Search tips
Search criteria 


Logo of emermedjEmergency Medical JournalVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Emerg Med J. 2007 May; 24(5): 375–382.
PMCID: PMC2658509


The College of Emergency Medicine Inaugural Scientific Meeting, 29 November—1 December 2006

Roderick little prize session—Thursday 30 November 2006. 13:45–15:15 (Wise Suite)

001 The PRIMA study: presentation ischaemia modified albumin in the emergency department

E. M. Keating, J. R. Benger, R. Beetham, S. L. Bateman, S. L. Veysey, J. Kendall, R. Pullinger.

ObjectivesTo investigate the diagnostic accuracy of presentation ischaemia modified albumin (IMA), in addition to cardiac Troponin I (TnI), as a strategy to rapidly ascribe low risk to patients with chest pain attending an emergency department (ED), and to determine whether IMA has the potential to reduce ED transit time.

MethodsWe did a prospective observational study within two EDs of a similar size in England. Consecutive adult patients presenting with features of possible ischaemic cardiac chest pain and a normal ECG were eligible. The index test (measurement of IMA and TnI at presentation) and reference standard (delayed TnI measurement, taken at least 8 h after pain onset) were applied to all recruited patients. All clinicians were blinded to the results of the index test. Assays were carried out in a single laboratory using standard techniques.

Results399 patients were recruited. 277 patients had a result for both the index test and reference standard. The sensitivity was 97.6% (95% CI 87.4 to 99.9), negative predictive value 97% (95% CI 84.2 to 99.9) and specificity 13.6% (95% CI 9.5 to 18.7). Sensitivity analysis showed similar findings in three alternative scenarios. ROC analysis indicated that a different “cut‐off” value for IMA would not improve the properties of the test. The median potential time saving (n = 268) was 6 h and 10 min.

ConclusionThe diagnostic accuracy of presentation IMA in this study does not support its use as an effective risk stratification tool for ED chest pain patients. The sensitivity is insufficiently high, with a small number of false negatives undermining the safety of the test. Frequent false positives produce a low specificity that limits practical value.

002 Crossover study comparing a self‐inflating bag‐valve‐mask with the Mapleson C anaesthetic breathing system for emergency pre‐oxygenation

R. Stafford, J. Benger, J. Nolan.

BackgroundPatients undergoing rapid sequence induction of anaesthesia in the emergency department need effective pre‐oxygenation to maximise the time before arterial desaturation occurs, and provide the best chance of successful intubation. A self‐inflating bag‐valve‐mask is often used for pre‐oxygenation, and is both reliable and easy to use, but has theoretical disadvantages when compared to the Mapleson C anaesthetic breathing system.

MethodThis crossover study in healthy volunteers sought to determine the optimum pre‐oxygenation strategy by comparing the efficacy and tolerability of a self‐inflating bag‐valve‐mask with the Mapleson C breathing system. 20 subjects were asked to breathe for 3 min through first one device and then the other, with a 30‐min washout period. The end tidal oxygen concentration and ease of breathing were recorded.

ResultsThere was a statistically significant difference in performance between the two devices, with the Mapleson C providing better pre‐oxygenation as measured by end tidal oxygen concentration at 3 min (mean 74.2% for bag‐valve‐mask (95% CI 72.4–75.9; SD 3.8) vs 86.2% for Mapleson C (95% CI 84.5–88.0; SD 3.7): p<0.0001 on paired t testing). The 95% CI for the difference between the mean values of end expiratory oxygen concentration at 3 min was 10.0%–14.2%. The participants also rated the Mapleson C as significantly easier to breathe through (p = 0.01).

ConclusionThe Mapleson C breathing system provides more effective pre‐oxygenation than a bag‐valve‐mask, and is more readily tolerated. However this must be weighed against the increased level of skill required to effectively use the Mapleson C, its reliance on an oxygen supply and the complications that can occur with this device.

003 Use of a prophylactic antiemetic with morphine in acute pain: randomised controlled trial

E. J. Harrison‐Hansley, M. Bradshaw, A. Sen.

ObjectiveThe aim of this study was to compare the incidence of nausea and vomiting in patients with acute pain treated with prophylactic metoclopramide or placebo.

MethodA randomised controlled trail was carried out on patients requiring morphine for acute pain in the emergency department (ED) setting. Children under the age of 12, patients who had been vomiting or had already received prehospital analgesia, and those unable to give consent were excluded. All patients were given either Metoclopramide (10 mg) or placebo (normal saline) followed by intravenous morphine. Pain scores (measured on a visual analogue scale) before and after morphine administration, all incidents of nausea or vomiting, the dose of morphine and the patients' demographic data were recorded. Fisher's exact test was used for comparing the two groups of patients.

ResultsA total of 259 patients were recruited. There were 123 patients in the metoclopramide group and 136 patients in the placebo group. The overall incidence of nausea or vomiting in the whole study population was 2.7% (1.6 in the metoclopramide group and 3.7% in the placebo group). The difference between the two gropes was not statically significant (Fisher's exact test  = 0.451; p = 0.3; z test static  = 1.02; 95% Cl –6% to 2%)

ConclusionsWhen intravenous morphine is administrated for acute pain, the overall incidence of nausea and vomiting is low, regardless of whether these patients are given prophylactic metoclopramide or not.

004 ACE genotype may predict mortality from major trauma

A. D. Kehoe, K. Eleftheriou, T. Coats, M. Heron.

IntroductionThe endocrine renin‐angiotensin system (RAS) plays a key role in circulatory homeostasis. Angiotensin converting enzyme (ACE) also degrades bradykinin, giving it a role in inflammation, tissue growth and metabolic efficiency. The deletion (D) rather than insertion (I) polymorphism of a 227 base pair fragment of DNA within the ACE gene is associated with higher ACE activity and excess mortality in adult acute lung injury and paediatric meningococcal infection and worse functional outcome from traumatic brain injury.

ObjectiveTo determine if ACE genotype is associated with mortality from major trauma.

MethodForty one subjects were selected from patients with major trauma admitted to the Royal London Hospital (RLH) over 1 year via the London Helicopter Emergency Medical Service (HEMS). ACE genotype was available in 36. The Injury Severity Score (ISS), Revised Trauma Score (RTS) age, sex and outcome data for each subject were obtained through the RLH trauma database. The primary outcome measure was survival to discharge. ACE genotype was determined from leucocyte DNA using well‐described techniques.

