schoolwould provide useful information on what isused to treat acute asthma in the community.Method—76 schools were visited with 19 485pupils aged 4–12 years.Results—It is known that 10–15% of school-children suVer with asthma but in this studyonly 962 (4.9%) of total schoolchildren tookbronchodilators to school. In the age group4–12 years virtually all children should beusing a spacer or BAI (for example, BricanylTurbohaler, Evans Clickhaler, Ventolin Easi-breathe) as most children aged less than 12would not have the coordination to use anaerosol correctly without a spacer. In thisstudy only 548 (2.8%) children were usingspacers or a BAI whereas 414 (2.1%) childrenwere using bronchodilators without spacers.Steroid inhalers should be used twice a dayprophylactically (at home). A total of 147children were found to have steroid inhalers inschool presumably to treat acute exacerba-tions of asthma as many children did not alsohave a bronchodilator.Conclusion—Many children are using inappro-priate medications or inappropriate devices.Although GPs would prescribe spacers manychildren do not use them as they are quitecumbersome to carry around and children donot like to be seen to be diVerent to theirpeers. A&E should encourage the use of BAIsand although they are more expensive thanaerosols “‘the most expensive inhaler is theone that does not work”.

Characteristics and management of chil-dren and adolescents presenting to acci-dent and emergency with deliberate selfharmA NADKARNI, A PARKIN, N DOGRA, D STRETCH, P A

EVANS

Greenwood Institute of Child Health, WestcotesHouse, Westcotes Drive, Leicester LE3 0QU

Objectives—The aim of this study was to studythe characteristics and management of chil-dren and adolescents presenting to accidentand emergency (A&E) with deliberate selfharm.Method—Descriptive analysis of data collectedby reviewing the notes of all children and ado-lescents aged 16 years and under, presentingduring the period of study (1 January to 31December) with a history of deliberate selfharm.Results—100 children (18 boys, 82 girls) wereresponsible for 117 episodes of deliberate selfharm. Some 69% were accompanied byimmediate family, while 21% children hadpresented alone. Some 49% presented be-tween 5 pm and midnight. Twenty five percent had prior or current contact with Child &Adolescent Mental Health Services(CAMHS). Assessment included investigationof physical complications (88.9%), reasons forthe self harm (83.8%), social circumstances(68.4%) and previous history of self harm(54.7%). After assessment, 72 were dis-charged from A&E and 36 admitted to paedi-atric or medical wards. Of those not admitted,

40 were referred to the local CAMHS for fur-ther assessment, 24 being seen in the A&E inthe first instance and 26 given a first appoint-ment in the CAMHS outpatient clinic.Conclusions—Children and adolescents pre-senting with deliberate self harm to A&Efrequently present alone or are accompaniedby people who are not family membersmaking assessment and treatment diYcult.The time of presentation is usually out ofhours, further complicating this process.Many are already known to CAMHS. Assess-ment tends to focus on the physical conse-quences of the attempt. Although guidelinesproduced by the Royal College of Psychiatristsrecommend that admission to hospital isdesirable, this did not occur in two thirds ofthe cases. The guidelines also recommendassessment by CAMHS in all cases but thiswas followed in only one third of casespresenting to A&E.

The management of alcohol and drugmisuse in children and young people—aSalford Health Action Zone (HAZ) initia-tiveM MALOBA, I A SAMMY, C TOVEY, D W YATES, J

MCKENZIE, T WHITFORD, B LEE, R THOMAS

Accident and Emergency Department, Hope Hos-pital, Stott Lane, Salford M6 8HD

Introduction—A report by a joint workingparty of the Royal College of Physicians andthe British Paediatric Association1 made thefollowing recommendations: (1) There shouldbe a well defined multi-disciplinary youngpeoples alcohol and drug service in eachdistrict to which referrals can be made. (2)Research into alcohol problems in the youngis exceedingly small in relation to the size ofthe problem. Retrospective audits of intoxi-cated teenagers has shown that their treatmentat accident and emergency (A&E) depart-ments and their subsequent follow up isinconsistent. Typically such children attendA&E late at night where they are allowed tosober up before discharge. Most A&E depart-ments would not routinely screen for drugmisuse or carry out a formal psychiatricassessment or arrange follow up.Method—Funding has been obtained from theHAZ Initiative for a school nurse (HAZFellow) who will follow up referrals fromA&E. All young people aged less than 16 (or ifin full time education under the age of 19)attending A&E as a consequence of alcohol ordrug misuse will be referred to the HAZFellow. Links have been established betweenan alcohol counsellor for the young and theDrug Advisory Service. The HAZ Fellow willrefer to these agencies if indicated.Results—Data being collected include: (1)Extent of problem—we will have data that willprovide information on the numbers ofpatients attending A&E and the time spentwith the HAZ Fellow with each patient. (2)Recurrences—data will be collected aboutrecurrent A&E attendances before and afterintervention. (3) School attendance—data willbe collected before and after intervention. (4)Preliminary cost analysis. The cost/benefit ofscreening for drugs and the domino eVect onother services will need to be considered.Conclusion—This study is at an early stage andso it is too early to draw conclusions.

