2007 CAEP/ACMU Scientific Abstracts - Oral Presentations: 1-30

2007 Scientific Abstracts

CJEM 2007;9(3):183-226

Abstracts: 1-30, 31-44, 46-48, 100-130, 131-159, 161-189, 191-216

Oral Presentations

1 BETA-BLOCKERS post acute myocardial infarction: systematic review and meta-analysis

Al-Reesi A, Al-Zadjali N, Fergusson D, Stiell IG, Al-Thagafi M, Al-Shamsi M. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: Acute myocardial infarction (AMI) remains a major cause of death. We wished to determine the mortality benefit of very early intervention of beta blockers compared to control group in patients surviving AMI. Methods: We conducted a systematic review of randomized controlled clinical trials that assessed short term mortality and compared beta-blockers to a control group in patients within 72 hours post AMI. We searched these databases: MEDLINE (1966-2006), EMBASE (1980-2006), Cochrane Central Register Of Controlled Trials, Health Star (1966-2006),Cochrane Database for Systematic Reviews, ACP Journal Club (1991-2006), Database of Abstracts of Reviews Of Effect (<1st quarter 2006), and Conference Papers Index (1984-2006). Two blinded reviewers extracted the data and assessed 4 quality components (randomization, double blinding, reporting of withdrawals and allocation concealment). We calculated Pooled odds ratios using a random effect model, and performed sensitivity analyses to explore the stability of the overall treatment effect. Results: We included 18 studies (13 met the high quality score) which enrolled 74,643 participants and had 5,095 deaths. Compared to control group, beta-blockers did not result in a statistically significant decrease in short term mortality when added to other interventions in the acute phase post AMI (OR=0.95, 95% CI 0.90 to 1.01). Six weeks OR for mortality including studies with high quality score was 0.96 (95% CI 0.91 to 1.02). We found similar results including studies which enrolled patients within 24 hours post AMI only. A subgroup analysis, excluding high risk patients with Killip class II and above, showed beta-blockers resulted in a significant reduction in short term mortality with OR=0.93 (95% CI 0.88 to 0.99). Conclusions: Acute intervention with beta-blockers does not result in a statistically significant short term survival benefit post AMI. Beta-blockers should be optional in AMI patients who do not have signs of heart failure. Key Words: Myocardial infarction, Beta blockers

2 TELEVISION public service announcements improve bystander CPR rates

Vaillancourt C, Hale K, Stiell IG, Wells GA. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: Television public service announcements (PSA) are commonly used to promote CPR classes to a large number of viewers at once, but they may lack the educational content to actually teach CPR to viewers. This study prospectively evaluated the impact of CPR education PSAs on bystander CPR rates for out-of-hospital cardiac arrest victims. Methods: This before-after controlled trial took place over 3 phases in a city with population 750,000, EMS with ALS and BLS-D paramedics, and dispatch-assisted CPR instructions. We collected patient and system data for out-of-hospital cardiac arrest using the Utstein style. We measured bystander CPR rates for 6 months before and 6 months after a 3-month period of CPR education PSAs. The 30-sec PSAs were aired 174 times by two television networks during prime time, each PSA reaching audiences up to 167,000. In addition to promoting CPR classes, our 30-sec PSAs demonstrated when to call 9-1-1 and how to perform CPR on a cardiac arrest victim. Analyses included Chi-Square, logistic regression, and descriptive statistics with 95% CIs. Results: Patients in phase 1 n=108, 2 n=52, and 3 n=104 had similar characteristics: mean age 67.7 (range 30-97), male 71.2%, witnessed arrest 48.9%, initial rhythm asystole 50.2%, VF/VT 32.3%, PEA 17.5%, 9-1-1 call to vehicle at scene 5min:34sec, survival to hospital discharge 4.9%. Bystander CPR rate was 14.6% during the 12-month control period compared to 28.9% during the 3-month PSAs intervention (Chi-Square 5.9, p=0.02). We adjusted the effect of PSAs on bystander CPR rates for the victim's age, gender, witnessed status, initial rhythm, and EMS time intervals. The adjusted Odds Ratio was 2.9, 95%CI 1.2-7.0, p=0.02, Goodness-of-Fit=0.62. We were underpowered to study the effect of PSAs on survival: OR 0.3, 95%CI 0.04-2.6. Conclusions: We urge all cities to introduce CPR education PSAs as means to improving bystander CPR rates. The impact of PSAs on survival to hospital discharge requires further studies. Key Words: Cardiac arrest, CPR, Education, Public health

3 THE Ottawa aggressive protocol for ED management of acute atrial fibrillation

Stiell IG, Clement CM, Dickinson G, Symington C, Perry JJ, Vaillancourt C. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: There is no consensus as to the optimal emergency department (ED) management of acute atrial fibrillation (AAF) or atrial flutter (AAFL). Our objective was to examine the efficacy and safety of the Ottawa Aggressive Protocol to convert and discharge ED patients with AAF/AAFL. Methods: This 5-year cohort study included consecutive visits to a university hospital ED for adults presenting with acute-onset AAF/AAFL and who were managed with the Ottawa Aggressive Protocol. Patients were identified from the National Ambulatory Care Reporting System (NACRS) database. The Aggressive Protocol was overseen by the attending emergency physicians and included: 1) IV procainamide as infusion of 1 gram over 1 hour; 2) electrical cardioversion if necessary, by ED staff; 3) discharge from the ED with outpatient cardiology follow-up. Outcomes included conversion, adverse events, and relapse. We conducted descriptive data analyses with 95% CIs. Results: Characteristics of the 660 eligible patient visits were mean age 64.5 years, mean heart rate 113.4, mean duration symptoms 8.9 hours, AAF 95.2%, AAFL 4.9%. Overall, 96.8% of patients were discharged home from the ED and 90.3% were discharged in normal sinus rhythm. The respective discharge rates were 97.0% and 93.5% for those in AAF and 93.8% and 87.5% for those in AAFL. All patients received procainamide with a conversion rate of 58.3% (AAF 59.9%, AAFL 28.1%). Electrical cardioversion was attempted in 36.8% of visits with a success rate of 91.7% (AAF 91.0%, AAFL 100%). Adverse events occurred in 7.6% of cases: hypotension 6.7%, bradycardia 0.3%, AAF relapse within 7 days 8.6%, Torsades de Pointes 0%, cerebrovascular accident 0%, mortality 0%. Conclusions: This is the largest reported study of AAF/AAFL in the ED and demonstrates that the Ottawa Aggressive Protocol is extremely effective for the rapid cardioversion and discharge of patients by ED physicians. This protocol is safe and could lead to a significant decrease in hospital admissions. Key Words: Atrial fibrillation, Atrial flutter, Cardioversion, Procainamide

4 FATIGUE and quality of CPR by older bystanders using the new 30:2 chest compression to ventilation guidelines: a randomized cross-over trial