ResultsACE genotype predicted mortality. The presence of one or more D alleles was associated with a mortality of 46.4% compared to 7.1% for IIs (p = 0.05). Age (p = 0.02) and RTS (p = 0.01) also predicted mortality whereas ISS did not (p = 0.3). ACE genotype was associated with ISS and RTS but not age. Using logistic regression models to adjust for ISS, the odds ratio of mortality for subjects with the D allele is 8.07 (p = 0.057).

ConclusionThis study suggests an association between ACE genotype and mortality from trauma with the presence of one or more D alleles conferring an increased risk. This finding persists and approaches significance when adjustment for ISS is made. If these findings are replicated in larger studies it may help define high‐risk subgroups that would benefit from early intensive management.

005 South West validation of the Elbow Extension Test (SWEET) trial in adults

A. V. Appelboam, A. Reuben, J. R. Benger, G. Lloyd.

IntroductionThere are no established guidelines as to which patients with elbow injuries require an x ray. Two previous small studies have suggested that the presence of full elbow extension may rule out significant bony injury and has been proposed as a simple test to exclude the need for an x ray. The aim of this study was therefore to assess this test within routine UK emergency department practise.

MethodsWe conducted a multicentre prospective validation study of the elbow extension test in adults. Consecutive patients presenting with elbow injuries who met the inclusion/exclusion criteria were recruited in four centres within the South West of England over a 2‐year period. Sitting patients were asked to actively extend their elbow in the supine position. All patients who could not fully extend were x rayed. Those patients who could fully extend their elbow (compared with uninjured limb) were not x rayed but given simple analgesia, and followed up within 7–10 days. Patients who had continued pain, loss of function or other concern were reviewed and x rayed.

Results1090 patients were recruited. 235 were excluded (injury>72 h, no history of trauma, multitrauma). The remaining 855 patients underwent the elbow extension test and of 287 patients who could fully extend their elbow, 6 patients had a fracture. 219 fractures were identified in the 568 non‐extenders. In this population therefore the elbow extension test had a sensitivity of 97.3% (95% CI 94.3 to 98.9), a specificity of 44.6% (95% CI 40.7 to 48.5) and a negative predictive value of 97.9% (95% CI 95.2 to 99.1).

ConclusionIn this study, the ability to fully extend the elbow following acute elbow injury ruled out significant fracture in over 97% of patients. The elbow extension test may be a useful screening test, particularly in minor injury units where x ray facilities may not be available.

006 South West evaluation of the Elbow Extension Test (SWEET) in children

A. Reuben, A. Appelboam, J. Benger, G. Lloyd.

IntroductionThere are no established guidelines as to which children with elbow injuries require an x ray. A previous study, which included children, suggests that full elbow extension may rule out significant bony injury. Elbow extension has been proposed as a simple test to assess the need for an x ray but has not been fully evaluated in children.

MethodsWe conducted a multicentre, prospective observational study of the elbow extension test in children. Children between the ages of 3 and 15 years with elbow injuries were consecutively recruited in four centres within the South West of England over a 14‐month period. In the test, the seated patient was asked to actively extend their elbow in the supine position. All patients who could not fully extend at first assessment were x rayed. Those children who could fully extend their elbow (compared with uninjured limb) were x rayed at the treating clinician's discretion but all those not x rayed were followed up within 7–10 days. Patients who had continued pain, any loss of function or other concern were reviewed and x rayed.

Results416 patients were recruited. 73 were excluded. The remaining 343 patients underwent the elbow extension test and of 117 patients who could fully extend their elbow, 3 patients had a fracture (1 lat epicondyle, 2 supracondylar) and 4 had an isolated effusion. Of the 226 non‐extenders, 154 had no fracture, 14 had isolated effusions, and 72 patients had a fracture. In this population therefore the elbow extension test had a sensitivity of 96.0% (95% CI 87.9 to 98.9), specificity of 42.5% (95% CI 36.5 to 48.7) and negative predictive value of 97.4% (95% CI 92.1 to 99.3). However, the potential missed fractures were significant injuries.

ConclusionIn this study, the ability to fully extend the elbow following acute elbow injury was not sufficiently sensitive to reliably rule out significant bony injury.

Free paper session 1—Thursday 30 November 2006. 11:15–12:45 (Harris Suite)

007 Chest pain presenting to the emergency department: to risk stratify with GRACE or TIMI?

R. Lyon, A. C. Morris, D. Caesar, S. Gray, A. Gray.

IntroductionThere is a need to rapidly risk stratify patients presenting to the emergency department (ED) with undifferentiated chest pain. The Global Registry of Acute Coronary Events (GRACE) and the Thrombolysis in Myocardial Infarction (TIMI) scoring systems predict outcome of adverse coronary events in patients admitted to specialist cardiac units. This study evaluates the relation between GRACE score and outcome in patients presenting to the ED with undifferentiated chest pain and establishes whether GRACE is preferential to TIMI in risk stratifying patients in the ED setting.

MethodsDescriptive study of a consecutive sample of 1000 ED patients with undifferentiated chest pain presenting to Edinburgh Royal Infirmary, Scotland. GRACE and TIMI scores were calculated for each patient and outcomes noted at 30 days. Outcomes included ST and non‐ST myocardial infarction, cardiac arrest, revascularisation, unstable angina with myocardial damage and all‐cause mortality at 30 days. Score and outcome were compared using receiver operator characteristic curves (AUC‐ROC).

Results: The GRACE score accurately risk stratifies patients presenting to the ED with undifferentiated chest pain (AUC‐ROC 0.80 (95% CI 0.75 to 0.85)). See table. The TIMI score was found to be similarly accurate in risk stratifying the study cohort with an AUC‐ROC of 0.79 (95% CI 0.74 to 0.85). It was only possible to calculate a complete GRACE score in 76% (n = 760) cases as not all the data variables were routinely measured in the ED.

ConclusionsGRACE and TIMI are both effective in accurately risk stratifying patients presenting to the ED with undifferentiated chest pain. The GRACE score is more complex than the TIMI score and in the ED setting TIMI may be the preferred scoring method.

