2008 CAEP/ACMU Scientific Abstracts - Poster Presentations: 44-73

2008 Scientific Abstracts

CJEM 2008;10(3):255-295

Abstracts: 1-4, 5-24, 25-43, 44-73, 74-103, 104-131

Poster Presentations

44 THE use of the Ottawa Ankle Rules in children: a survey of physicians’ practice patterns

Dowling SK, Wishart I. University of Calgary, Calgary, AB

Introduction: Despite a number of recent studies assessing the diagnostic accuracy of the Ottawa Ankle Rules (OAR) in children, we are unaware of any studies that have examined physicians practice patterns in this setting. The primary purpose of this survey was to assess whether Canadian pediatric emergency physicians (PEPs) reported using the OAR in their clinical practice. Secondary outcomes included PEPs management and diagnosis of Salter-Harris I ankle fractures. Methods: A self-administered piloted mail survey was distributed to 215 Canadian PEPs using a modified Dillman technique (maximum of 3 mailings for nonresponders). All physician members of Pediatric Emergency Research Canada (PERC), a national organization of PEPs, were surveyed. Outcomes were analyzed using descriptive statistics. Results: The survey response rate was 68.9% (144/209). Six respondents were excluded based on a priori exclusion criteria. The respondents were primarily full-time PEPs (73.2%) and 38.9% had completed a pediatric emergency medicine fellowship. Of those physicians surveyed, 87.5% (126/144) reported applying the OAR in children to determine whether to order radiographs for acute ankle or midfoot injuries. Of these physicians, 65.1% reported using the OAR always/usually. Sixty-five percent of physicians surveyed felt that all ankle fractures are clinically significant. Although the physicians most commonly order the radiographs (63.1%), 34.8% of the time, x-rays are requisitioned by triage nurses or the orthopedic technicians. Management of SH-I fractures the included immobilization by 85.1% of emergency physicians, with the majority arranging follow-up with the orthopedic surgeons. Conclusion: Although there are limitations with self-reported surveys, it appears that the majority of Canadian PEPs are applying the OAR when assessing children with acute ankle and midfoot injuries. Future studies should assess how well the self-reported usage correlates with actual clinical usage of the OAR in the clinical setting. Keywords: Ottawa Ankle Rule, pediatric ankle injury, survey

45 EMERGENCY medicine observerships for preclerkship students: a structured experience

Penciner R. Division of Emergency Medicine, Department of Family & Community Medicine, University of Toronto, Toronto, ON

Introduction: Medical students are expected to make residency and career decisions early in their undergraduate medical education. In medical school curricula there is limited exposure to emergency medicine (EM) in the preclerkship years. The purpose of this study was to develop a formalized EM observer ship program for preclerks and to survey the students following their participation.Methods: A structured EM observership program was developed at the University of Toronto Medical School in 2007. All firstand second-year students were eligible to participate. Nine ED teaching sites were recruited with each site recruiting interested preceptors. The observership consisted of two 4-hour shifts with 1 preceptor at 1 site. Specific expectations of the students were outlined. A convenience sample was used for the period Feb. 26 to Nov. 4, 2007, to conduct an anonymous online survey of the students experience following the observership. Results: During the study period, 82 students completed 99 observerships at 9 sites with 54 preceptors. Seventy-one (71) students completed the survey (response rate 86.6%). Only 22 (31%) students participated in prior informal EM observerships. Most students (94.4%) found the process of setting up the observership easy. Overall, most students (98.6%) found the experience to be worthwhile. Most students (95.8%) viewed the preceptors as good role models and thought the patients were accepting of the students’ presence as observers (98.6%). As a result of the observership, 47 (66.2%) students reported that their attitudes/ interest to EM has changed and most (83.1%) plan on exploring other opportunities in EM (such as electives). As a result of the observership, most students (78.6%) would like to find a mentor in EM and 57 (80.3%) students feel they have the opportunity to contact the preceptor for further discussion (mentoring). Conclusion:Formal EM observerships are viewed by students to be worthwhile and have the potential to impact positively on students’ career choices. Keywords: undergraduate medical education, observerships, curriculum

Winner of the CAEP Resident Research Abstract Award

46 WHAT do emergency medicine workers think and know about crisis resource management? Results from a needs assessment survey