Midzic I, Vaillancourt C, Taljaard M, Chisamore B. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: The 2005 International Consensus on CPR changed the chest compression:ventilation ratio from 15:2 to 30:2 to minimize interruptions and increase the number of compressions. We sought to measure bystander fatigue and CPR quality after 5 minutes of CPR using the new and the old CPR ratios in a population of older lay persons. Methods: This randomized cross-over study took place at a senior's center and a tertiary care hospital. Participants were aged 55 or greater with no significant physical limitation (Frailty score ≤3/7). Participants completed two 5-min CPR sessions (using 30:2 and 15:2 ratios) on a recording manikin, separated by a 5-min rest. We used concealed-blocked randomization to determine the order of ratio. Metronome feedback maintained a compression rate of 100/min. We measured changes in heart rate (HR), mean arterial pressure (MAP), venous lactate (VL), and Borg Exertion Scale (range 6-20). CPR quality measures were number of chest compressions and number of good compressions (fully-released, depth≥38mm). Analyses included Paired t-Test, mixed-effect regression, and descriptive statistics with 95% CIs. Results: The 42 enrolled participants were: mean age 66.0 (range 55-84), female 69.0%, past CPR training 66.7%, and mean initial HR 70.4, MAP 91.3, VL 1.5, and Borg score 9. Bystander fatigue was similar for each CPR ratio: mean difference between groups in increased HR 1.5 (95% CI -1.5-4.5), MAP 1.5 (-1.8-4.8), VL 0.2 (-1.1-1.4), Borg 0.2 (-0.2-0.8). Participants attempted more chest compressions per session using the 30:2 vs the 15:2 ratio (382.2 vs 303.6, mean diff. 78.6; p<0.0001), but completed a similar number of good compressions (128.5 vs 126.6, mean diff. 2.0; p=0.85). The number of good compressions/min declined significantly more rapidly over time for the 30:2 ratio; p=0.02. Conclusions: In a population likely to perform CPR, the new 30:2 ratio resulted in significantly more rapid CPR quality decline and no additional benefit over the old ratio. Key Words: CPR, Public health

5 PROSPECTIVE multicenter study of admissions to Canadian hospitals for acute COPD

Rowe BH, Willis V, Stiell I, Young B, Stenstrom R, Campbell S, Abu-Laban R, Akhmetshin E. Department of Emergency Medicine, University of Alberta, Edmonton, AB

Introduction: Hospitalization rates after ED treatment of acute COPD vary across jurisdictions, and the limited previous research has involved health systems with universal access and high use of preventive medications. Our objective was to determine the factors associated with hospitalization after ED treatment for COPD in Canada. Methods: 19 Canadian EDs enrolled patients (pts) over the study period. Enrolled pts underwent a structured ED interview and telephone interview 2 weeks later. Inclusion criteria were MD diagnosis of COPD, age >35, and no evidence of asthma. Admission was defined as an acute visit that resulted in a formal admission to that hospital. Data were analyzed using chi-square, t-test, M-W test, and logistic regression. Results: Of 501 subjects, 242 (48%) were admitted to the hospital; site admission proportions ranged from 20-80%. Pts who were admitted did not differ from those discharged in age or gender. Longer ED length of stay (LOS) increased the chances of hospital admission (p<.001) and LOS for admitted patients varied (median: 5.85 days; IQR: 3.97, 8.50). Significant predictors of admission in multivariate testing were ED usual site for problem COPD care (OR=0.33; 95% CI: 0.12-0.95), ever taken corticosteroid for COPD (OR=0.26; 95% CI: 0.08-0.9), diabetes (OR=4.2; 95% CI: 1.05-16.92), activity limitations in past 2 weeks (OR=1.92, 95% CI: 1.17-3.16), pCO2, mmHg (OR=1.1; 95% CI: 1.03 1.170) and ED length of stay (OR=1.07 per hour; 95% CI: 1.02-1.13). Conclusions: Admission to these Canadian hospitals for acute COPD from the ED varies dramatically. COPD admissions contribute to ED overcrowding through event frequency and prolonged in-ED care. Admission is associated with a variety of pre-treatment factors, and efforts to reduce admissions seem warranted. Key Words: COPD, Overcrowding

6 WHICH alarms first during procedural sedation: the pulse oximeter or the capnograph?

Messenger DW, Sivilotti MLA, VanVlymen J, Dungey PE, Murray HE. Department of Emergency Medicine, Queen's University, Kingston, ON

Introduction: The role of end-tidal CO2 (ETCO2) monitoring during procedural sedation and analgesia (PSA) in the emergency department (ED) is unclear. Many EDs lack such monitors. Proponents claim capnography detects respiratory depression before pulse oximetry. We tested this claim during a randomized clinical trial of low-dose ketamine versus fentanyl during ED PSA with propofol. Methods: A planned secondary analysis of a trial involving titrated propofol for orthopedic reduction or abscess drainage in ASA class I or II adults. All subjects were monitored by continuous pulse oximetry and combined oral/nasal sampled CO2 (Smart Capnoline Oral Nasal Cannula, Nellcor, Pleasanton, (CA) via a Lifepak 12 (Medtronic, Redmond, (WA). By protocol, supplemental oxygen was withheld unless hypoxemia (SaO2 < 92%) developed. Hypercapnia was defined as a 10 mmHg rise in ETCO2, an absolute ETCO2 >50 mmHg, or a 30s loss of waveform. Results: Of 63 subjects enrolled (31 male, age 39±18 yr), 36 developed hypoxemia and 17 hypercapnia (12 due to loss of waveform). In 12 of these 17 hypercapnic patients, hypoxemia detected by pulse oximetry preceded hypercapnia by a median of 187s. Of the remaining 5, 3 had isolated loss of waveform, and none received specific intervention. All hypoxemic patients responded quickly to supplemental oxygen, interruption of propofol administration, and/or stimulation, with only 1 patient requiring brief bag-valve-mask ventilation. Satisfaction with level of sedation was very high among both subjects and physicians. Conclusions: As ED PSA becomes widely accepted and practiced, safety concerns remain paramount. In selected, low-risk patients breathing room air, oxygen desaturation usually precedes changes in capnography and is readily reversible. The use of continuous capnography during PSA with propofol did not affect the management of these patients. These conclusions should not be extrapolated to patients pretreated with supplemental oxygen, or other settings/monitoring configurations. Key Words: Procedural sedation, Capnography, Oximetry, Propofol

7 CHALLENGING the dogma: topical proparacaine is safe and effective for the outpatient management of acute traumatic corneal injuries

Ball IM, Desai N, Seabrook J, Allan L, Anderson A. Division of Emergency Medicine and Department of Surgery, University of Western Ontario, London, ON

Introduction: Acute traumatic corneal injuries cause significant patient discomfort. The dogma forbids prolonged use of topical anesthetic agents in this patient population. There is little or no evidence for such an approach. Recent studies in the ophthalmology literature show that dilute topical Proparacaine is safe and effective as an analgesic when used in the post-operative setting. Our goal was to evaluate the safety and efficacy of dilute topical Proparacaine when prescribed in the emergency department for outpatient management of acute traumatic corneal injuries. Methods: A convenience sample of patients seen in two tertiary care emergency departments were randomized by a computer generated table of random numbers to either 0.05% Proparacaine drops or to placebo. Emergency department physicians, patients and the ophthalmologist were blinded to treatment allocation. Patients were instructed to use 1-2 drops PRN for 5 days. They recorded their pain on a visual analog scale 5 minutes prior to and 5 minutes after "study drug" administration. All patients were provided with topical antibiotics. They were also prescribed oral Acetaminophen with Codeine for breakthrough pain. Patients were followed in an ophthalmology outpatient clinic on days 1, 3 and 5 post-injury to ensure adequate wound healing. The study was powered to detect a 20% improvement in analgesia. Our primary endpoint was mean pain reduction. Secondary endpoints included patient satisfaction and wound complications. Results: Thirty-four patients were randomized and completed the trial. The results were analyzed with the Mann-Whitney U Test. There was a statistically significant improvement in both pain reduction 3.5 vs 1.4 (P= 0.007) and patient satisfaction 7.2 vs 3.8 (P= 0.027) in the treatment group versus the control group. There were no wound complications in any patient studied. Conclusions: Dilute topical Proparacaine is safe and effective for outpatient analgesia of acute traumatic corneal injuries when prescribed in the emergency department. Key Words: Corneal abrasion, Analgesia, Proparacaine

8 YOUTH violence secondary prevention initiatives in emergency departments: a systematic review

Snider C, Lee J. Division of Emergency Medicine, University of Toronto, Toronto, ON