Table thumbnail
Abstract 007

008 Is “see and treat” the solution to long waiting time in the emergency department? A randomised controlled trial

M. Sakr, M. Chaudhry, G. Ward, M. Samadi.

IntroductionThe emergency departments are under enormous pressure to achieve government targets. The NHS plan dictated the 4 h target to be achieved in at least 98% of patients. There is also a strong recommendations to introduce “see and treat” following apparent success in some departments.

AimsTo examine the impact of see and treat on waiting time.

MethodCluster randomised controlled trial comparing Manchester triage, fast track and the see and treat as a model of care. Randomisation was done on a daytime basis including weekends between 09:00–17:00. The trial continued for 12 weeks. Data were collected on waiting times, number of patients breached, and the number of days when the consultant was diverted to minor cases.

ResultsThe trial comprised 7113 patients. The table compare the three groups regarding basic patients' criteria, waiting time, time seen until discharge, total visit time and number of days consultant was diverted to deal with minor injuries.

DiscussionIn our study patients in the Manchester triage type of care spent less time in the department and the total visit time was also shorter. The see and treat model made the waiting time to be seen shorter but the total time was not as short as other models. There were more tendencies to divert the consultant to deal with minor cases when the Manchester triage was in operation.

ConclusionThe Manchester triage model of care tended to cause instability in the department which led to early involvement of the consultant in minor injury cases. This involvement led to more favourable outcomes when the Manchester triage was in operation.

Table thumbnail
Abstract 008

009 Risk stratification of Syncope in the emergency department: the ROSE study pilot

M. J. Reed, A. J. Gray, D. E. Newby, A. J. Coull, K. Jacques, R. J. Prescott.

AimThis study was conducted as a pilot for the ROSE study. It will aim to be the largest international derivation and validation study of undifferentiated syncope based in the emergency department (ED) and the first looking at short‐ and medium‐term outcome.

MethodsNinety nine consecutive adult patients were recruited over a 3‐month period between 7 November 2005 and 7 February 2006. All patients underwent a standardised assessment using 31 historical variables, 28 examination variables and 26 ECG variables. All patients also had a FBC, urea, creatinine and electrolytes, glucose, CRP and near‐patient BNP tests. Admitted patients underwent a formal Troponin I at least 12 h post syncope. Endpoint measures were defined serious outcomes at 1 week, 1 month and 3 months.

ResultsNinety nine patients were enrolled. 32 patients were high risk, 51 medium and 16 low according to ESC guidelines. 72 patients had BNP measured, 25 (35%) were 100 pg/ml or greater, and 3 were over 1000 pg/ml. 44 patients were admitted to hospital and 55 were discharged from the ED. 30 of those admitted had troponin I measured. Only one of these was raised (14.40 ng/ml) and this was thought to be due to an AMI. 66 patients had CRP measured, 16 were raised (>5 mg/l). Five patients had died by 3 months, and 2 others had a serious outcome (1 AMI, 1 complete heart block requiring pacemaker). All 7 had been admitted and all were high risk. Causes of death were lung CA at 6 days, unknown causes at 6 days, 32 days and 79 days and a leaking AAA repair at 80 days. All patients with BNP levels over 1000 had a serious outcome (2 died, 1 AMI).

ConclusionsThe ESC syncope guidelines are good at predicting serious outcome in syncope patients. A BNP level over 1000 may have excellent positive predictive value and specificity for serious outcome. Troponin may be worth measuring in adult patients presenting to the ED with syncope. Further work to determine the value of troponin and BNP in syncope is merited.

010 Randomised controlled trial of mechanical support treatment of severe ankle sprains

M. Cooke, J. Hutton, J. Marsh, R. Nakash, R. Jarvis and S. Lamb on behalf of the CAST ankle team.

BackgroundCochrane reviews highlight a lack of good quality evidence for managing severe ankle sprains.

ObjectivesTo estimate the clinical and cost‐effectiveness of below knee cast, Aircast ankle brace and Bledsoe boot in comparison to Tubigrip.

DesignA randomised controlled trial, reflecting a model of practice used in the majority of UK hospital emergency departments (EDs). Patients aged 16 or over with acute severe ankle sprain, unable to weight‐bear, with no fracture were recruited from eight UK EDs. All patients had two days' initial rest and elevation pre‐randomisation. Below knee cast was applied for 10 days; other supports were worn for as long as the patient felt that they were needed. 584 patients were recruited with follow‐up by postal questionnaire at 4 weeks, 12 weeks and 9 months.

Main outcome measuresPrimary outcome was recovery of pre‐injury function, and secondary outcomes included recovery of normal occupation and avoidance of residual symptoms, and need for further treatment. Disease‐specific measures (Foot and Ankle Outcome Score (FAOS) and Functional Limitations Profile (FLP) and generic measures (SF‐12 and EQ‐5D).

ResultsThe below knee cast gave significantly better results for pain (difference from Tubigrip 7.7 points (95% confidence interval (CI) 2.37 to 13.03)), other symptoms (difference from Tubigrip 6.89 points (95% CI 1.72 to 12.06)) and activities of daily living (ADL) (difference from Tubigrip 4.47 points (95% CI 0.75 to 8.19)). Aircast brace and Bledsoe boot were not significantly better than Tubigrip. Differences persisted at 12 weeks, and by 9 months there were no significant differences.

ConclusionsA short period of immobilisation in below knee cast, applied only after the initial swelling has had time to subside, produces better short‐term results in terms of return to pre‐injury function than Tubigrip. Functional treatment using either Aircast brace or Bledsoe boot produces broadly similar outcomes, and a result slightly better than Tubigrip. Overall, Tubigrip is the least effective treatment. The below knee cast is also the most cost‐effective treatment in terms of improvements in patient quality of life. Acceptability for below knee cast was variable.

Collaborative Ankle Support Trial (CAST) ISRCTN37807450. This study was funded by the NHS Health Technology Assessment Programme.

011 Emergency department overcrowding: evaluation of the impact of a triage liaison physician role

B. R. Holyrod, M. Bullard, P. Moon, D. Gordon, K. Latoszek, S. Allen, S. Blitz, B. H. Rowe.

ObjectiveDespite opinions that suggest some interventions to reduce emergency department (ED) overcrowding are effective, the current evidence base is limited. The use of a triage liaison physician (TLP) has been identified as a potentially effective intervention. The aim of this study was to evaluate the implementation of a TLP shift at an academic tertiary care ED.