Hicks CM, Denny CJ, Bandiera GW. Sunnybrook Health Sciences Centre, Toronto, ON

Introduction: Emergency department (ED) resuscitation requires the coordinated efforts of an interdisciplinary team. Human errors are common and have a negative impact on patient safety. Although Crisis Resource Management (CRM) skills are utilized in other clinical domains, emergency medicine (EM) caregivers currently receive no formal team training education. Objective: To assess and compare attitudes towards CRM training among EM staff physicians, nurses and residents at two Canadian academic teaching hospitals. Methods: EM staff physicians, residents and experienced nurses were asked to complete a web survey, which included Likert scales and short-answer questions. Focus groups and pilot testing informed survey development. Thematic analysis was performed on the qualitative data set and compared to quantitative results. Descriptive and inferential statistics were calculated to summarize the findings. Results: The response rate was 75.7%. There was strong consensus regarding the importance of core CRM principles (effective communication, team leadership, resource utilization, problem solving and situational awareness) in ED resuscitation. Problems with coordinating team actions (58.8%), communication (69.6%) and establishing priorities (41.3%) were among factors implicated in adverse events. Interdisciplinary collaboration (95.1%), efficiency of patient care (83.9%) and decreased medical error (82.6%) were proposed benefits of CRM training. Communication between disciplines is a barrier to effective ED resuscitation for 94.4% of nurses and 59.7% of physicians (p = 0.008). Residents reported a lack of exposure to (64.3%) yet had interest in (96.4%) formal CRM education using human patient simulation. Conclusion:Nurses rate communication as a barrier to teamwork more frequently than physicians. EM residents are keen to learn CRM skills. An opportunity exists to create and novel interdisciplinary CRM curriculum to improve EM team performance and mitigate human error. Keywords: crisis resource management, team training, needs assessment survey

47 PREDICTING the development of severe bronchiolitis among an emergency department cohort: a multicentre study

Plint AC, Johnson DW, Correll R, Stiell IG, Patel H, Gouin S, Bhatt M, Black K, Turner T, Whitehouse S, Joubert G, Wells GA, Spruyt J, McGahern C, Klassen TP; for Pediatric Emergency Research Canada (PERC). Departments of Pediatrics and Emergency Medicine, University of Ottawa, Ottawa, ON

Introduction: Bronchiolitis is the most common lower respiratory tract infection of young children. While many studies have described risk factors for admission, no study has prospectively examined the risk factors for developing severe bronchiolitis among emergency department (ED) patients. The objectives of this study were 1) to describe the occurrence of severe disease among children with bronchiolitis presenting to the emergency department; and 2) to predict which infants with bronchiolitis are at risk for developing severe bronchiolitis. Methods: Prospective cohort study at 8 Canadian pediatric EDs in which children < 12 months of age presenting with bronchiolitis were assessed for 22 standardized demographic, historical, environmental and physical exam variables. Main outcome measure was severe bronchiolitis defined as need for intubation, apnea, death and admission to a pediatric ICU. Results: There were 1554 patients enrolled. Among the study sample, 482 (31%) of infants were admitted. Thirty-eight infants (2.4%) developed severe bronchiolitis with 36 (2.3%) admitted to ICU, 11 intubated (0.7%) and 7 (0.5%) developed apnea. No infants in the cohort died. Mean age was 22 weeks (SD 13), 39% of the cohort was female, 13% were born at < 37 weeks gestation, 3% had chronic cardiopulmonary disease and 37% were exposed to second-hand smoke. In the univariate analysis age, gestational age, heart rate, respiratory rate, respiratory distress assessment instrument score, oxygen saturation on room air were significantly associated (p < 0.05) with the development of severe bronchiolitis. Variables found to be strongly associated with the outcome of severe bronchiolitis will be further analyzed by multivariate partitioning techniques. Conclusion: This large emergency department cohort of infants with bronchiolitis is the first study to prospectively follow ED patients with bronchiolitis to determine the risk of developing severe bronchiolitis and to examine a wide range of predictor variables. Keywords: bronchiolitis, clinical prediction guide, respiratory emergencies

48 MEASURING tissue oxygen saturation (StO2) via spectrometer in healthy children

Goldman RD, Shah Aziz A, Marques L, Rogovik AL. Pediatric Research in Emergency Therapeutics (PRETx) Program, Division of Pediatric Emergency Medicine, BC Childrens Hospital, Department of Pediatrics, University of British Columbia, and Child & Family Research Institute (CFRI), Vancouver, BC