1 Anonymous. Alcohol and the young. J R Coll Phy-sicians Lond 1995;29:470-4.

Abstracts 447

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

Presentation to accident and emergencywith crying or screaming and likelihoodof child protection registrationA K FLETCHER, D P BURKE

Department of Accident and Emergency, SheYeldChildren’s Hospital, Western Bank, SheYeldS10 2RX

Aim—To determine whether children agedless than 2 who present to accident and emer-gency (A&E) with crying or screaming as theonly complaint, are more likely to be placedon the child protection register in later yearsthan children who do not attend with cryingor screaming alone.Methods—The SheYeld Children’s HospitalA&E database was examined for five yearsfrom 1 January 1992. Children were identifiedwho presented at triage with crying orscreaming as their sole complaint. Controlswere taken from children who presented withany other complaint. Matches were made forsex, postcode and date of birth. All nameswere checked against files that contained datesof past or present child protection registration.In January 2000, the children’s age rangedfrom 3 to 10 years. The mean follow up periodwas six years (SD one year seven months).Results—From 1 January 1992 until 31December 1996, 450 children made 462attendances to A&E with crying or screamingas their only complaint. Of these, 12 had beenplaced on the child protection register. Ten ofthe 450 control children had been registered.The odds ratio of subsequent child protectionregistration if a child presents in SheYeld withcrying or screaming alone is 1.21 (95% confi-dence intervals 0.52 to 2.82).Discussion—Presentation of young childrenwho cry or scream for no clear reason is rela-tively common. It has been suggested suchchildren are at risk of abuse. Although childprotection registration is not the same asabuse, it is the closest surrogate marker wehave. This study shows there is no increasedlikelihood of child protection registration forchildren who present with crying or screamingalone and prejudices against parents of thesechildren, if held, are inappropriate.

Children and pain in the neck—should wetake them seriously?A NATARAJAN, D P BURKE, J A FERNANDES, Y G

YASSA

Accident and Emergency Department, PaediatricOrthopaedic Department,SheYeld Children’sHospital, Western Bank, SheYeld S10 4TH

It is not uncommon for children to present tothe accident and emergency (A&E) depart-ment with neck pain. Minor trauma, viral orupper respiratory infections are the common-est causes. Clinicians managing these childrenare often inclined to treat them as torticolliswith analgesics and reassure their parents thatthese are self limiting ailments.

Less common but more sinister underlyingcauses may present as neck pain in childhood.The aim of this paper is to draw attention tothese unusual conditions masquerading asneck pain and to emphasise the need for earlyrecognition of such conditions.

We describe a series of five patients whopresented with pain in the neck to our A&Edepartment over a period of 12 months. Ourdepartment serves 34 000 new patients a year.The cases we report emphasise that thepresence of neck pain especially if persistentand associated with abnormal neck posturingor gait abnormalities should act as a possibleindicator of other less common but potentiallymore sinister diagnoses such as posterior fossa

tumour, eosinophilic granuloma or discitis.Appropriate evaluation of such cases isdiscussed.

Children and magnets—an almost fatalattractionSIMON MCCORMICK, P O B BRENNAN, R N SHAWAS,JANET G YASSA

Accident and Emergency Department, PaediatricSurgical Department,SheYeld Children’s Hospi-tal, Western Bank, SheYeld S10 4TH

Children imitating adults often seem toindulge in harmless fun, however at timesthese activities have more serious implica-tions. The recent fashion of body piercingamong adults has encouraged children to havesimilar procedures or imitate them by apply-ing small magnets to sustain the jewellery usedin piercing in position.

We present a series of 24 cases presented tothe SheYeld Children’s Hospital over a periodof eight weeks. The children imitated bodypiercing by using small powerful magnetsacross parts of their body including nose, ears,penis and tongue. Some swallowed themagnet while attempting to use them.

Several of them have had complications,including one who had near fatal surgicalcomplications.