MethodsThe project was conducted between 21/11/2005 and 09/02/2006 where physicians worked 9 h (11:00–20:00), performed triage and patient care when needed, assisted triage RNs, answered incoming calls, and managed administrative matters. Within each of three 2‐week blocks; 7 days were randomised to a TLP or control (C) shifts. Outcomes included length of stay (LOS) times in the ED, proportion of patients who left without being seen (LWBS), and ambulance diversion. Median and interquartile ranges (IQR) are reported for continuous data. Proportions were compared using odds ratios (OR) and 95% confidence intervals (CI) or χ2; continuous variables were analysed using generalised linear models.

ResultsPatient volume, sex, age, triage levels, mode of arrival and disposition were similar between the TLP and C shifts. TLP physicians assessed a median of 14 patients (IQR: 13, 17), received 15 telephone calls (IQR: 14, 20), and spent from 17–81 minutes on the telephone. Overall, the TLP reduced LWBS by 25% (5.7% vs 7.24%; p = 0.02) and reduced overall ED LOS by 36 min (4:21 vs 4:57). LOS reduction varied from 2 min (CTAS level 2) to 53 min (CTAS 1); however, variability was not directly related to CTAS level. In addition to TLP coverage status, LOS was influenced by sex, triage level, disposition and age.

ConclusionsOur results indicate that a TLP physician provides valuable services to an overcrowded ED and that the implementation of this type of shift can reduce LOS and LWBS statistics in an ED.

Free paper session 2—Thursday 30 November 2006. 13:45–15:15 (Harris Suite)

012 Wavelet transform based ECG analysis is the most accurate predictor of defibrillation outcome in ventricular fibrillation

G. R. Clegg, J. N. Watson, P. S. Addison, P. A. Steen, C. E. Robertson.

BackgroundRecent studies show that CPR administered before attempted defibrillation in out‐of‐hospital cardiac arrest can improve the rate of return of spontaneous circulation (ROSC) and survival to hospital discharge. This benefit is only seen in patients whose downtime exceeds 4–5 min. There is a need to identify patients who will benefit from a “CPR‐first” treatment strategy rather than immediate defibrillation. We report on a method for the prediction of successful defibrillation, which shows greater accuracy than existing ECG analysis techniques.

MethodsThe wavelet transform is a particularly valuable analytical tool because of its ability to elucidate spectral and temporal information from a signal simultaneously. Here we derive a wavelet transform‐based marker (COP: Cardioversion Outcome Prediction) from the modulus‐maxima of the wavelet transform scalogram. We compare COP to other well‐documented shock outcome predictors: median frequency (MF) of fibrillation, spectral energy (SE) and AMSA (AMplitude Spectrum Analysis).

ResultsOptimum specificities for sensitivities around 95% for the four reported methods are: 63% (SD 4%) at 97% (SD 2%) (COP), 42% (SD 15%) at 90% (SD 7%) (MF), 12% (SD 3%) at 94% (SD 5%) (SE), and 56% (SD 5%) at 94% (SD 5%) (AMSA) with successful defibrillation defined as the rapid (<60s) return of sustained (>30s) spontaneous circulation. This markedly better performance by COP at specificity values around 95%, (required for implementation of the technique in practice) is a result of enhanced partitioning of pertinent information in the time‐frequency plane.

ConclusionCOP identifies patients for whom attempted defibrillation would be futile, and for whom an alternative therapy (eg, CPR‐first) should be considered. COP is methodologically distinct from other Fourier‐based analytical approaches (eg, MF, SE or AMSA) and delivers better results when compared directly with these techniques.

013 Validation of a transcutaneous CO2 monitor in the emergency department

R. Eager, J. McVicar.

IntroductionA number of transcutaneous CO2 monitors have been trialled over the last decade with variable results. The recent development of a monitor via an ear clip has shown useful benefits in the controlled environments of intensive care units and sleep labs. However, use within the emergency department has not been studied to date.

MethodologyWe recruited 51 patients within the emergency department about to undergo arterial blood gas sampling. After informed consent, patients were connected to the monitor and the CO2 level on the monitor was read at the same time as the blood gas sample was taken. A number of clinical parameters were noted. The correlation coefficient and the limits of agreement of the two measurements were determined using SPSS (v13).

ResultsPatients were recruited with a broad spectrum of clinical problems. The mean PCO2 was 5.5 KPa (2.27–9.43). The mean difference between measurements was 0.02KPa (95% CI −0.11 to 0.15 KPa). The Pearson's correlation coefficient was 0.94 (p<0.001). No relation could be identified between the difference in measurement and other patient variables (age, PR, RR, temp, BP, haemoglobin level, arterial pH). The figure shows the limits of agreement (−0.88 to 0.92KPa) for the two measurements by the Bland–Altman method.

figure em47217.f1
Abstract 013.

ConclusionsAlthough the mean difference was small and the correlation coefficient large, the limits of agreement show that the monitor would not be accurate enough to replace CO2 measurement by blood gas analysis. However, as this monitor allows for continuous CO2 measurement, the limits would suffice for the use of the machine to monitor trends in the arterial CO2 in selected patients. This study has a larger number of patients but also the widest limits of agreement than any previous study. Although we could identify no relationship, there may be factors in our patients which adversely affect the accuracy of the monitor.

014 The structure, processes and initial outcomes of chest pain units established in the ESCAPE multicentre trial

S. Goodacre on behalf of the ESCAPE Research Team.

BackgroundThe ESCAPE (Effectiveness and Safety of Chest pain Assessment to Prevent Emergency admission) trial is a multicentre evaluation of the role of chest pain units (CPUs) in the NHS. Seven out of 14 participating hospitals have been randomly allocated to set up a CPU. We describe the structure, processes and initial outcomes at the seven intervention sites.

MethodsEach hospital set up a CPU based upon standardised protocols to rule out acute coronary syndrome (ACS) in low‐risk patients and provide exercise treadmill testing to assess risk of coronary heart disease. Details of all patients managed on the CPU were recorded over the first year of operation. Patients were then followed up for 30 days to identify any adverse events (defined as re‐admission to hospital for more than 48 h, deaths or non‐fatal myocardial infarction).