Background: Near-infrared spectroscopy (NIRS) offers a noninvasive monitoring method for quantifying the percent of tissue oxygen saturation (%StO2) in cases where pulse oximetry (SpO2) is less reliable or not measured, such as in cases of shock or trauma. StO2 in adults is measured in the thenar area but in children, owing to small size, this location may not be suitable for measurement. Objective: The aim of this study was to evaluate a new NIRS system in children and to determine the best area to measure StO2. Methods: Prospective study in children 0–17 years arriving at the emergency department (ED) at the Hospital for Sick Children in Toronto, Ontario. StO2 on several areas including shoulder (deltoid muscle), palm (thenar eminence), forearm, calf and mid-triceps was measured with a 25-mm probe and SpO2 documented. We included patients in the 3 lower triage categories (urgent, semiurgent and nonurgent), with no respiratory distress or any presenting symptom related to a chronic respiratory or cardiovascular disorder. Results: We recruited a total of 316 patients and conducted 983 measurements of StO2. The mean age was 6.8 (SD 4.4) years (range 0.1–17) and 53% were males. Average StO2 was in the normal range on the bicep (82.7, SD 10.1%) and deltoid (82.2, SD 12.8%) muscles and significantly (p < 0.05) lower on other areas. While significantly more children moved their limbs (77 [25%] v. 46 [15%], p < 0.001) or struggled (31 [10%)] v. 20 [6.4%], p = 0.02) during the StO2 measurements, less than 5% in the StO2 group reported any type of pain or cried which did not differ from SpO2. Conclusion: Bicep and deltoid muscles are the most appropriate areas to measure StO2 using the 25-mm transducer in children of different ages. Monitoring peripheral tissue oxygenation using noninvasive NIRS that allows very early application in the pediatric ED, could become a regular part of ED care of critically ill patients to institute prompt therapy and guide resuscitation, avoiding organ damage. Keywords: oxygen saturation, spectrophotometry, pediatrics

49 PHYSICIANS’ clinical interpretation of the Ottawa Ankle Rule for children

Hedrei P, Gravel J, Gouin S, Grimard G. CHU Sainte-Justine, University of Montréal, Montréal, QC

Background: The Ottawa Ankle Rule (OAR) is a clinical decision rule developed to identify, based on a short physical exam, the need for radiography for patients with an ankle trauma. Objective: Determine the appropriateness of interpretation of the OAR by physicians evaluating children with ankle trauma. Methods: This was a prospective cohort study performed in a university-affiliated pediatric emergency department (ED). Participants were children less than 16 years with an acute ankle trauma. Before study recruitment, all physicians were formally instructed in the use of the OAR. For each participant, the ED physician completed a standardized datasheet that described thoroughly the physical exam. He then interpreted the need for radiography according to the OAR rule with the aid of a standardized diagram. Ankle radiography was left at the physicians’ discretion. All standardized datasheet were reviewed by a single reviewer to assess for the need for radiography based on the described physical exam and the OAR guidelines. The primary outcome was the correlation between the physicians interpretation of the OAR (x-ray or not) and the need of radiography according to the description of the physical exam. This correlation was evaluated using a κ score. Results: There were 272 participants included in the study with a total of 80 ankle fractures. All patients had a completed standardized datasheet with no missing data. Physicians evaluated that 227 (0.83) patients needed radiography according to the OAR while the description of the physical exam suggested that 213 (0.78) would need it. There was a good correlation between the physicians interpretation and the physical exam has demonstrated by a κ score of 0.66 (95% CI 0.60–0.72). Discordance in the interpretation was not associated with a missed diagnosis of fracture Conclusion: Pediatric emergency physicians use the Ottawa Ankle Rule appropriately in children with ankle trauma. Keywords: Ottawa Ankle Rule, pediatrics, interobserver agreement

50 HYPERGLYCEMIA at presentation predicts functional outcome in nonlacunar infarctions

Bellolio MF, Gilmore RM, Vaidyanathan L, Enduri S, Suravaram S, Decker WW, Rabinstein AA, Stead LG. Mayo Clinic College of Medicine, Rochester, MN

Introduction: Assess the functional outcome in patients presenting with and without hyperglycemia in lacunar and nonlacunar stroke subtypes. Methods: A consecutive cohort of 1013 patients presented to an academic emergency department with an acute ischemic stroke from December 2001 to December 2005 was included. All patients had serum glycemia levels measured in the emergency department. The variables collected were age, gender, stroke severity at presentation (National Institutes of Health Stroke Scale [NIHSS] from 0 to 42 points), functional outcome at hospital dismissal (modified Rankin score, from 0 to 6 points) and stroke subtype categorized in lacunar versus nonlacunar. Glycemia followed a normal distribution and was analyzed as continuous variable with parametric tests. All p values were 2-sided. Results: A total of 1013 patients were included. The mean and standard deviation age was 72.3 (SD 14.5) years; 53% were male. The median NIHSS at admission was 5 (interquartile range 3–11), with a mean of 8. The mean functional outcome was 3.0 (SD 1.6). A total of 27% of the cohort had diabetes and 74% hypertension. Overall, 152 (15%) had lacunar strokes. In nonlacunar strokes glycemia was a strong predictor of functional outcome at hospital dismissal (p< 0.001); the higher the glycemia level the worse the functional outcome. However, in lacunar strokes, glycemia was not a predictor of functional outcome at hospital dismissal (p = 0.98). Conclusion:Hyperglycemia at presentation predicts functional outcome in nonlacunar infarctions; however, does not predict the outcome of patients presenting lacunar infarctions. Hyperglycemia at emergency department presentation does not confer a worse prognosis in patients with lacunar strokes. Keywords: stroke, hyperglycemia, functional outcome