We describe the details of the patients, theprocedures used to detect the magnets and themanagement of the diVerent complicationsencountered.

The accident and emergency process

Should ambulant patients be directed toreception or triage first?STEVE GOODACRE, FRANCIS MORRIS, BINIAM TES-FAYOHANNES, GARETH SUTTON

Accident and Emergency Department, NorthernGeneral Hospital, Herries Road, SheYeldS5 7AU

Objectives—Developments in triage have leadto patients being directed to triage rather thanreception upon arrival in accident and emer-gency (A&E). We aimed to determine whetherattending triage or reception first, was prefer-able in terms of timeliness, clarity and safety.Methods—We compared two consecutive fourweek periods during which patients weredirected to attend triage first and thenreception first. Observers recorded their ac-tions on arrival. Questionnaires recorded thepatients’ perceptions of each strategy. Highpriority triage categories were audited duringthe reception first phase.Results—1850 patients were observed in thetriage first phase with a mean door to triagetime of 10.6 minutes, triage to reception 5.3minutes and door to reception 15.8 minutes.Altogether 1522 patients were observed in thereception first phase with a mean door toreception time of 0.5 minutes, reception totriage 12.4 minutes and door to triage of 12.9minutes. Patients were more likely to presentto the appropriate place during the receptionfirst phase (88% v 34%) and reported shorterwaiting times and better understanding ofinstructions. No case given triage category oneor two suVered an adverse outcome resultingfrom delay associated with attending recep-tion first.Conclusion—Directing patients to attend re-ception first is timely and less confusing. It issafe provided booking in does not delayassessment by more than 15 minutes.

The changing management of acutemedical patientsN CONTRACTOR, I SAMMY

Accident and Emergency Department, Hope Hos-pital, Stott Lane, Salford M6 8HD

Introduction—Emergency medical admissionsin Britain are rising. A report by the Councilsof International Hospitals in 1996 recom-mended the development of acute assessmentunits. Some of the recommendations in theRoyal College of Physicians Working Partyreport include appointment of a doctor whohas overall management of the unit or aweekly consultant in charge, and twice dailyconsultant ward rounds.Objectives—To assess if medical assessmentunits that are already established meet therecommendations of the report in 1996 andthe new report from the Royal College of Phy-sicians.Method—A list of all hospitals with a CCUand/or ITU in Britain was composed from thedirectory of emergency and special care units.It was assumed these hospitals would be treat-ing acute medical patients. The questionnairewas composed of 13 questions and thenrequested general comments.Results—So far, 40.5% of units have a nameddoctor in charge of the unit, and the majorityof these doctors are from internal medicine.Although, some units are managed by acci-dent and emergency and ITU doctors. Of theremaining units, only 10.4% of these rotatethrough a consultant in charge for the week.Some 5.8% of units have no consultant wardrounds, 42% have one ward round and 50.7%have two.Conclusions—Many hospitals have already de-veloped medical assessment units. The mainarea of deficiency regards medical staV. Themajority of departments do not have a doctorin overall charge of the unit. This is likely tohave adverse eVects in both eYciency and riskmanagement. Also, many departments do nothave additional staV to on call doctors. Medi-cal assessment units are still in the early stagesof development, and as more research is done,continued improvements can be made.

Out of hospital cardiac arrest—factorsaVecting survival. A 3.5 year prospectiveobservational study at Glasgow RoyalInfirmaryFRANK WESTERDUIN, RUDY CRAWFORD

Accident and Emergency Department, GlasgowRoya Infirmary, 84 Castle Street, Glasgow G4OSF

A prospective observational study of allcardiac arrests brought to Glasgow RoyalInfirmary Accident and Emergency (A&E)department over a 3.5 year period (November1996 to May 2000).

Approximately 600 cardiac arrests with anunderlying cardiac cause were managed overthis period. Specific features were examined.VF was the presenting rhythm in 46%. Overallsurvival to discharge from hospital was 5.5%.There were no survivors to hospital dischargewho had not had a return of spontaneous cir-culation prior to arrival at hospital. There wasno improvement in survival in arrests treatedby paramedics as compared with techniciancrews. The majority of arrests (50%) occurredin the home. The majority of arrests (75%)were witnessed and bystander CPR wasperformed in half of these. There were foursurvivors out of seven arrests (57%) occurringat a football ground with onsite first aidersequipped with automated external defibrilla-tors, ambulance crews and medical cover.

448 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

This represents 12.5% of the total survivors todischarge from hospital. The pre-hospitaltimes in relation to survival were examined.