ResultsThere was substantial variation in the location, staffing and operating hours of the CPUs. Activity also varied substantially between the CPUs, from less than 100 patients managed over the year to over 500. To date, 1573 patients have been managed on the CPUs, with 1303 (83%) being discharged after assessment. There were 23 (1.8%) adverse events among those discharged.

ConclusionsCPUs have been successfully established at all seven intervention hospitals in the ESCAPE trial, but with substantial variation in activity levels. Most patients are discharged after assessment with a low rate of adverse events. CPUs appear to be safe and feasible at a variety of NHS hospitals.

015 The role of systemic warming of “at risk” surgical patients during the initial hospital phase

K. S. Satheesan, D. Whetter, A. Melling, D. Emerton, D. Leaper.

IntroductionThe physiological effects of perioperative hypothermia cause significant morbidity and mortality. Conversely, maintenance of normothermia is beneficial, by improving perfusion, tissue oxygenation and immunological responses.

MethodsAdults presenting to the Accident and Emergency (A&E) department with abdominal pain, as elderly fallers (>65 years old) or with a suspected fractured neck of femur are randomised to: (1) intervention group (IG) with systemic warming using a resistive carbon polymer mattress/overblanket (Inditherm PLC) set at 40°C; (2) control group (CG) in which patients receive standard treatment only. Core temperature, pain, anxiety and comfort scores are recorded before and after warming using visual analogue scales (0–100) with demographic, clinical and physiological variables. 30‐day mortality and morbidity, and length of hospital stay are recorded by a trained blinded observer.

ResultsTwenty three patients were randomised (2 rejected, 11 IG, 10 CG). Age, sex ratio, BMI, transport methods were comparable. The were no difference in the core temperature, pain score, anxiety score and comfort score on arrival to A&E between IG and CG. Pain and anxiety improved significantly in the IG during the first hour in A&E. On A&E departure core temperature, pain, anxiety and comfort all showed a significant difference between IG and CG (Mann–Whitney test). 60% (CG) and 73% (IG) were hypothermic on arrival; 7% (CG) and none from IG was hypothermic on departure. 44% of CG (none from IG) were hypothermic following their transfer from A&E to the admission ward. There were 2 infectious complications and 2 non‐infectious complications all from the CG. There were no deaths

ConclusionsThese early results suggest that systemic warming of “at risk” surgical patients during their initial admission phase increases their core temperature and prevents hypothermia. It may also contribute towards greater clinical gains particularly the reduction of pain, anxiety, discomfort and infectious complications.

016 Public perceptions and experiences of heart attack, cardiac arrest and CPR

R. Donohoe, K. Haefeli, F. Moore.

IntroductionCoronary heart disease is a leading cause of mortality and a common cause of out‐of‐hospital cardiac arrest. The London Ambulance Service NHS Trust attends around 10000 cardiac arrests each year. Of the cardiac arrests considered viable for resuscitation during 2004/5, 79% were of presumed cardiac aetiology. The overall rate of survival for this group of patients was 4%. Although bystander cardiopulmonary resuscitation (CPR) can significantly improve chances of surviving out‐of‐hospital cardiac arrest, it was initiated in just 30% of cases.

AimsTo investigate public knowledge of cardiac arrest and myocardial infarction, explore perceptions of survival, ascertain levels of CPR training, and investigate barriers to CPR.

MethodsAn interview survey was conducted with 1011 Greater London residents. Eight focus groups were also conducted to explore the issues in greater depth and validate trends from the survey.

ResultsThe majority of respondents overestimated the proportion of patients who survive out‐of‐hospital cardiac arrest in London. Few had received CPR training and most were hesitant about performing the procedure on strangers. Barriers to CPR included fear of harming patients and concerns about contracting diseases. Perceived willingness to perform CPR in emergency situations was greater when the procedure did not involve mouth‐to‐mouth resuscitation.

LimitationsAs the research was limited to Greater London residents it may be useful to replicate the research in other regions, both within the UK and abroad, to validate the results of the study.

ConclusionsKnowledge of CPR and reactions to cardiac emergencies reflect relatively low levels of CPR training in London. Publicising cardiac arrest survival figures may be instrumental in prompting members of the public to train in CPR and motivate those who have been trained to intervene in cardiac emergencies. Concerns about mouth‐to‐mouth resuscitation also support the need for further research into the effectiveness of chest compression‐only CPR.

017 Cardiac rhythm analysis during resuscitation without stopping CPR: a wavelet‐based methodology

G. R. Clegg, J. Watons, P. Addison, N. Uschaipichat, P. Steen, C. Roberton.

BackgroundCessation of cardiopulmonary resuscitation (CPR) for even a short period of time (eg, to analyse the ECG tracing and deliver a shock) markedly reduces the likelihood of successful cardiac resuscitation. We have developed a wavelet‐based ECG analysis tool for the interrogation of the underlying cardiac rhythm during CPR. Use of this approach could significantly reduce “hands‐off” periods during resuscitation with subsequent improvement in the rate of return of spontaneous circulation after countershock.

MethodsPartitioning the time frequency plane of the ECG signal allows the differentiation of CPR components from the underlying cardiac rhythm. We extracted segments of CPR from an ECG data set derived from 110 patients and containing 878 shocks. The pre‐ and post‐CPR rhythms were identified and wavelet markers derived from the data. These were then tested for correlations within the CPR segments to identify those which could identify the underlying cardiac rhythm.

ResultsUsing a wavelet‐based technique we found enhanced differentiation of the underlying rhythm signals during CPR when compared to other signal processing methodologies.

ConclusionWavelet‐based analysis of the ECG during CPR holds tangible advantages over conventional techniques (eg, Fourier and adaptive filtering) allowing effective partitioning of the signal component information. Our technique enables the robust differentiation of rhythm components during CPR thus allowing the underlying rhythm to be categorised as shockable or non‐shockable. This methodology could also be employed to predict the outcome of defibrillation therapy and indicate the efficacy of the ongoing CPR.