51 MANAGEMENT of transient ischemic attack and stroke in British Columbia emergency departments: the Current Practice Indicator project

Harris DR, Innes GD, Grafstein E, Stenstrom R, Holmes A. Providence Health Care, St. Paul’s Hospital, University of British Columbia, Vancouver, BC

Introduction: Stroke is the third leading cause of death and the leading cause of adult disability in Canada. We sought to measure the performance of BC EDs in the management of patients with TIA and stroke, when measured by evidence-based clinical indicators.Methods: This was a retrospective medical records review study performed in adherence to published guidelines. Charts were randomly sampled from selected EDs in all 6 health regions (HRs) in BC, with an ED discharge diagnosis of TIA or stroke. Where available, ED administrative databases were utilized; otherwise, medical records were hand searched. Fourteen data abstractors were given standardized training and were blinded to study hypotheses. Adherence to 19 performance indicators was measured using standardized data collection forms. Ethical approval was obtained prior to study inception. Interrater reliability was calculated. Results: There were 1352 charts audited from 48 emergency departments in BC. Of these, 685 (50.1%) charts were stroke patients; 503 (37.2%) and 164 (12.1%) were TIA and, query TIA patients, respectively. Percent agreement ranged from 61% to 91% between HRs. Of charts, 430/1352 (31.8%) arrived within 2.5 hours of symptom onset, 23/685 (3.4%) received tPA, and of those, 7/23 (30.4%) received tPA within 1 hour of arriving to the ED. Only 618/1352 (45.7%) had a blood glucose checked on arrival and 1137/352 (84.1%) had an electrocardiogram performed in the ED. Of patients, 547/1352 (40.5%) were mobilized within 24 hours and only 133/1352 (9.8%) had a swallowing screen in the ED. Upon discharge, 509/1262 (40.3%) were prescribed an antiplatelet and 177/381 (46.5%) patients with atrial fibrillation were prescribed anticoagulants. Only 51/1352 (3.8%) patients or caregivers were given stroke education on discharge. Conclusion: The care for patients with TIA and stroke in emergency departments in BC could be optimized. Various knowledge translation implementation strategies may be employed to improve care for stroke and TIA patients in the emergency department.Keywords: transient ischemic attack, performance indicators, medical records review

52 PHYSICIAN attitudes towards clinical decision rules for acute respiratory disorders for patients > 50 years old in the ED: an international survey

Perry JJ, Goindi R, Symington C, Taljaard M, Brehaut J, Schneider S, Stiell IG. University of Ottawa, Ottawa, ON

Introduction: There are no widely used guidelines to determine which older patients with acute respiratory conditions require admission. This survey assessed the need for clinical decision rules (CDR) to determine the need for hospital admission for patients > 50 years for 3 respiratory conditions: chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF) and community acquired pneumonia (CAP). Methods: We surveyed ED physicians from 3 countries (the United States, Canada and Australia) by a random sample of their respective ED physician associations (ACEP, CAEP and ACEM). We used a modified Dillman technique with up to 3 mailed surveys plus a prenotification letter. Physicians were asked a series of questions regarding the need for CDRs for ED patients > 50 years to direct the need for admission for COPD, CHF or CAP, and the required sensitivity of each rule for death within 14 days. The current use of some form of walk test in the ED to assess for possible discharge was also assessed. Results: There were 801 responses from 1493 surveys received with rates of 55%, 60% and 46% for Australia, Canada and the United States, respectively. The median years in clinical practice was 11 (IQR: 7, 17) and 69% were male.