Introduction: Youth violence continues to trouble Canadians. We hypothesized that effective secondary violence prevention interventions exist for injured youth in the emergency department (ED), and conducted a systematic review to identify 1) outcome measures used for ED based interventions, and 2) success rates of interventions. Methods: Eight databases (MEDLINE, EMBASE, PUBMED, CINAHL, COCH, ACP, DARE and CENTRAL) were searched. Search terms included combinations and synonyms of "Youth", "Violence", "Interventions" and "Emergency" or "Hospital". Studies were included if they: 1) described and evaluated an intervention, 2) were healthcare-based and 3) targeted youth injured by violence. Results: Thirteen articles were selected by two blinded investigators from 171 abstracts. After full text review, 5 articles with 3 intervention programs were included in the review. Eight articles were excluded: adults only (n=1); primary preventions only, (n=6); and no evaluations performed (n=1). All interventions included involved case management programs in the ED with the violently-injured patient. A randomized control trial demonstrated a significant reduction in re-injury rates (cases 8.1% vs. controls 20.3%, Chi2 = 3.87, p=0.05). A retrospective cas-condemonstrated a 70% relative risk reduction in re-arrest. A retrospective observational study enrolled 72% of potential pediatric patients injured due to violence with only 1% of these youth returning for injuries due to violence (vs. historically reported 44% recidivism rates). Conclusions: All three studies reviewed were positive, although they had limited sample size and lacked intention to treat analyses. These consistently promising preliminary findings support continued development and investigation of ED-based programs to reduce the terrible toll of youth violence. Key Words: Youth, Violence, Injury prevention

9 THE acceptability of clinical decision rules: validation of the Ottawa acceptability of decision rules scale (OADRS)

Brehaut JC, Graham I, Eagles D, Stiell IG. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: To develop a scale measuring the acceptability of a clinical decision rule (CDR). Development of high-quality CDRs is extremely resource-intensive, with no guarantee of widespread use. A tool to examine acceptability of early-stage CDRs would be useful. Methods: A 12-item scale (6-point Likert responses) was developed, pilot-tested, and included in larger surveys of practicing emergency physicians from four countries (AUS, CAN, UK, US; total n = 1290), for two CDRs: the Canadian C-Spine Rule (CS) and the Canadian CT-Head Rule (CT). In addition to typical item and internal consistency analyses, construct validity was measured in part by testing 4 hypotheses: 1) Scale scores should vary predictably by country, the highest scores coming from Canada; 2) CDR users should give higher scores than non-users; 3) Among users, OADRS scores should be higher for consistent users; 4) Among non-users, scores should be higher for those who would consider using the rules in the future. Results: Internal consistency was high, ranging from 0.79 to 0.83 for CS, and from 0.81 to 0.87 for CT. Multi-factor ANOVAs with scale scores as the outcome showed 1) Canada scored highest on overall scale scores (range 0-5; Mean scores: AUS 3.36, CAN 3.91, UK 3.52, US 3.38; p < .001); 2) Scores were higher among users than non-users (CS means: 3.83 vs. 2.96, p<.001; CT means 3.77 vs. 3.01, p <.001); 3) Scores were higher among consistent users (CS means: Always 4.30; Most of the time 3.93, Some of the time 3.23; CT means: Always 4.34; Most of the time 3.92, Some of the time 3.36; p < .001 in both cases); and 4) Among non-users, scores were higher for those who would consider use (CS means: Yes 3.40; No 2.57; CT means: Yes 3.45; No 2.64, p < .001). Interactions were generally small or non-significant. Conclusions: The OADRS might serve as an "early-warning system' for CDR producers wanting to know whether a CDR will be considered acceptable by the target audience. Key Words: Decision rules, Head injury, Cervical spine injury

10 THE epidemiology of suicide post emergency department visit

Grafstein E, Stenstrom R, Harris D, Hunte G, Innes G. Department of Emergency Medicine, Providence Health Care & St. Paul's Hospital, Vancouver, BC

Introduction: Existing studies on suicide have a lack of specificity for emergency health care providers because they focus on data from all patient populations rather than on data for emergency department visits. The study objective was to characterize the epidemiology of suicide in patients presenting to the ED. Methods: A retrospective study in an urban, academic emergency department with an annual census of +60,000 visits. We reviewed all patients presenting to the ED between January 1, 2000 and November 1, 2005. Using the emergency administrative database, we linked with provincial vital statistics data and the regional emergency visit database from the same time period to ascertain which patients presenting to the ED had committed suicide and whether or not there were other related patient encounters at other local emergency departments. Results: There were 130,619 patients who made 300,625 visits to the ED. 8327 patients had 15,412 visits (5.1%) related to mental health. There were 127 (0.97%) deaths attributable to suicide in patients with previous ED visits. There were 35 suicides in 3,501 patients (5,547 visits) with a presenting complaint of suicidal ideation or attempt (1%). 93 of the 127 (73.2%) were male. Average age for this cohort was 42.0 (SD 14.0). Only 74 (58.2%) of these deaths had a presenting complaint or ED discharge diagnosis related to mental health and/or substance misuse including alcohol. 36 of the 127 patients made visits to other regional hospitals. 42 of 127 (33.7%) patients committed suicide within 30 days of an ED encounter and 55 (43.3%) within 60 days. 42 of 127 (33.1%) had a previous psychiatric related admission and 38 (30%) had only a single ED encounter with no other regional ED visits. Conclusions: This represents one of the first reports of suicide in an emergency department population. Although the overall percentage of patients committing suicide is low, there is still about one patient a month committing suicide within 60 days of an ED visit. Key Words: Suicide, Population health

11 IDENTIFICATION and classification of drug-related problems in the emergency department

Zed PJ. Departments of Pharmacy and Emergency Medicine, Queen Elizabeth II Health Sciences Centre; Capital District Health; College of Pharmacy and Department of Emergency Medicine, Dalhousie University, Halifax, NS

Introduction: Clinical pharmacy services and the practice of pharmaceutical care is directed at identifying and resolving drug-related problems (DRPs) to achieve optimal patient outcomes. The Emergency Department (ED) is an area with a patient population at risk for many DRPs both prior to presentation and within the ED. Our objective was to determine the incidence and type of DRPs identified in the ED of a large tertiary-care, Canadian teaching hospital which provides clinical pharmacy services, and to describe the proportion of DRPs in the ED relative to the remainder of the hospital. Methods: A prospective, observational study was conducted on a population of adults presenting to hospital over an 18-month period. Clinical pharmacists utilized a PDA-based tool for documenting DRPs and pharmacist interventions during routine patient care practice and documentation was based on the pharmaceutical care model. DRPs were categorized according to one of eight established categories as defined by Hepler and Strand. Primary outcomes were reported as proportions presented as percentages with 95% confidence intervals. Results: 39,874 DRPs were documented throughout the institution over the 18-month evaluation period. The intensive care unit identified the most DRPs at 20.7% (95%CI 20.3-21.1%) followed by the bone marrow transplant service at 17.3% (95%CI 16.9-17.6%) and ED at 16.5% (95%CI 16.1-16.8%). The most common types of DRPs in the ED were need to add drug 53.3% (95%CI 52.0-54.5%), unnecessary drug 10.1% (95%CI 9.4-10.8%), high dose 9.3% (95%CI 8.6-10.0%) and adverse drug reaction 7.2% (95%CI 6.6-7.9%), wrong/suboptimal drug 6.6% (95%CI 6.0-7.2), low dose 6.5% (95%CI 6.0-7.2%) and non-compliance 3.2% (95%CI 2.8-3.7%). Conclusions: Pharmacist identified DRPs are common in the ED with a volume of DRPs comparable to other priority areas for clinical hospital pharmacists. Hospitals without clinical pharmacy services in the ED should consider implementation to address this high risk patient population. Key Words: Pharmacy, Pharmacist, Adverse drug effects, Drug related problem, Hospital care