Free paper session 3—Thursday 30 November 2006. 15:45–17:00 (Harris Suite)

018 Comparison of paediatric major incident primary triage tools

L. Wallis, S. Carley.

ObjectivesTo determine the sensitivity and specificity of paediatric major incident triage scores. The Paediatric Triage Tape (PTT), Careflight, jumpSTART and START systems were tested.

Methods3461 children presenting to a South African emergency department with trauma were scored using the four different methods. The sensitivity and specificity of the four scores was calculated against the Injury Severity Score (ISS), New Injury Severity Score (NISS) and a modification of the Garner criteria (a measure of need for urgent clinical intervention). We also performed a Bayesian analysis of the scores against three different types of major incident. The primary outcome of each score was determined against its ability to determine whether a patient was a priority 1 casualty or not.

ResultsNone of the tools showed high sensitivity and specificity. Overall the Careflight score had the best performance in terms of sensitivity and specificity. The performance of the PTT was very similar. In contrast the jumpSTART and START scores had very low sensitivities which meant that they failed to identify patients with serious injury and in the models of major incidents would have missed the majority of seriously injured casualties.

Table thumbnail
Abstract 018 Accuracy scores for each triage method in three hypothetical incidents.

ConclusionThe Careflight or PTT methods of triage should be used in paediatric major incidents in preference to the jumpSTART or START methods.

019 CRASH2 trial: the first 1000 patients

T. Coats, B. Hunt, H. Shakur, I. Roberts on behalf of the CRASH2 collaborators.

The CRASH2 trial is a randomised controlled trial of an antifibrinolytic versus placebo in haemorrhage following injury. The trial centres are distributed worldwide, so that the answer will be generally applicable to trauma care. The intervention is cheap, and therefore is potentially applicable even in low income countries (the source of much of the current pandemic on death and disability due to injury). The trial entry criteria are pragmatic, in that patients are eligible if the clinician feels that the patient is experiencing, or is at risk of experiencing, significant haemorrhage. For the sample size calculation estimates have to be made of the event rate of the primary outcome measure; however, until patients are randomised it is not know if these estimates are correct. This abstract presents the data on event rates and characteristics of the first 1000 patients randomised in the CRASH2 trial. The primary outcome measure is death, with an estimated rate of 20% being used for the sample size calculation. The actual death rate in the first 1000 patients is 20%. The most important secondary outcome measure is blood transfusion, where an initial estimate was made that 75% of the patients would need blood. The actual percentage of patients transfused is 60%. Other characteristics of the randomised patients are that they are mainly young (68% <35 years), and mainly male (90%). Encouragingly most of the patients (70%) are randomised within 3 h of injury. These results confirm that the trial is recruiting the correct type of patients and is on course to recruit its total of 20 000 patients by December 2009. Any additional collaborators are welcome at

020 ABC is dead, long live <C>ABC: training with novel haemostatic techniques in military emergency medicine

LtCol R. J. Russell, LtCol P. Mahoney, ColT. J. Hodgetts, LtCol M. Russell.

Uncontrolled external haemorrhage is the most likely cause of prehospital avoidable death in combat and security operations. Bleeding from limb wounds accounts for over half the potentially preventable deaths in combat and 10% of all battlefield deaths. Every serviceman is taught the basic principles of haemorrhage control: pressure, apply a field dressing, elevation. But for catastrophic haemorrhage this will not be enough. A comprehensive review, led by the Academic Department of Military Emergency Medicine (ADMEM), has led to the introduction of new haemostatic products and algorithms:

  • Elasticated Field Dressing
  • Combat Application Tourniquet
  • Haemostatic Agents—Quikclot and Hemcon

To ensure a safe transition to the new techniques, ADMEM developed a training package, which was assessed by the use of pre‐ and post‐course questionnaires.

MethodsAll personnel attending the course completed questionnaires before the start and at its completion. The pre‐course questionnaire asked about previous training and practice and confidence with the old equipment. The post‐course questions sought opinions on the effectiveness of the new products, the training and whether there was any change in the level of confidence when dealing with severe haemorrhage. The results were analysed using McNemar's test with Yates continuity correction.

ResultsNinety seven pre‐course and 86 post‐course questionnaires were completed. Candidates agreed strongly that the new techniques and equipment were improvements. Confidence levels when dealing with a traumatic amputation and a deep ragged groin wound both increased significantly (p<0.001).

DiscussionThe results show that the new techniques, products and training package achieved high degrees of approval. The results of this project have led to a change in the resuscitation paradigm used by the Defence Medical Services. Control of catastrophic external haemorrhage now takes the highest priority. ABC is dead—long live <C>ABC.

021 Do ECGs of discharged patients need to be reviewed by senior medical staff in emergency departments? A three month prospective audit

S. Boyce, K. J. Thomson, P. T. Gran.

BackgroundECGs from all patients discharged from the emergency department (ED) are reviewed by senior medical staff in the department with the relevant ED notes 2–4 days after discharge. Looking through these takes time and removes a senior member of staff from clinical duties.

ObjectiveTo determine if the practice of senior medical staff reviewing ECGs of patients discharged from ED highlighted any significant clinical problems.

MethodA prospective audit over a 3‐month period carried out by two specialist registrars in the ED of a city centre teaching hospital. All ECGs from discharged patients were reviewed and the following data recorded: nature of ECG abnormality, specialty and grade of doctor attending the patient, abnormality noted by the attending doctor and any clinical action taken by senior medical staff.

Results1067 ECGs were reviewed over the 3‐month period. 101 ECGs (9.5%) were considered to be abnormal. All abnormalities were noted in the ED notes by the attending doctor irrespective of specialty or grade. No clinical action was required by senior medical staff after reviewing the ECGs. The ECG abnormalities noted are shown in table 1.

ConclusionThe practice of senior medical staff reviewing ECGs of patients discharged from the ED yielded no significant clinical problems and can be safely discontinued.

Table thumbnail
Abstract 021 Table 1 ECG abnormalities
Table thumbnail
Abstract 021 Table 2 ECG abnormalities correlated with specialty and grade of doctor

022 Intravenous ketamine for adult emergency department procedural sedation

L. Chan, C. Reid.

BackgroundKetamine is a sedative agent producing a unique dissociative state. It is widely used in developing countries, in prehospital situations and for paediatric procedural sedation in emergency departments. Use for procedural sedation and analgesia in the adult emergency department population may be limited by fear of the potential side effects.