In addition to continuing to highlight theimportance of urgent action by the public,steps need to be taken to reduce the delay todefibrillation. Early defibrillation has beenshown to improve survival in cardiac arrest.Studies show that the best survival rates areachieved in two or three tiered EMS responsesystems compared with a single response ascurrently exists in the UK. The involvement ofthe police, fire brigade and where appropriatethe voluntary ambulance service as firstresponders equipped with automated externaldefibrillators in Britain would be a major stepforward in improving our survival rates in car-diac arrest.

Does live television football aVect attend-ances at accident and emergency depart-mentsS H BOYCE

Crosshouse Hospital, Kilmarnock

Aim—To determine if live terrestrial televisionfootball broadcasts aVect patient attendanceat an accident and emergency (A&E) depart-ment.Method—Glasgow Rangers football club, theScottish league champions, were competing inthe European Champions League season1999/2000. The competition was played forsix weeks, either on a Tuesday or Wednesdayevening. The matches were broadcast live onScottish television, the programme commenc-ing at 1930 and ending at 2200. Patientsattending the A&E department during thesehours on each match night were notedprospectively. Only self presenting patientswere noted. All specialty receiving patientswere excluded. The patients attending on allother week nights during these hours werenoted and a weekly average figure calculated.Results—Attendance was reduced on four ofthe six match nights. The reduction was mostmarked on the night of the first and last game,both deemed to be crucial matches, withattendance being 63% and 51% respectivelyof the weekly average.Conclusion—The results demonstrate thatthere is an apparent decrease in patientattendance during match nights, although notexclusively. Other factors, for example,weather, time of year, also require considera-tion.

A retrospective study on ENT presenta-tion and its appropriateness of manage-ment in accident and emergency depart-mentsM DE, A ANDE

Crosshouse Hospital, Accident and EmergencyDepartment, Kilmarnock KA2 OBE

Objectives—To determine the management byaccident and emergency (A&E) staV ofpatients attending an A&E department withENT problems and to identify problems withpatient management or patient disposal.Method—This is a retrospective analysis of thepatients presenting with ENT problems toA&E department. One hundred patients wererandomly selected from patients attendingover a six month period. They were assessedon time of attendance, mode of presentation,management and appropriateness of suchmanagement.Result—The ages of patients seen ranged from2–94 years. The vast majority (86%) ofattendances were between the hours of 9 am

and midnight. The commonest presentingsymptom was a nose bleed (26%). Some formof treatment was carried out in 68% of allcases. No treatment was required in 32% ofthe cases, where the problem had resolved orwhere no significant problem was detected.The use of other hospital resources (that is,radiology and other laboratory services) wasuncommon and referral to ENT departmentoccurred in less half the patients (44%).Review of the casenotes suggested that themanagement was entirely appropriate in 86%.Conclusion—This study reveals that the major-ity of the patients were managed satisfactorilywith appropriate follow up. It could be arguedhowever, that in view of the infrequent use ofacute hospital resources that a significantnumber of these patients could be treated bytheir own GPs, thereby reducing the workloadon hospital A&E departments.

Trauma (2)

The flow of patients following a majorincident, the Omagh bombMARK G JENKINS, ALAN MCKINNEY


Aim—To chart the flow of the patientsinvolved in the Omagh bomb as they weretransferred around the hospitals of NorthernIreland for ongoing and definitive care.Methods—All hospital attendances followingthe Omagh bomb were identified. The flow ofpatients transferred from the Tyrone CountyHospital were traced using their original notesand notes from their subsequent hospitaladmissions.Results—Three hundred and nine patientsmade immediate contact with the HealthServices following the Omagh bomb. Sevenhospitals were involved in the immediate andaftercare of the patients involved in the bomb.Flow charts will be used to illustrate theimmediate transfers (first 24 hours), second-ary transfers (2–3 days) and tertiary referrals.Conclusion—In this major incident, the initialreceiving hospital was inundated. The bombscene was clear of living casualties within 20minutes. Due to this and The Tyrone CountyHospital’s geographical isolation, a triage sys-tem was used. Patients were referred on toneighbouring, area and regional tertiary carecentres appropriately for continuing anddefinitive care.