12 NATIONAL survey of CCFP(EM) program directors regarding emergency medicine ultrasonography

Woo MY, Nussbaum C, Lee AC. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: Emergency medicine ultrasonography (EMUS) is used by emergency physicians in Canada. The purpose of this study was to survey CCFP(EM) program directors regarding current and future EMUS training. Methods: We conducted a web-based survey of all CCFP(EM) residency training directors in Canada using a modified Dillman technique. Two academic emergency physicians reviewed the validity and reliability of the survey. The survey was pilot tested by another academic CCFP(EM) residency trained emergency physician. Research ethics approval was obtained. The web based questionnaire consisted of 17 yes/no and open questions designed to assess perceptions regarding current and future EMUS education. Descriptive statistics with 95% CIs were used. Results: The survey response rate was 100% (18/18). Currently 84.2% of respondents use EMUS in their program. Of those, 68.8% offer an introductory EMUS course of which 54.6% are mandatory. All programs recommend attending an introductory EMUS course. 37.5% of the programs have over 90% of their attending staff using EMUS, while 43.8% of programs have less than 30% of staff using EMUS. 77.8% of directors think that an introductory course in EMUS should be mandatory. 64.7% believe that residents are able to acquire sufficient experience to use EMUS independently to make practice decisions prior to the completion of their residency. 88.9% believe that EMUS should be a part of the scope of practice for emergency physicians. Only 61.1% believed that questions about EMUS should be on the CCFP(EM) exam. Open responses indicated that funding, resources, and standardization were issues that needed to be addressed. Conclusions: Formal EMUS training for CCFP(EM) programs is being introduced in Canada. The majority of directors think that an introductory course in EMUS should be mandatory. However, fewer directors believe EMUS should be on the CCFP(EM) certification exam until further funding, resources, and standardization of EMUS programs are in place. Key Words: Ultrasonography, Education and training, Curriculum

13 DEVELOPMENT of an emergency based curriculum on homelessness: what do medical trainees need to know?

Spence JM, Bandiera G, Hwang SW. St. Michael's Hospital, University of Toronto, Toronto, ON

Introduction: Exposure alone to homelessness may negatively impact attitudes of medical trainees (MTs). Many homeless persons (HPs) will seek routine care in the emergency department (ED) as a result of barriers to access to care. Previous work has shown that HPs perceive attitudes of health care workers to be a significant barrier to care. Objectives: To define and develop an ED-based curriculum in the care of homeless persons. Methods: The study was conducted at an urban Canadian ED (55,000 visits/yr; 15% HPs). 2 focus groups (N=19) were conducted with adult and youth community service providers (CSPs) and 3 (N=15) with medical trainees and ED staff (MT-ED). At the suggestion of focus group participants, further interviews were conducted at drop-in centers and shelters, with HPs(N=6) who seek care from many sources. CSPs (N=5) who could not attend sessions provided written comments. Participants discussed experiences, expectations, and curriculum needs. Focus groups were transcribed verbatim, coded and recoded using an iterative process to define common themes corresponding to key objectives. Results: Codes were grouped as: 1) attitudinal issues, 2) resource/system challenges, 3) curriculum needs. CSPs were concerned about perceptions of pathways to homelessness and issues of substance abuse. Major system issues included ED wait times, lack of follow-up, discharge planning, and feedback to CSPs. Curriculum goals cited lack of general knowledge about homelessness, associated medical and psychosocial problems. MT-ED described addressing homeless issues in the ED as opening "Pandora's Box". Trainees felt ill-equipped and uncomfortable asking social history. MT-ED requested curriculum on continuum of care and discharge planning, and approaches to history and physical. Conclusions: Homeless curriculum needs to include pathways to homelessness, approaches to medical and psychosocial problems, with a greater emphasis on continuum of care and discharge planning. Key Words: Homelessness, Education, Trainees, Curriculum

14 ESSENTIAL dimensions of performance for emergency medicine residents

Blouin D, Dagnone JD. Department of Emergency Medicine, Queen's University, Kingston, ON

Introduction: Personal interviews are regarded as the most important screening tool in resident selection. Interrater reliability is significantly greater for structured interviews than for unstructured ones. The content of structured interviews is most often developed based on performance dimensions felt to be important to succeed in a particular line of work. The dimensions of performance essential to succeed as a resident in Emergency Medicine have not yet been studied. This project aims at determining the dimensions of performance necessary to succeed as a resident in an Emergency Medicine program. Methods: This is a prospective study. A focus group of 5 representative members of the Emergency Medicine department at Queen's University used the "critical incident technique' to generate scenarios of poor and excellent resident performances. The 5 members compose 20% of the department workforce. As recommended in the critical incident literature, 2 faculty members, different than the initial 5, independently reviewed each scenario and labeled the performance dimensions that they felt were reflected in each. These 2 faculty members provided their own labels. All labels assigned to a particular scenario were pooled and reviewed by the 2 faculty members until consensus was reached. For analysis purposes, only labels applying to at least 2 scenarios were considered. Results: The focus group generated 51 critical incidents. Twelve dimensions of performance applied to at least 2 scenarios. The four main dimensions most frequently reflected by the scenarios were (1) professionalism, 24% of scenarios, (2) insight, 11%, (3) humanity, 9%, and (4) reliability, 9%. Conclusions: The main performance dimensions considered essential to succeed as a resident in the Emergency Medicine program at Queen's University are presented. These dimensions reflect the values sought by this specific program. As an initial step in structuring admission interviews, other programs should identify the dimensions of particular importance to them. Key Words: Residency, Selection criteria, Performance

15 DESIGNING advanced EMS subspecialty training: a survey of Canadian EMS directors

Frank JR, Youssef F, Woo MY, Maloney J, Lee AC. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: Prehospital care is an emerging subspecialty for Emergency Medicine. While EMS fellowships are well developed elsewhere, they are nascent in Canada. We conducted a systematic needs assessment survey to define the essential elements for a Canadian EMS subspecialty training curriculum. Methods: We developed a questionnaire from a review of literature and existing EMS program documents. This was validated and pilot tested by EMS experts. Using Dillman methods we conducted a phone and email survey of all Ontario Base Hospital EMS Directors (n=30). Directors were asked to rate essential competencies for Base Hospital physicians from a list of 18 domains, as well as the degree of need for EMS fellowships in Canada. Results were analyzed using descriptive statistics. Results: Response rate was 66.7% (20/30). Ninety-five percent were male, and 50% served small urban environments. Most were certified in EM as CCFP-EM (50%), or FRCPC EM (20%). Only 30% of Directors had formal training for their role. They were responsible for an average of 355 paramedics (87-1000), and 6 supervised aeromedical services. Six domains of competence were rated as specifically important for EMS directors: quality control, human resource management, finance, legal issues, EMS system design, and history of EMS in Canada. Twelve other domains were rated essential for all Emergency physicians: paramedic levels, dispatch and triage, disaster preparedness, WMD, HAZMAT, injury prevention, community education, aeromedical issues, telemedicine, research methods, medical control, and therapies used by EMS. Small urban Directors were less likely to endorse the need for formal EMS fellowships (2.4/5) than urban Directors (3.1/5) (p>0.05). Ninety percent suggested EMS training should be 1 year or less. Conclusions: This needs assessment identifies 6 essential domains of competence exclusive to EMS directors and 12 essential for all EM physicians. These can be used to guide future Canadian EMS fellowship development. Key Words: Prehospital care, Emergency medical services, Administration

16 EXAMINERS' assessments of resident performance on in-training examinations agree moderately with written examination results and with each other