Table 1, Abstract 52. The percentage of physicians who support the need for a rule and who would use such a rule stratified by disease and country

	% support needed for rule; % would use rule
Country	CHF	CAP	COPD
Overall	65; 93	69; 92	63; 92
Australia	65; 92	73; 92	64; 91
Canada	77; 94	75; 93	75; 93
United States	48; 91	55; 92	45; 90

The median sensitivity for death within 14 days was 97%–98% for all conditions. A walk test is used by 45%, 60% and 64% of Australian, Canadian and US physicians. Conclusion: This international survey determined that ED physicians would adopt highly sensitive CDRs for all 3 conditions. This survey supports the development of CDRs to determine the need for hospital admission for ED patients > 50 years with COPD, CHF or CAP. Keywords: clinical prediction guide, acute respiratory disorder, survey research

53 USE of personal protective equipment in Canadian pediatric emergency departments

Reid SM, Ken J, Farion KJ, Suh KN, Audcent T, Barrowman NJ, Plint AC. Department of Pediatrics, University of Ottawa, Ottawa, ON

Introduction: This study examined the attitudes, knowledge, selfreported behaviors and perceived barriers to compliance with infection control practices and the use of personal protective equipment (PPE) in Canadian pediatric emergency departments (EDs). Methods:A self-administered survey tool consisting of 21 questions was developed and validated de novo for this study. The survey was mailed to all individuals listed on the Pediatric Emergency Research Canada Database of physicians practising pediatric emergency medicine in Canada. Results: There were 187 physicians surveyed and 125 (67%) responded. Respondents had a median 9 years of experience (range 0–32 yr) and 50 respondents (41%) had completed a pediatric emergency medicine fellowship. Fifty-four percent reported that they had either never received PPE training or had not been trained in the previous 2 years. Respondents scored a mean of 4.92 out of 11 questions correct (SD 1.66) on knowledge-based questions although 53% reported being very or somewhat comfortable with their knowledge of transmission-based isolation practices. Only 11% reported always or usually wearing a mask when assessing febrile respiratory patients. For scenarios assessing self-reported use of PPE, respondents reported correct PPE use in a mean of 1.01 out of 6 scenarios (SD 0.95). There was no statistically significant correlation between knowledge and reported use of PPE (p = 0.08). Respondents report they would be more likely to use PPE appropriately if patients were clearly identified prior to physician assessment, equipment was easily accessible, and PPE was made a priority in their ED. Conclusion:Knowledge and self reported adherence to recommended infection control practices among Canadian pediatric emergency physicians is suboptimal. Early identification of patients requiring PPE, more convenient access to PPE, and improved education regarding isolation and PPE practices may improve adherence to these important guidelines.Keywords: personal protective equipment, infection control, pediatrics

54 OCCURRENCE of override among nurses using the Canadian Canadian Triage and Acuity Scale Paediatric Guidelines

Gravel J, Gouin S, Manzano S, Arsenault M, Amre D. CHU Sainte- Justine, Université de Montréal, Montréal, QC

Background: The Canadian Triage and Acuity Scale Paediatric Guidelines (PaedCTAS) is a triage tool that suggests triage level to children visiting an emergency department (ED) based on a brief clinical assessment. Guidelines of the PaedCTAS stipulate that nurses can override the suggested triage level with caution based on clinical judgment. Objective: To measure the frequency and reproducibility of override for nurses using a computerized version of the PaedCTAS in a pediatric ED. Methods: This was a prospective cohort study among a children visiting a single tertiary care pediatric ED. Recruitment occurred during 40 shifts of 8 hours from June to September 2007. To be included, children had to be initially triaged from level II (emergent) to V (non urgent) on the PaedCTAS by a regular triage nurse using a computerized version of the PaedCTAS (Staturg from Statdev). All participants were triaged a second time by a research nurse immediately after their first triage using the same tool. Both triages were performed blinded to each other. Research nurses were regular ED nurses performing extra hours for research. Final triage levels assigned by the triage and research nurses and those suggested by Staturg were recorded. The two primary outcome measures were the percentage of override and the interrater agreement for these overrides measured by the κ score. Results: There were 499 patients recruited to participate in the study. The percentages of overrides were 23.2% (116/499) for the regular triage nurses and 21.8% (109/499) for the research nurses. These overrides were equally distributed among increase and decrease in triage level assigned. The interrater agreement between the nurses for overrides was poor (κ score of 0.319 SD 0.098). Conclusion: There was a high prevalence of override among nurses using a computerized version of the PaedCTAS. The interrater agreement between nurses for these overrides was poor. Keywords: PaedCTAS, overrides, interobserver agreement