ObjectiveTo describe our experience using ketamine for procedural sedation in adults in the emergency department, the patient demographics, indications for use, success of these procedures and complications rate.

MethodsProspectively completed sedation charts and case notes of patients receiving intravenous ketamine for procedural sedation that were aged 16 years and over were examined and patient demographics, diagnosis, drug doses and method of administration and complication rates were described.

Results131 patients received ketamine for procedural sedation. Of these 111 case records and sedation charts were located and scrutinised. 48.6% of patients were female. The age range of our patients was 16–91 years old (mean 47.7, median 48.5). Procedures using ketamine were predominantly fracture and joint reductions (91%). The other procedures were for tube thoracostomy (2.7%), cardioversion (0.9%), wound debridement (1.8%), application of splint (0.9%), suturing and examination under sedation (0.9%). Complications were documented in seven cases (6.3%).

ConclusionsKetamine is a safe and effective agent for procedural sedation of adult patients in the emergency department. It has a low associated complication rate and is well tolerated in the elderly. Concerns over emergence phenomenon, respiratory, cardiovascular and cerebral complications have not been borne out by this study.

023 Emergency physicians can reliably assess patient cardiac output using the non‐invasive Ultrasonic Cardiac Output Monitor

J. M. Y. Lam, C. O. Tang, M. K. Tam, C. A. Graham, T. H. Rainer.

AimsCardiac output is important for diagnosing shock and monitoring the response to therapy. The Ultrasonic Cardiac Output Monitor (USCOM) is a new, non‐invasive, stand‐alone device for measuring cardiac output using continuous wave Doppler technology. The aims of this prospective study are firstly to determine the number of proctored studies necessary for skill acquisition, and secondly to evaluate the reliability of USCOM cardiac output assessments in the emergency department.

MethodsTwo emergency physicians from a teaching hospital emergency department underwent one hour of training from an USCOM Ltd credentialed trainer. Each independently performed USCOM cardiac output assessments on a convenience sample of 15 emergency department patients. Records of the acoustic image were reviewed by two independent, accredited trainers using the Freemantle Protocol, a validated six image‐scoring criteria, as the gold standard. Skill acquisition was assessed at the 5th, 10th and 15th examinations. Two trained emergency physicians then each performed blind protocolised examinations on 44 emergency department patients. Inter‐assessor reliability was evaluated.

ResultsDuring the training period, the average image score improved from 3.5/6 at the 5th assessment to 4.5/6 at the 10th assessment to 6/6 (maximum possible) at the 15th assessment. Subsequent analysis of 88 cardiac output assessments in 44 emergency department patients demonstrated excellent inter‐assessor correlation (r = 0.93, 95% CI (0.88–0.96).

ConclusionEmergency physicians can be trained to obtain reliable cardiac output estimations on emergency department patients with the USCOM over the course of 15 patient assessments.

Free paper session 4—Friday 1 December. 11:30–13:00 (Wise Suite)

024 Validation of serum S‐100B as a predictive marker for mild head injury outcome

W. Townend, K. Muller, K. Waterloo, T. Ingebrigtsen, F. Lecky.

IntroductionThe search for objective indicators of brain injury severity after head trauma has included serum biomarkers. One such marker is protein S‐100B. It has been shown to relate to measures of outcome and inventories of symptoms after both mild and severe injuries, but not to measures of neuropsychological performance. Studies of S‐100B to date have used cut‐offs derived from the dataset on which they are applied to describe diagnostic properties. Such an approach will likely yield biased results. This current study applies cut‐off points from previously published data to a cohort of mildly head injured recruited in two centres.

MethodPatients with mild head injury (GCS 13–14 and 15 with altered consciousness at the time of injury) were prospectively recruited in two teaching hospitals—Hope Hospital, Salford, and the University Hospital of Northern Norway. Patients underwent blood testing for S‐100B within 12 h of injury and were followed up at 6 months using the Extended Glasgow Outcome Score (GOSE). Patients in Salford were also followed up at one month.

Results158 patients were recruited in 12 months (90 in Salford). Follow‐up rates were 50/90 (56%) at one month and 88/158 (56%) at six months. At one month, 15/50 (30%) and at six months 10/88 (11%), were disabled by their injuries. An S‐100B concentration of 0.32 μg/l identified the most severely disabled patient at one month—GOSE 4 (sensitivity 100%, specificity 8%, area under ROC curve 0.94). There was no correlation between S‐100B and GOSE score at six months. The 95% confidence intervals for the area under the ROC curve for S‐100B and outcome score at 6 months crossed 0.5.

ConclusionS‐100B does not perform as well diagnostically when previously derived cut‐off points are applied. It may still be a sensitive test for disabling symptoms in the early post‐injury period.

025 Exposure or expert? Setting standards for postgraduate education through a Delphi technique

S. Carley, J. Shacklady, P. Driscoll, D. Kilroy, M. Davis.

ObjectivesThe aim of this study was to determine if a syllabus of upper limb anatomy for emergency clinicians derived by an expert panel reflected clinical practice as experienced by those at whom the syllabus would be directed.

MethodsA three‐round Delphi was performed using an expert group. We compared this to a measure of the exposure to anatomical concepts in the day‐to‐day practice of trainees in emergency medicine. Round one consisted of a comprehensive list (404 items) of anatomical structures and relationships for the upper limb. Participants were asked to score these items on a three‐part scale: (1) essential to know; (2) useful to know; (3) not useful in clinical practice. Having defined the opinions of the experts we invited a group of 35 emergency medicine trainees to grade the same anatomical structures and relationships using the following criteria: (1) almost never used in my clinical practice (eg, once a year or less); (2) occasionally used in my clinical practice (eg, once a month); (3) regularly used in my clinical practice (eg, weekly).

Results404 separate anatomical facts relating to the upper limb were reviewed by both groups. There was poor agreement (κ = 0.348) between the expert group and the trainees. The table shows the results of the expert group in terms of final modal scores.

Table thumbnail
Abstract 025

ConclusionWe have shown disparity between what an expert group believes trainees should know and what trainees are actually exposed to in clinical practice. We believe this demonstrates that curriculum development must strike a balance between important (expert) information and common (exposure) information. We have shown how an expert:exposure matrix may be used to inform curriculum development.