Bandiera GB. Division of Emergency Medicine, University of Toronto, Toronto ON

Introduction: Residents in Canadian residency programs undergo in-training exams to assess progress and prepare them for the certification exam. Little is known about the inter-rater reliability of the current format of oral in-training exam, or about the concurrent validity of examiners' assessments of residents' knowledge as measured by written exam. Program directors are challenged with providing reliable in-training assessments of their residents, while resource constraints favour minimizing resources and faculty time. We sought to determine the degree to which examiners' scores agree with each other and with a concurrent written exam. Methods: Residents in one FRCP(EM) program underwent four sets of in-training examinations over two years. Exams included two oral exams with two examiners each, a bellringer exam, and a three hour short answer question written exam. Each examiner used a 5-point scale to independently assign each resident a separate score for knowledge base, patient management, and examsmanship. Correlations between examiners' assessments (interrater reliability) and between residents' knowledge base marks on the oral exam and their written exam marks were measured using a Pearson correlation coefficient. Examiner's agreement on pass/fail decisions were measured using Cohen's kappa. All exams were standardized and a template was used for marking purposes. Results: Fifty-seven sets of exam scores were reviewed. The average correlation between examiners for all marks was 0.66. The average kappa for pass/fail decisions was 0.61. The average correlation between exam scores and examiners' assessments of knowledge base was 0.43. Conclusions: Examiners demonstrate reasonable agreement on pass/fail decisions, but do not agree closely enough to support the use of only one examiner. Oral and written exams appear to measure different forms of knowledge supporting the continued use of both formats. Sources of variance among scores is worthy of further study. Key Words: Emergency medicine training, Education

17 CLINICAL problems and procedures encountered by fourth year medical students during their emergency medicine clerkship

Penciner R, Siddiqui S, Lee S. Division of Emergency Medicine, University of Toronto, Toronto, ON

Introduction: In order to fulfill the Liaison Committee on Medical Education (LCME) accreditation standards, the emergency medicine (EM) clerkship at the University of Toronto piloted an electronic procedural logging application using handheld computers. The purpose of this study was to determine what clinical problems and procedures were encountered by fourth year medical students during their emergency medicine clerkship rotation. Methods: Fourth year medical students (n=199) at University of Toronto were expected to participate in the electronic logging project during their 4 week EM clerkship between September 2005 to April 2006. Novel software was designed for handheld computers. Following each clinical encounter, students entered data on their handheld computer utilizing check boxes and drop-down menus. Following synchronization of their handheld computers, data was transmitted to a central server. Results: A total of 46 medical students (23%) participated in the pilot electronic logging project. A total of 2930 encounters with a mean 63.7 per student were entered. Students logged an average of 8.3 different patients per shift. The majority of patients encountered (49%) were between the ages of 17-49 years. There was a wide range of patient acuity encountered. The majority of students encountered patients with problems/diagnoses from all major categories. Twenty-eight percent of students encountered patients in the category of resuscitation. The most common problems encountered were lacerations, fractures, sprains, chest pain and abdominal pain. The most common procedures performed were suturing, splinting and slit lamp use. Conclusions: Fourth year medical students during their emergency medicine rotation are encountering a wide variety of clinical problems and performing procedures with large variability as demonstrated by electronic logs. More standardization of their clinical encounters in emergency medicine will need to be addressed in order to meet the LCME accreditation standards. Key Words: Emergency medicine training, Education, Procedures

18 EFFICACY and impact of intravenous morphine before surgical consult in children with right lower quadrant pain suggestive of appendicitis: randomized controlled trial

Bailey B, Bergeron S, Gravel J, Bussières JF, Bensoussan A. Division of Emergency Medicine, CHU Sainte-Justine, Université de Montréal, Montréal, QC

Introduction: The evidence supporting the use of analgesia in children with abdominal pain suggestive of appendicitis is limited. Methods: All children between the age of 8 and 18 years presenting to a paediatric emergency department with a significant right lower quadrant abdominal pain and a presumptive diagnosis of appendicitis were eligible to be enrolled in a randomized double-blind placebo controlled trial if the initial pain was at least 5/10 on a verbal numeric scale. Patients received either 0.1 mg/kg of morphine (maximum 5 mg) or placebo. The primary outcomes were: 1) the difference in pain using a visual analog scale at baseline and 30 minutes after the completion of the intervention, and 2) the time delay between arrival in the emergency department and the final surgical decision (surgery or discharge from surgery care). Results: Ninety patients with a suspected diagnosis of appendicitis were randomized to receive either morphine or placebo. Of these, 58/87 (67%) had a final diagnosis of appendicitis. Both groups were similar in terms of demographic, history, physical findings, probability of appendicitis, and initial pain score. There was no difference in the decrease of pain between the morphine and placebo groups: 24 ± 23 mm and 20 ± 18 mm, respectively (difference 4 mm [95% CI -5, 12]). There was also no difference in the time delay between arrival in the emergency department and the surgical decision: median of 269 minutes (95% CI 240 to 355) for morphine, and of 307 minutes (95% CI 239 to 415) for placebo (difference -34 minutes [95% CI -105 to 40]). Conclusions: The use of morphine in children with right lower quadrant abdominal pain with a presumptive diagnosis of appendicitis did not delay the surgical decision. In our group of patients, however, morphine was not more effective than placebo in diminishing their pain. Key Words: Paediatrics, Abdominal pain, Analgesia, Appendicitis, Morphine

19 EVALUATION of a rapid nucleic acid hybridization test in detecting Group A streptococcal pharyngitis and comparison to throat culture in a pediatric emergency department

Ouellet C, Bailey B, Scrivo C, Laferrière C. Division of Emergency Medicine, Department of Pediatrics, CHU Ste-Justine, Université de Montréal, Montréal, QC

Introduction: The sensitivity of rapid antigen detection test based on latex agglutination or ELISA varies and may be poor in some cases. Newer rapid streptococcal tests using nucleic acid hybridization appear promising and may eventually replace the throat culture. The objective of this study was to compare the validity of the Gen-Probe® Group A Streptococcus Direct Test (GADST) to the throat culture in detecting Group A streptococcal pharyngitis in a pediatric emergency department (PED). Methods: Children (<18 years) presenting to the PED of a tertiary care hospital were eligible to be included in the study if a throat culture was deemed clinically required for pharyngitis from 03/05 to 07/05 and from 10/05 to 06/06. Two throat swabs were collected simultaneously; one for culture and the other for GADST. In case of discrepancy, a Todd-Hewitt broth was performed and the result of this was considered the true result. Sensitivity (Sn), specificity (Sp), positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) and their 95% confidence interval were calculated. Results: A total of 399 children aged 6.3 +/- 5.0 years were included. Clinical symptoms and signs were: sore throat (73%), fever (65%), exudate (33%), adenopathy (51%), and absence of cough (50%). The rate of positive throat culture was 39%. For all children, Sn of GADST in our study was 92% (87, 96); Sp 75% (69, 80); PPV 71% (64, 77); NPV 94% (89, 96); LR+ 3.71 (2.97, 4.64) and LR- 0.10 (0.06, 0.18). After distribution of discordant results, the Sn of the GADST was 99% (96, 99), Sp 83% (78, 88), PPV 81% (75, 86), NPV 99.5% (95% CI 97.0, 99.9), LR+ 6.04 (4.48, 8.15), and LR- 0.008 (0.001, 0.055). Conclusions: GADST in our study did not perform as well as anticipated. The numbers of false positives preclude us from recommending its use instead of standard throat culture within our population. This confirms the need to evaluate the performance of a new diagnostic test in one's setting before adopting it. Key Words: Paediatrics, Emergency medicine, Strep pharynigitis, Diagnostic testing