55 CHARACTERIZING wheelchair-related falls — identifying risk factors and targets for future intervention

Sillberg VAH, Woo MY, Symington C, Perry JJ. University of Ottawa, Ottawa, ON

Introduction: Wheelchair-related falls are an avoidable cause of injury. This study assessed the characteristics of patients injured owing to a fall from a wheelchair, the incidence of significant injuries and need for hospitalization. Methods: A formal health records review was performed from January 2002 to March 2006 in a tertiary referral centre. Research ethics approval was obtained. We included patients presenting to the emergency department with a complaint related to a fall from a wheelchair. A single reviewer identified cases and abstracted data into a standardized database. The primary outcome was risk of serious injury (fracture, dislocation, intracranial bleed, death, other life-threatening injury). The secondary outcome was need for hospital admission. Analysis included χ2 and t tests.Results: The analysis included 194 patients; 228 patients were identified using ICD codes. Eighteen patients did not meet inclusion criteria and 16 charts were not found. Patients had a mean age of 64.6 years (range 18–104) and 62.9% were female. Of falls, 31.4% occurred during a wheelchair transfer and 37.6% occurred in a supervised residence. The most common location of injury involved the hip and lower limb (43.3%), and head (34.0%). The most common injury was fractures (38.7%) and soft tissue injuries (30.4%). Seven patients (3.6%) suffered an intracranial bleed and 1 patient (0.5%) died. Overall, 44.3% of patients sustained a serious injury; 24.2% were admitted for a mean hospital length of stay of 23.6 days (range 0–460). There was an increased risk of admission in patients with coronary artery disease (p = 0.01), diabetes (p = 0.03) and peripheral vascular disease (p = 0.02). Conclusion: There is a significant risk of serious injury resulting from wheelchair-related falls. Further study and intervention strategies are required in this patient population. Keywords: wheelchair-related falls, injury, risk factors

56 DIAGNOSTIC accuracy of clinical prediction rules to exclude acute coronary syndrome in the emergency department

Hess EP, Thiruganasambandamoorthy V, Wells GA, Erwin P, Jaffe A, Montori V, Stiell IG. University of Ottawa, Ottawa, ON

Introduction: To determine the diagnostic accuracy of clinical prediction rules (CPRs) to exclude acute coronary syndrome (ACS) in the ED setting. Methods: MEDLINE, EMBASE, Web of Science and the Cochrane Database of Systematic Reviews were searched with the assistance of an expert librarian. Additional studies were identified by searching reference lists and contacting content experts. Articles were selected for review based on the following criteria: 1) enrolment of consecutive ED patients; 2) incorporation of variables from the history or physical examination, electrocardiogram and cardiac biomarkers; 3) stress testing or coronary angiography not incorporated as part of prediction rule; 4) based on original research; 5) prospectively derived or validated; 6) does not require use of a computer; and 7) data available to construct a 2 × 2 contingency table. Quality assessment and data extraction were performed using a standardized data extraction form. Analyses included descriptive statistics and calculation of diagnostic test performance characteristics with 95% CIs using MetaDiSc software.Results: Eight studies met inclusion criteria, encompassing 7937 patients. None of the studies verified the results of the CPR by using a reference standard on all or a random sample of patients. Data were not pooled owing to between-study heterogeneity. Sensitivities and specificities ranged from 94% to 100% and 13% to 57%, and positive and negative LRs from 1.1 to 2.2 and 0.01 to 0.17, respectively. Of the 2 CPRs with reported sensitivities of 100%, the sensitivity of 1 has not been consistently reproduced in the literature, and the other incorporated variables not available in all patients. The upper bound of the 95% CI of the negative LR included 0.1 in all studies. Conclusion:Current prediction rules for ACS have substantial methodologic limitations and do not exclude ACS with sufficient certainty to be clinically useful. Future methodologically-sound studies are needed to guide clinical practice. Keywords: clinical prediction guide, acute coronary syndromes, systematic review

57 COMPARING the predictive validity of memory based triage to a computerized emergency triage tool

Dong SL, Bullard MJ, Meurer DP, Villa-Roel C, Holroyd BR, Rowe BH. University of Alberta, Edmonton, AB