Revised trauma scores and mode oftransfer of patients after a major inci-dent, the Omagh bombMARK G JENKINS, ALAN MCKINNEY


Aim—To correlate revised trauma score ofpatients involved in the Omagh bomb andtheir subsequent method of transfer.Method—All patients involved in the Omaghbomb transferred from the Tyrone CountyHospital for further/ongoing treatment wereidentified. From their notes pulse, blood pres-sure, respiratory rate and Glasgow coma scalewere recorded. These same parameters werenoted before and after transfer. Mode oftransfer was also recorded. Their subsequentrevised trauma score was calculated.Results—Three hundred and nine patientsmade immediate contact with the hospitalservices after the explosion. Sixty sevenpatients required transfer to area and regionalhospitals. Forty two went by ambulance, 6 bycar and 19 by helicopter (table 13). Twopatients had deterioration in their revised

trauma score after transfer by helicopter.These will be discussed.Conclusion—In a major incident, there seemsto be a group of patients that “self triage”,they know they are well enough to transferthemselves. Helicopters transfer patientsquickly but there is a risk of deterioration.

Injury severity scores and mode of trans-fer of patients after a major incident, theOmagh bombMARK J JENKINS, ALAN MCKINNEY


Aim—To correlate injury severity scores ofpatients involved in the Omagh bomb andtheir subsequent method of transfer.Method—All patients involved in the Omaghbomb transferred from one of the TyroneCounty hospital for further/ongoing treatmentwere identified. Injuries were recorded fromtheir notes. Anatomical injury scores and sub-sequent injury severity scores (ISS) werecalculated. The mode of transfer from thehospitals to area and regional centres wereidentified from notes and Northern IrelandAmbulance Service records.Results—Three hundred and nine patientsmade immediate contact with the hospitalservices after the explosion. Sixty sevenpatients required transfer to area and regionalhospitals. Forty two went by ambulance, 6 bycar and 19 by helicopter. The patients withinthe car group, self triaged and only one of thisgroup was seriously injured with an ISS of 14(range 1–14, median 3). Ambulance patientshad an ISS range of 1–34 and helicopterpatients had an ISS range of 6–75. An ISS of>16 indicates serious trauma, 9 helicopterpatients and 11 ambulance patients had thesescores.Conclusion—Helicopter transfer for the geo-graphical isolation was an ideal method oftransfer for the patients from the Omaghbomb and staV from Tyrone County hospitaltried to do this. Such transfer is hampered byproblems of staV availability to transferpatients and a desire to fill available space inthe helicopters even if this meant sending lessseriously injured to make up the numbers.

Deaths among car drivers and theirpassengersM PARRIS, J P WYATT, D BEARD, A BUSUTTIL

Forensic Medicine Unit, University of Edinburgh,Scottish Trauma Audit Group and Accident andEmergency Department,Royal Infirmary,Lauris-ton Place, Edinburgh EH3 9YW

Deaths among car drivers and theirpassengers in Lothian and Border regions ofsouth east Scotland between 1994 and 1998were identified in a prospective studyinvolving collaboration between ForensicMedicine, Scottish Trauma Audit Group andaccident and emergency (A&E). The circum-stances surrounding each death were exam-ined using police reports held by the Procura-tor Fiscal. Ambulance and hospital recordsprovided information about the treatmentprovided and time of death. Detailed necrop-sies were performed using a standardised

Table 13

Revisedtrauma score Car Ambulance Helicopter

5 0 0 16 0 2 17 0 2 08 6 38 17

Abstracts 449

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

fashion from which a list of all the injuriessustained was prepared, enabling them to bescored using the Abbreviated Injury Scale(1990 revision).

One hundred and fifty nine deaths (122male, 37 female) occurred during the five yearperiod. Various forms of “human error” wereimplicated in causing the majority of deaths,which included combinations of the following:excessively fast driving, driving under theinfluence of alcohol, attempting dangerousovertaking manoeuvres and failure to wear anavailable seat belt. Most of those who diedwere found dead at scene. There were 72 peo-ple with unsurvivable injuries (AIS = 6, InjurySeverity Score = 75) causing immediate deathat scene, which mainly involved injury to thebrain, brainstem, thoracic aorta and upperspinal cord.

The results of this study confirm thecontinuing role of road traYc collisions incausing premature death in south east Scot-land. The predominance of prehospital deathsand the frequency of unsurvivable injuryunderlines the importance of injury preven-tion measures in the future prevention of suchdeaths. Given the background to many of thecollisions it is clear that eVorts need to beaimed at changing the behaviour of both cardrivers and their passengers.