20 WHAT is the impact of advanced life support on the management and outcomes of out-of-hospital seizures in children?

Osmond MH, Stiell IG, Nesbitt L, Clement CM, Campbell S, Munkley D, Luinstra-Toohey L, Maloney J, Wells GA, for the OPALS Study Group. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: The Ontario Prehospital Advanced Life Support (OPALS) Study is designed to evaluate EMS interventions for critically ill patients. The OPALS Pediatric Study tested the impact on children with out-of-hospital seizures of adding a full ALS program to existing BLS-D EMS systems. Methods: This multicenter before-after controlled clinical trial was conducted in 17 communities and enrolled all children (<16 years old) with out-of-hospital generalized seizures who were seizing during EMS attendance during the 36-month BLS-D phase and the subsequent 36-month ALS phase. Paramedics were fully trained to ALS standards including endotracheal intubation and administration of IV drugs. Chi-square and t-test analyses were performed. Results: There were 417 children enrolled during the BLS-D (N=183) and ALS (N=234) phases. The groups were well matched for gender (52.5% male) but the BLS group was younger (mean age 4.2 vs 5.6 years), had more febrile seizures (35.5% vs 22.2%; P=.02) and less chronic non-febrile seizures (39.3% vs 49.6%; P=.02). During the ALS phase, intubation was attempted for 5 children with 40% success; intravenous access was attempted for 123 (52.6%) children with 65% success; and anticonvulsants were given to 124 (53.0%) children (rectal diazepam 33.8%; IV diazepam 19.2%). No anticonvulsants were used during the BLS-D phase. From the BLS-D to the ALS phase, there was a decrease in number of patients still seizing on arrival to the ED (44.8% vs 32.1, P=0.008) and an increase in the use of bag valve mask ventilation (6.7% vs 19.2%, P=.0002). There was no difference in any other outcome including hospital discharge (98.9 vs 99.1%; P=.85), hospital admission (45.0% vs 42.2%; P=.59), or ICU admission (16.9% vs 11.6%; P=.34). Conclusions: This first controlled trial to evaluate full ALS programs for out-of-hospital seizures in children shows a substantial number of children receiving prehospital anticonvulsants and fewer children still seizing on arrival to the ED. Key Words: Emergency medical services, Paediatrics, Seizures, Advanced life support, Diazepam, Anticonvulsants

21 COMPARISON of two clinical scores in the evaluation of acute asthma in pre-school children

Robidas I, Gouin S, Gravel J, Guimont C, Chaput G, Chalut D, Amre D. Divisions of Pediatric Emergency Medicine & Research Institute, Université de Montréal, Division of Emergency Medicine, Université Laval, Division of Pediatric Emergency Medicine, McGill University, Montréal, QC

Introduction: To evaluate the accuracy and responsiveness of two clinical asthma scores: the Preschool Respiratory Assessment Measure (PRAM) and the Pediatric Asthma Severity Score (PASS), during acute asthma. Methods: A prospective cohort study was conducted in a pediatric Emergency Department (ED). All the patients between the age of 18 months to 7 years who presented for an exacerbation of asthma while one of the research assistants was available (Feb-Dec 06), were approached. The exclusion criteria were: chronic cardio-respiratory problems, clinical diagnosis of bronchiolitis, pneumonia, laryngitis, varicella, whooping cough and transfer of a patient. The variables of interest were: length of stay ≥ 6 hours in the ED or admission, oxygen saturation, physician's judgment of severity, admission rate, utilization of systemic corticosteroid and > 5 bronchodilator treatments. Clinical findings were assessed at the start of the ED visit, after 90 minutes of treatment, and at the time of physician's decision regarding patient disposition. Results: During the study period, 1853 patients were seen in the ED for an asthma exacerbation. Of these patients, 213 were approached and 5 refused to participate. The mean age was 3.3 years (SD 1.4) and 63% were boys. Significant correlations were seen between the physician's judgment of severity and PRAM (r = 0.42-0.65) and PASS (r = 0.43-0.65). Moderate correlations were found between a length of stay ≥ 6 hours or admission and PRAM (AUC 0.69, 0.59-0.79) and PASS (AUC 0.70, 0.60-0.80) calculated at the start of the ED visit. When the scores were recalculated after 90 minutes of treatment, the correlation increased for PRAM (AUC 0.81, 0.73-0.90) but remained similar for PASS (AUC 0.72, 0.62-0.82). The scores were shown to be responsive, with a 26.7% relative improvement in score from start to 90 minutes of treatment for the PRAM and 26.9% for the PASS. Conclusions: The PRAM and PASS scores are valid measures of asthma severity in children and show both discriminative and responsive properties. Key Words: Paediatrics, Asthma, Clinical severity scores

22 CLINICAL and laboratory assessment of dehydration severity in the emergency department: diagnostic test properties

Parkin PC, Macarthur C, Khambalia A, Goldman RD, Friedman JN, Divisions of Pediatric Medicine and Pediatric Emergency Medicine, The Hospital for Sick Children, Department of Pediatrics, University of Toronto, Toronto, ON

Introduction: We have previously described the development of a clinical dehydration scale which includes the following 4 clinical characteristics: general appearance, eyes, mucous membranes, and tears. The objective was to evaluate the diagnostic test properties of the scale and laboratory measures in the ED. Methods: Percent dehydration was calculated (weight change from pre- to post-rehydration) for 93 children (1-36 months) with acute gastroenteritis. The clinical score and laboratory tests (pH and bicarbonate) were obtained. Sensitivity and specificity for each clinical score (0 to 8) was calculated, and receiver operating characteristics (ROC) curves were used to obtain cutoffs for no dehydration (<3%), mild dehydration (>3%) and moderate to severe dehydration (>6%). Likelihood ratios (LR) were calculated for each of the 3 dehydration categories and for laboratory tests for dehydration >6%. Results: ROC curve analysis resulted in the following dehydration severity categories: score = 0 (none); score = 1 to 4 (mild); score = 5 to 8 (moderate to severe). The final diagnostic test properties are shown in the table. Conclusions: The clinical dehydration scale, pH and bicarbonate demonstrate meaningful likelihood ratios for severity of dehydration. Children with a clinical score of 0 may be considered to have no dehydration; children with a clinical score of 1 to 4 may have mild dehydration; children with a clinical score of 5 to 8, pH < 7.32, or bicarbonate < 18 mmol/L may have moderate to severe dehydration. These variables may be added to a decision making algorithm to guide management (oral or intravenous fluid rehydration) and disposition decision (discharge home, short stay unit, hospitalization). Key Words: Paediatrics, Dehydration, Disposition, Gastroenteritis

Table 1, Abstract 22
Variable	N	Severity	Pre- dictor	LR+	LR-
Clinical score = 0	24/93	None	<3%	2.2	0.79
Clinical score = 1 to 4	56/93	Mild	>3% to <6%	1.3	0.67
Clinical score = 5 to 8	13/93	Moderate to severe	>6%	5.2	0.55
pH < 7.32	9/61	Moderate to severe	>6%	7.2	0.48
Bicarbonate < 18 mmol/L	9/61	Moderate to severe	>6%	11.6	0.35

23 EVALUATION of the validity of a computerized version of the Canadian Triage and Acuity Scale in a paediatric emergency department

Gravel J, Gouin S, Bailey B, Roy M, Bergeron S, Amre D. Division of Emergency Medicine, Hôpital Sainte-Justine, Montréal, QC

Introduction: The use of a standardized triage tool like the Paediatric Canadian Triage and Acuity Scale (paedCTAS) allows better comparison of patients. The use of a computerized version of the tool (Staturg from Statdev) could theoretically improve its validity. The objective of the study was to evaluate the inter-rater agreement among nurses using Staturg in comparison to the traditional paedCTAS. Methods: A two phases experimental study was conducted to compare the inter-rater correlation between nurses assigning triage level to written case scenarios using either the traditional paedCTAS or Staturg. Participants were nurses with at least 1 year of experience in paediatric emergency and they had to be trained and evaluated at triage. Each of the 54 scenarios was evaluated by each nurse first using either one of the strategies (CTAS or i-paedCTAS). Two to four weeks later, they evaluated the same scenarios using the other tool (crossover). The primary outcome was the inter-rater correlation measured using the weighted Kappa score. Results: Eighteen of the 29 approached nurses participated in the study. They all evaluated the 54 clinical scenarios. The computerized triage tool showed a better inter-rater correlation (Staturg kappa score of 0.54 (95% CI 0.54-0.56) vs. CTAS Kappa score 0.51 (95% CI 0.49-0.52 )). The computerized version was associated with a better correlation for the scenarios describing patients with the highest severity of triage (kappa 0.72 vs. 0.55 for triage priority 1 and kappa 0.70 vs. 0.50 for triage level 2). Both triage tools had similar accuracies as measured by the agreement between nurses and an expert panel (concordance in 57% for paedCTAS vs. 55% i-paedCTAS). Conclusions: A computerized version of the paedCTAS showed a statistically significant improvement in the inter-rater correlation for nurses evaluating triage level of 54 clinical scenarios but this difference has small clinical significance. Key Words: Paediatrics, Triage, Computerized decision support