Introduction: Emergency department (ED) triage prioritizes patients based on urgency of care. A web-based triage tool (eTRIAGE) has been developed using the 5-level Canadian Emergency Department Triage and Acuity Scale (CTAS) templates. Despite research on triage reliability, limited triage validation research has been published. This study examined 2 cohorts: patients triaged with CTAS using a memory-based tool and those triaged with the eTRIAGE program. The predictive validity of the triage scores based on patient outcomes of acuity and resource utilization were examined. Methods:In both cohorts, dedicated triage nurses assessed each patient on ED presentation and both were unaware of the study. Each patients triage score, resource utilization (need for specialist consultation, computerized tomography and ED length of stay [LOS]) and acuity (admission to hospital or death in the ED) were collected over two 6-month study periods: January–June 2003 (memory-based triage cohort) and January–June 2004 (eTRIAGE cohort). Using a regional database, the association between the triage scores and each categorical outcome were analyzed through univariate logistic regression models. A univariate ANOVA was used to evaluate the significance of ED LOS variation according to each CTAS score. Results: The memory (n = 30 078) and eTRIAGE cohorts (n = 29 447) involved patients with similar ages, sex ratios, mode of arrival and triage score distribution. Compared with CTAS I, the odds ratios for specialist consultation, CT scan, admission to hospital and death were significantly lower in CTAS II, III, IV and V (p < 0.001) in both cohorts. The ED LOS showed a significant variation among CTAS scores (p < 0.001) in both cohorts. Conclusion: Both memory-based triage and eTRIAGE demonstrated excellent predictive validity for resource utilization and patient acuity. CTAS scoring appears to be a valid and consistent method of estimating resource use in EDs and should be included in future ED health economic studies.Keywords: computerized triage, memory based triage, CTAS

58 USE of prehospital strategies to reduce time to reperfusion for STEMI patients varies widely among Canadian EMS operators

Vaillancourt S, Donovan L, Boothroyd L, Andrusiek D, Travers A, Trickett J, Sookram S, Canham H, Lefrancois D, Vermeulen MJ, Tu JV, Schull MJ; for the Canadian Cardiovascular Outcomes Research Team. Institute for Clinical Evaluative Sciences, Toronto, ON

Introduction: Timely reperfusion therapy for ST-elevation myocardial infarction (STEMI) is an important determinant of outcome, yet time to treatment targets are widely unmet in North America. Prehospital strategies can reduce time to reperfusion; we sought to determine the extent emergency medical services’ (EMS) use these strategies in Canada. Methods: A cross-sectional survey in 2007 of land EMS operators in British Columbia (BC), Alberta (AB), Ontario (ON), Quebec (QC) and Nova Scotia (NS) focused on the use of 4 prehospital strategies: 1) 12-lead electrocardiogram (ECG), 2) routine expedited emergency department (ED) transfer of STEMI patients (from a referring ED to a percutaneous coronary intervention (PCI) centre), 3) ED bypass (bypass of local EDs to transport directly to a PCI centre); and 4) prehospital fibrinolysis. Results: We surveyed 99 ambulance operators (provincial response rate varied from 87.5% in AB to 100% in all others), representing 15 681 paramedics and serving 97% of the combined provincial populations. Of operators, 68% (95% CI 59%–77%) had ambulances equipped with 12-lead ECGs, ranging from 40% in QC to 100% in AB and NS. Overall, 47% (95% CI 46%–48%) of paramedics were trained in ECG acquisition and 40% (95% CI 39%–41%) were trained in ECG interpretation. Only 18% (95% CI 10%–25%) of operators had ED bypass protocols; 45% (95% CI 35%–55%) had protocols for expedited ED transfer. Prehospital fibrinolysis occurred only in AB (50% of respondents). All EMS operators in BC, AB and NS had at least 1 of the 4 prehospital strategies, while one-third of operators in ON and QC had 0 of 4. In major urban centres, at least 1 of 12-lead ECG acquisition, bypass and expedited transfer was available, but there was considerable urban variation in and across provinces. Conclusion:There is substantial variation in prehospital STEMI strategies across these provinces. The key role of EMS providers in STEMI care is increasingly recognized, yet access to widely recommended protocols and technologies is not uniform. Keywords: prehospital, door-to-balloon times, door-to-needle times