Intravenous fluid use in accident andemergency departments: eVects of pub-lished studies in medical literatureE C CROWHURST, N O’CONNOR, P A EVANS


Background and methods—In view of thecontinuing controversy over the safety ofcolloids and albumin since the meta-analysespublished in the British Medical Journal in1998 we sought to assess prescribing practiceby means of postal questionnaires. These weresent to all accident and emergency (A&E)consultants (459) in the UK and Republic ofIreland around the time of the first meta-analysis with a follow up in 1999. Similarquestionnaires were then sent to all consult-

ants in burns units (205) and intensiverherapy units (ITUs) (1444). In the A&E sur-vey consultants chose the one fluid they woulduse: for burns and ITU, they ranked fluids(1=first choice, etc).Results—Overall response rates: 88% initialand 64% follow up A&E surveys, 55% burnsand ITU surveys. For trauma: A&Econsultants preferred crystalloids: Ringer’slactate (46% adults, 43% children), normalsaline (25% adults, 23% children). In ITUsthe first choice was Gelofusine (mean rank4.6) in adults and normal saline (4.5) inchildren. For burns: In adults: A&E con-sultants preferred crystalloids: Ringer’s lactate(37%), normal saline (25%). For children andall ages at the burns unit: albumin is thepreferred fluid (34% for children at A&E,rank orders 4.0 adult, 4.5 children at burnsunit).Conclusions—Since the meta-analyses consult-ants have changed their practice. Those whohave stopped using colloids: in A&E fortrauma: 11% adults, 13% children. In ITUpractice: 1.5% adults and 1.7% children. Forburns in A&E: 13% adults, 14% children. Atburns units: 7% changed their practice in viewof the colloid meta-analysis. Those who havestopped using albumin: in A&E for trauma:1.7% adults, 7.2% children. In ITU practice:31% adults, 18% children. For burns in A&E:12% adults, 16% children. At burns units:14% changed their practice in view of thealbumin meta-analysis. Thus the publishedmeta-analyses have had a marked impact onintravenous fluid use resulting in a significantreduction in the use of these fluids in clinicalpractice.

Studies on the eVects of resuscitation flu-ids on platelet aggregation in vitroE C CROWHURST, J GLENN, S HEPTINSTALL, P A

EVANS


Continuing controversy exists over fluids usedfor trauma and burns and their haemostaticeVects. In this study, eVects on platelet aggre-gation were determined in whole blood andcompared with results obtained after dilutingblood with autologous plasma. This acted as anon-crystalloid, non-synthetic colloid control.

The fluids or plasma were added to hirudi-nised whole blood from volunteers (n=10) inthe ratio 2:3 (to model the ATLS protocol).Aggregation was measured in response to ADP,collagen, adrenaline and ristocetin. Resultswere compared with those for undiluted bloodand expressed as mean (SEM) percentageaggregation. Ionised Mg2+ and Ca2+ weremeasured using an AVL Analyser (table 14).

Reducing the cell count by adding autolo-gous plasma to whole blood, reduced aggrega-tion in response to collagen and adrenaline, butnot ADP. Saline and Gelofusine enhancedaggregation in response to ADP, collagen andadrenaline, possibly as a consequence oflowered Ca2+. Haemaccel always inhibitedaggregation further, probably through in-creased Ca2+ and, like Gelofusine, abolishedristocetin induced responses. Compared withsaline and Gelofusine, albumin usually limitedenhancement of aggregation consequent tocreating a low Ca2+ environment. Therefore weanticipate that the eVects of these fluids inclinical practice will depend largely on theextent of changes in blood cell counts and therapidity with which homeostatic mechanismscorrect accompanying changes in plasma Ca2+.

Table 14

ADP Collagen Adrenaline Ristocetin Ca2+ Mg2+

Whole blood 58 (9) 91 (1) 38 (5) 28 (9) 1.16 0.46Plasma 55 (9) 57 (12)* 19 (3)* 13 (5) 1.16 0.50Saline 69 (4)+ 80 (7) 35 (6)+ 65 (11)+ 0.77+ 0.30+Haemaccel 32 (7)+ 42 (11) 11 (2)+ −8 (3)+ 2.88+ 0.36+Gelofusine 82 (3)+ 88 (2)+ 45 (8)+ 4 (3) 0.70+ 0.28+Albumin 70 (4)+ 74 (5) 26 (5)+ 38 (11)+ 0.67+ 0.26+

*p<0.05 cf whole blood. +p<0.05 cf dilution with autologous plasma.

450 Abstracts

www.jnlaem.com


http://emj.bm

j.com/

J Accid E

merg M


ovember 2000. D

ownloaded from

http://emj.bmj.com/

the millennium scientiﬁc meeting, manchester, 9–11...

Documents