24 A cluster randomized knowledge transfer trial in 4,457 minor head injury patients

Stiell IG, Clement CM, Grimshaw J, Brison R, Rowe BH, Schull M, Lee JS, Brehaut JC, Letovsky E, MacPhail I, Shah A, Ross S, McKnight RD, Dreyer J, Edmonds M, Rutledge T, Clarke A, Perry J, Wells GA, for the CCC Study Group. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: The Canadian CT Head Rule (CCHR) for imaging in minor head injury was previously derived (N=3,121) and validated (N=2,707). This knowledge transfer study sought to evaluate the effectiveness and safety of implementing the CCHR in multiple EDs. Methods: This matched-pair cluster randomized trial compared 12-month "before' and "after' periods at 6 "intervention' and 6 "control' EDs, stratified by teaching or community hospital status. We enrolled adults presenting with GCS 13-15 after acute head trauma and transient neurological impairment. We randomly allocated sites to intervention or control groups. During the after period at the intervention sites, we introduced active strategies to implement the CCHR, including education, policy, and CT requisition reminders. Outcomes included CT imaging rates and missed injuries. Chi-square procedures were used for cluster data analysis. Results: We enrolled 4,457 patients with mean age 38.9 years, male 70.5%, brain injury 5.7%, neurosurgical intervention 0.7%. Cases were similar, comparing control (N=1,932) to intervention (N=2,525) sites. Baseline CT rates ranged from 22.5% to 80.0%. From the before to after periods, the CT rate increased at the intervention sites (from 61.4% to 76.2%; P<0.01) and at the control sites (from 67.3% to 74.3%; P<0.01). There were no missed cases of important brain injury. Comparing before to after periods at the intervention and control sites, there was an increase in total minor head injury cases from 1003 to 1522 vs from 872 to 1060 (relative increase 51.7% vs 21.6%; P<0.01). There was also an increase in the total brain injury cases from 50 to 89 vs a decrease from 60 to 53 (78.0% vs -11.7%; P<0.01). Conclusions: This knowledge transfer trial failed to demonstrate an impact of the previously validated CCHR on CT imaging rates. Issues included physician compliance and a change in patient demographics. Future studies should identify implementation barriers and explore strategies to deal with them. Key Words: Head injury, Diagnosis, CT scan, Decision rules, Knowledge transfer

25 IS D-dimer testing appropriately used in the management of emergency department patients with suspected venous thrombo embolisms?

Smith C, Worster A, Mensah A, Mal S. Division of Emergency Medicine Research, McMaster University, Hamilton, ON

Introduction: The d-dimer is the most commonly ordered diagnostic test for suspected pulmonary embolus or deep venous thrombosis. In order for its results to be applied appropriately, a pre-test probability (PTP) must be performed. Without this, the test results can be misinterpreted and lead to unnecessary diagnostic imaging or inappropriate discharge. Therefore, we sought to determine if PTP was documented for emergency department (ED) patients on whom a d-dimer was performed for suspected venous thrombo embolism (VTE). Methods: In this medical record review, we used a random number generator to select 100 of 760 charts of patients at three EDs who had a SimpliRED d-dimer performed during a three-month period. Trained data abstractors, blinded to the study hypothesis, abstracted explicitly defined data from the ED chart. An independent abstractor assessed reliability of 15% of the charts. Results: Suspicion of VTE was documented in 97/100 charts. There was no documentation of PTP assessment for 62/97 (64%) cases. Of 10/62 (16%) without documented PTP and with positive d-dimers, 5/10 (50%) had no diagnostic imaging. Of 52/62 (84%) without documented PTP and with a negative d-dimer, 47/52 (90%) had no imaging. Conclusions: PTP is not documented in a majority of ED patients with suspected VTE. This, and the finding that 50% of patients with a positive D-dimer had no further testing, suggested that D-dimer testing is being used inappropriately. Key Words: D-dimer, DVT, Diagnostic test

26 The prehospital validation of the Canadian C-Spine Rule by paramedics

Vaillancourt C, Stiell IG, Beaudoin T, Maloney J, Anton A, Bradford P, Cain E, Stempien M, Lees M, Munkley D, Battram E, Banek J, Wells GA, for the EMS C-Spine Study Group. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: The Canadian C-Spine Rule (CCR) for radiography in alert and stable trauma patients was previously validated in a cohort of 8,283 patients. The CCR calls for evaluation of active neck rotation if patients have none of 3 high-risk criteria and at least 1 of 5 low-risk criteria. This study prospectively evaluated the accuracy, reliability, and clinical sensibility of the CCR when used by paramedics in the prehospital setting. Methods: This prospective cohort study took place in seven Canadian prehospital Emergency Medical Services and involved alert (GCS 15) and stable adult trauma patients at risk for neck injury. Advanced and basic care paramedics completed 15-item data forms and interpreted the CCR status for all patients who then underwent immobilization and radiography to determine the outcome, clinically important c-spine injury. 149 patients were independently examined by a 2nd paramedic. Patients were followed by a 14-day telephone interview. Analyses included sensitivity, specificity, kappa coefficient, and descriptive statistics with 95% CIs. Results: The 2,397 patients enrolled over 50 months had these characteristics: mean age 40.7 (range 16-103), female 51.7%, motor vehicle collision 62.7%, fall 19.6%, admitted to the hospital 9.9%, clinically important c-spine injury 0.5%, unimportant injury 0.2%, internal fixation 0.3%. The CCR classified patients for 12 important injuries with sensitivity 100% (95% CI 74-100), specificity 38.2% (36-40). The kappa value for paramedic interpretation of the CCR was 0.94 (0.96-0.98). Paramedics misinterpreted the rule in 6.0% of cases, did not evaluate range of motion when indicated in 3.3%, were comfortable applying the rule in 81.5%. 916 prehospital immobilizations could have been avoided using the CCR. Conclusions: Paramedics can apply the CCR with accuracy and identified all important cervical-spine injuries. The adoption of the CCR by paramedics could have a significant impact on the number of prehospital immobilizations. Key Words: Canadian C-Spine rule, EMS, Cervical-spine

27 CAN ED triage nurses reliably clear the c-spine in minor trauma?

Stiell IG, Clement CM, O'Connor A, Davies B, Leclair C, Mackenzie T, Beland C, Peck T, Sheehan P, Gee A, Perry J. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: Having ED triage nurses clinically clear the c-spine of minor trauma patients would avoid prolonged and unnecessary immobilizations. This validation study evaluated the reliability of ED nurse neck assessments using the Canadian C-Spine Rule (CCR). Methods: We conducted a prospective cohort study at 6 EDs (2 teaching and 4 community hospitals) and enrolled alert and stable, adult, trauma patients who presented with neck pain or on an ambulance backboard. Triage nurses were trained to use the CCR by means of a 1-hour interactive CD and 1-hour practical session and underwent a written test. Nurses evaluated each patient for the CCR, including assessment of neck tenderness and rotation (where applicable) and recorded findings on a data form. A second clinician (nurse or MD) performed interobserver assessments of the patients independently. Data analyses included the kappa coefficient and 95% CIs, with kappa > 0.60 considered to indicate substantial agreement. Results: Overall, 112 nurses enrolled 345 patients with mean age 42.4 years (range 16-100), female 60.3%, ambulance arrival 73.0%, MVC 59.4%, admitted 8.1%, c-spine injury 1.2%. According to the CCR interpretations, 47.5% of patients could have had their c-spine cleared. Interobserver assessments (61.7% by MDs) for overall need for immobilization by CCR showed simple agreement of 90.5% and a kappa of 0.81 (95% CI 0.74-0.88). Results did not change regardless of whether the second observer was a nurse or MD. The agreement for assessing the 9 CCR component findings had these kappas:

Age 65 - 0.97
Dangerous mechanism - 0.79
Paresthesias - 0.71
Rearend MVC - 0.76
Upright position - 0.78
Ambulatory - 0.71
Delayed neck pain - 0.66
Midline tenderness - 0.54
Able to rotate - 0.81

Conclusions: This highly successful validation study found that nurses can reliably assess the CCR and its components. Future studies should test the effectiveness and safety of a strategy of having triage nurses clear the c-spine of low-risk trauma patients. Key Words: Canadian C-Spine rule, Nurse, Cervical-spine

28 SENSITIVITY and specificity of the Canadian CT head rule and the New Orleans criteria in a US trauma center

Papa L, Stiell IG, Clement CM, Rees E, Clapp A, Light J, Ferguson K, Goldfeder B, Meurer D, Stair R, Luetke C, McCormick S. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: CT ordering rate for minor head injury in the US is high. This study compared the clinical performance of both Canadian CT Head Rule (CCHR) and New Orleans Criteria (NOC) decision rules for detecting the need for neurosurgical intervention and clinically important brain injury in a set of patients in the US. Methods: Prospective cohort study in a single US university Level I trauma center comparing the CCHR and NOC among minor head injury patients who presented to the ED with blunt trauma to the head resulting in witnessed loss of consciousness, disorientation or definite amnesia, and a GCS score of 15. The CCHR was also evaluated in all patients with GCS 13-15. The primary outcome measures were "need for neurosurgical intervention" and "clinically important brain injury" on CT. Data were analyzed using appropriate univariate measures as well as sensitivity and specificity with 95% CIs. Results: Among 314 patients with GCS 15, 3 (0.9%) required neurosurgical intervention and 11 (3.5%) had clinically important brain injury. As compared with the NOC, the CCHR had the same sensitivity (100 vs 100%; 95% CI 29-100%) but was more specific (80.7% vs 9.6%) for predicting need for neurosurgical intervention. For clinically important brain injury, the CCHR had similar sensitivity to the NOC (100 vs 100%; 95% CI 72-100%) but was more specific (34% vs 9.9%), and would result in lower CT rates (66.2% vs 90.4%). Among all 413 patients with GCS 13-15 the sensitivity and specificity of CCHR for neurosurgical intervention for 5 cases was 100% and 66.7% respectively and for 27 cases of clinically important brain injury it was 100% and 28.2% respectively. Conclusions: In a US sample of minor head injury patients with GCS 15, the CCHR and the NOC have equivalent high sensitivities for need for neurosurgical intervention and clinically important brain injury. The CCHR, however, has much higher specificity for important clinical outcomes and could result in much lower imaging rates. Key Words: Minor head injury, Canadian CT Head rule, New Orleans criteria

29 IMPACT of an electronic prescription writer on emergency department narcotic prescribing practice

Grafstein E, Stenstrom R, Harris D, Hunte G, Innes G. Department of Emergency Medicine, Providence Health Care and St. Paul's Hospital, Vancouver, BC

Introduction: BC physicians use special and time consuming duplicate prescriptions pads to prescribe narcotics. We built an electronic prescription writer (e-Rx writer) that also automatically creates narcotic prescriptions on the special duplicate pads and reduces physician prescribing time. Our objective was to determine whether the use of the e-Rx writer increased the number of narcotic prescriptions dispensed. Methods: A before and after study in an urban, academic emergency department. We obtained the records of prescribed controlled substances for each emergency physician (EP) from the College of Physicians and Surgeons for a 5 month period (June - October 2005) and the corresponding period one year later when the e-Rx writer was used. We included EPs who were half time clinical or greater and who did not write prescriptions outside the department. Only prescriptions that were presented to a pharmacy were included. Prescribing practices for controlled substances that do not require duplicate prescriptions (benzodiazepines and Tylenol #3) were also compared. We identified all patients with a potentially painful condition (PPC) seen by these physicians in the pre and post narcotic prescription writer period. Results: There were 25 EPs included in the analysis. Based on repeated measures analysis of covariance (using Tylenol #3 prescriptions in the pre and post period as the covariate) there was a significant increase in narcotic prescriptions per 100 patients with a PPC provided in the post e-Rx writer period (mean = 4.2 pre and 6.5 post F 1,24 = 20.3; P < 0.001) . There was no difference in the number of Tylenol #3 or benzodiazepine prescriptions pre and post e-Rx writer. Conclusions: The e-Rx writer for narcotics significantly increases the number of narcotic prescriptions provided in the ED. Key Words: Electronic prescription, Narcotic

Table 1, Abstract 29
Factor	Pre (mean)	Post (mean)	Difference (pre–post)	P-value
Narcotics/100 PPC patients	4.2	6.5	2.3	<0.001
Tylenol #3/100 PPC patients	12.0	12.6	0.6	0.62
Benzodiazepines/100 patients	1.5	1.2	-0.3	0.48

30 MULTIPLYING the serum acetaminophen by the aminotransferase to risk-stratify patients following acetaminophen overdose

Langmann C, Sivilotti MLA, Green TJ, Yarema MC, Johnson DW for CAOS Study Group. Department of Emergency Medicine, Queen's University, Kingston, ON

Introduction: Early risk prediction following acetaminophen (APAP) overdose is challenging when time of ingestion is unknown. Because serum APAP concentrations fall more slowly and serum aminotransferase (AT) rise more quickly in patients with more severe hepatic injury, we have recently proposed the APAPxAT multiplication product may allow early risk-stratification. We describe this novel measure in patients developing hepatotoxicity following acute APAP overdose. Methods: This is a proof-of-concept analysis using an existing dataset, namely the derivation subset of the Canadian Acetaminophen Overdose Study, a large multicenter hospital record review of all admissions for APAP overdose. Only acute overdoses developing an AT>1000 IU/L despite treatment with N-AC were retained for this analysis. At each time point when serum AT was measured (AST or ALT, whichever greater), the corresponding serum APAP was recorded or estimated based on first-order elimination kinetics. Results: Of 3500 hospital admissions, 94 developed AT>1000 IU/L following an acute overdose (median time to N-AC 15.5 hr), and 62 did so within 48 hours of the ingestion ("Early Hepatotoxicity"). A total of 707 discrete timed AT were available for calculation. The table below lists the median [IQR] APAPxAT product (in mM x IU/L) at various time intervals relative to the initiation of N-AC. Conclusions: The APAPxAT product uses readily available lab tests, is simple to calculate, does not require graphical interpretation or plotting, and remains elevated (usually well above 10 mM x IU/L) during the first 12 to 24 hours of N-AC therapy. Moreover, it does not require certain knowledge of the time of ingestion, nor that the overdose be taken at a single point in time. As such, it has many attractive properties for early risk prediction following APAP overdose. Key Words: Acetaminophen overdose, Aminotransferase, Hepatotoxicity

Table 1, Abstract 30
Time post N-AC (hr)	Early hepatotoxicity	Late hepatotoxicity
0±6	113 [61.9, 383]	63.5 [41.6, 126]
12±6	59.0 [18.2, 238]	14.1 [4.42, 29.5]
24±6	23.5 [9.50, 39.4]	12.0 [3.91, 16.2]
36±6	40.0 [20.3, 54.6]	8.16 [5.90, 11.2]