026 Rapid sequence induction by emergency physicians reduces the time to intubation in critically ill emergency patients

P. R. Atkinson, A. A. Boyle, V. Ahmed.

BackgroundCritically ill patients not in cardiac arrest at emergency departments are traditionally intubated by anaesthetic staff who may work in areas remote from the emergency department. We measured time to intubation before and after the introduction of emergency physician performed rapid sequence induction.

SettingAddenbrooke's Hospital emergency department, UK seeing 70 000 patients/year.

MethodsBefore and after observational study. Detailed and accurate timing of the time of arrival, the decision to intubate and the time of intubation were recorded over two separate fortnight periods in 2001 and 2005. In 2001 all the intubations were performed by anaesthetists, in 2005 all the intubations were performed by emergency physicians.

ResultsDuring the first fortnight, in 2001, 13 patients were intubated with mean time between decision to intubate and actual intubation of 26 min. During the second fortnight, in 2005, 17 patients were intubated with a mean delay of 10 min. This difference was statistically significant (p = 0.015) There was complete agreement about the need for intubation between the anaesthetists and the emergency physicians (κ = 1.0) There were no failed intubations in either group.

InterpretationRapid sequence induction performed by emergency physicians results in quicker time to intubation in critically ill patients. Whether this results in improved outcomes is not proven by this study.

027 Can UK emergency physicians safely scan for abdominal aortic aneurysms?

B. M. Dent, R. J. Kendall, P. R. T. Atkinson.

IntroductionInternationally, studies have shown that emergency physicians can accurately perform aortic ultrasound scans with relatively little training. Within a tight clinical governance framework, focused aortic ultrasound was introduced into the practice of a UK emergency department. We review 12 months of practice.

MethodAll patients who underwent a focused aortic ultrasound by an emergency physician during a 12‐month period from January to December 2005 were included. The patient details were obtained from the departmental ultrasound case log. Outcome data were obtained from paper and electronic patient records. The follow‐up period ranged from 5 to 17 months.

ResultsOver the 12‐month period 122 scans were performed by 18 individuals. Of these, 25 scans (20%) were positive for an abdominal aortic aneurysm (AAA); (diameter >3 cm). In 15 cases this represented a new diagnosis. Ruptured aneurysms represented 48% (12/25) of all positive scans of which four underwent emergency repair. In the remaining 13 patients the AAA was an incidental finding. Of the 97 patients with a normal emergency department scan, 38 subsequently underwent further imaging of their aorta. In this group no aneurysms were detected within 12 months. In one case a 4.3 cm AAA was detected at 14 months follow‐up. This shows a sensitivity of 96% and specificity of 100% for detection of an AAA in patients with follow‐up. In the 59 patients who did not undergo further imaging, none had died from a ruptured AAA at follow‐up.

ConclusionsOur results suggest that UK emergency physicians can accurately and safely undertake focused abdominal aortic ultrasound, identifying those patients with an AAA. We have devised a flow chart incorporating ultrasound for the suggested management of patients presenting with a suspected AAA.

028 Paediatric coma guideline: a consensus‐based pathway for the management of a child with a reduced level of consciousness

C. Downing.

ObjectiveThe causes of coma in children are legion but most are rare. Junior medical and nursing staff frequently provide front‐line treatment. We produced a clinical algorithm for the management of a child who presents acutely to hospital with a reduced level of consciousness.

OptionsAssessment, immediate treatment and investigations are examined.

OutcomesFour principles underpin the guideline. First, use the ABCD approach. Second, recognise and treat important underlying causes early. Third, have a low threshold for giving early intravenous antibiotics. Fourth, form a definitive diagnosis systematically.

EvidenceThe algorithm is based on statements of recommendation produced by the Delphi process.

ValuesThe Delphi process represents expert consensus of medical and nursing opinion from various UK departments of emergency medicine and paediatrics. This places value on overall safety and on identifying early even rare causes of a reduced level of consciousness.

Benefits, Harms and CostsWe anticipate the introduction of local guidelines that reflect expert consensus opinion, standardising medical care and providing a benchmark for audit. A potential disadvantage is that safe guidelines may encourage over‐invasive investigation and/or intervention. Financial costs and measures of clinical effectiveness have not been considered.

Recommendations in briefFor each child presenting with a reduced level of consciousness, treat the greatest threat to life first by following ABCD. Perform blood glucose, capillary blood gases, and urine stix immediately, and send venous blood and (non‐sterile) urine to the laboratory. Consider Cefotaxime 80 mg/kg iv/io stat. If no obvious cause is apparent after 1 h then treat for infection, arrange head CT, and consider rare causes.

ValidationThe guideline has undergone implementation in an emergency department and a paediatric assessment unit. It remains to be validated formally.

Sponsors: None.

029 Impact of LUCAS device use on arterial blood gas results during human CPR

A. Kidd, C. E. Robertson, A. Oglesby, M. McKechnie.

Patients and MethodsPatients with out‐of‐hospital cardiac arrest treated with the LUCAS device in the resuscitation room at Edinburgh Royal Infirmary were studied prospectively. The ambulance crews delivered conventional CPR with manual chest compressions before arrival in hospital. On arrival the LUCAS device provided active chest compression/decompression CPR at a rate of 100/min. Ventilation was provided via a tracheal tube, using conventional IPPV and an FiO2 of 1.0. Blood gas measurements were taken during the resuscitation process.

ResultsTwenty eight patients were studied. The mean time between ambulance call and commencement of LUCAS CPR in the resuscitation room was 40 min (range 10–71 min). Mean PaO2 values after 5 min of LUCAS CPR were 27.6 kPa (range 8.4–70.9 kPa). Mean PaCO2 value was 5.88 kPa (range 1.3–14.1 kPa). PaO2 and PaCO2 levels were maintained for up to 20 min during the resuscitation attempt. Hydrogen ion concentrations slowly increased.

ConclusionsManual chest compression is commonly performed suboptimally. We report, for the first time, blood gas values in patients in whom resuscitation was performed using the LUCAS device. Our results are commensurate with those which we have previously reported using mechanical CPR with the Thumper device. The implications for use of the LUCAS device are discussed.

Articles from Emergency Medicine Journal : EMJ are provided here courtesy of BMJ Group