59 THE state of EM ultrasonography adoption: a national survey using a novel evaluation tool (ETUDE)

Woo MY, Frank JR, Lee AC. University of Ottawa, Ottawa, ON

Introduction: Emergency Medicine Ultrasonography (EMUS) first appeared in the 1980s in North America but the extent of its adoption is unknown. We developed a tool to characterize EMUS adoption and applied it to a national survey of emergency physicians (EPs). Methods: We used Rogers Diffusion of Innovations theory to develop a novel Evaluation Tool for Ultrasound skills Development and Education (ETUDE), which assessed level of training, past, current and potential use of EMUS. ETUDE content and weighting was validated by 3 independent experts in EMUS. We defined a series of patterns to categorize respondents into innovators, early adopters, majority and those not interested in EMUS. We applied the tool to a web-based survey of a sample of the Canadian Association of Emergency Physicians membership database using a modified Dillman technique. We used descriptive statistics and correlations to analyze the data. Results: The 296 respondents (36.4% of 814 surveyed) had a median age of 40 years and were 72.5% male. EMUS was available to 53.5% of respondents after 2004. EPs adoption scores using ETUDE followed a normal distribution of innovators (18.0%), early adopters (34.5%), majority (28.7%), and those not interested (18.8%). Respondents varied in their level of training in EMUS: 21.0% had no training, 51.3% had introductory training, 21.4% were credentialed and 6.3% had advanced training. Respondents endorsed always using EMUS currently and in the future for FAST (current 41.8%/future 88.4%), first trimester pregnancy (current 23.3%/future 73.7%), suspected AAA (current 32.7%/future 92.6%), cardiac (current 30.7%/future 87.5%) and central venous catheterization (current 17.0%/future 80.3%). Conclusion: We developed a novel validated Evaluation Tool for Ultrasound skills Development and Education (ETUDE) and used it to characterize the current state of adoption of EMUS in Canada. ETUDE can be used to evaluate uptake of EMUS over time and to compare adoption in different countries. Keywords:ultrasound, educational instruments, survey research

60 PREVALENCE and characterization of community-acquired methicillin resistant Staphylococcus aureus colonization in highrisk individuals in Toronto

Borgundvaag B, Svoboda T, McGeer A, Willey B, Rostos A, Gnanasuntharam P, Tyler A, Porter V, Kreiswirth N. Mount Sinai Hospital, Toronto, ON; Centre for Research on Inner City Health, St. Michaels Hospital, Toronto, ON; University of Toronto, Toronto, ON

Introduction: Community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) has emerged as a common pathogen causing skin and soft tissue infections (SSTIs). In many parts of North America CA-MRSA has now replaced methicillin-susceptible S. aureus (MSSA) as the primary cause of SSTIs. There are several commonly cited risks for CA-MRSA infection, yet little is known about colonization rates in high-risk individuals. Methods: Between Jul. 10 and Aug. 18, 2007, 295 consenting male residents of a Toronto community shelter provided information regarding risk factors for MRSA colonization and swabs from their nares, axilla and any visibly open sores. Swabs were enriched and selectively cultured for MRSA and MSSA, which were identified using standard methods. MRSA were typed by SmaI PFGE and SCCmec type and presence of PVL was determined by PCR. Results: Overall, 110 (37%) and 12 (4.1%) of residents screened positive for MSSA and MRSA, respectively. MRSA were of 5 distinct types, the largest cluster included 7 residents positive from 11 sites, including a postfrostbite foot infection. This unusually resistant (R; R to clindamycin and fucidin) cluster was closely related to the PVL-positive epidemic community strain, CMRSA-10 (USA-300). Two other residents carried the typical CMRSA-10/USA-300 strain responsible for the surge of CA-MRSA in North America. The remaining 4 MRSA (all PVLneg) included 2 CMRSA-2 (USA-800) with community-acquired SCCmec-IVa (R to blactams only), 1 CMRSA-2 (USA-100) with SCCmec-II (the common nosocomial strain) and 1 CMRSA-1 (USA-600) also carrying the community SCCmec-IVa cassette.Conclusion: The most common isolate of CA-MRSA found in highrisk individuals was an unusually drug resistant variant of the most common strain of CA-MRSA. The implications of colonization with this strain are yet to be determined. Keywords: CA-MRSA, SSTIs, colonization rates

61 PROLONGED TASER drive stun exposure in humans does not cause worrisome biomarker changes

Ho JD, Dawes DM, Lapine AL, Bultman LL, Miner JR. University of Minnesota Medical School and Hennepin County Medical Center, Minneapolis, MN; Lompoc District Hospital, Lompoc, CA

Introduction: The TASER (TASER International Inc., Scottsdale, Arizona) electronic control device (ECD) is used to control dangerous behaviour in 2 ways. The primary method is probe deployment. The secondary method is the Drive Stun (DS). This is the first study of the human effects of the DS. ECDs are scrutinized since occasional deaths occur unexpectedly following their use. Some deaths have followed a DS. Cases of custodial sudden deaths occur when no ECD has been used, but a causal relationship is hypothesized.Methods: Volunteers underwent 24 hour monitoring. After informed consent, a health history and baseline bloodwork was obtained. Subjects then received either a 15-second or 2 consecutive 5-second DS applications. Applications were to the upper trapezius region using a TASER X-26 ECD. Bloodwork was obtained after exposure and again at 8 and 24 hours after exposure. Samples were analyzed for BUN/creatinine ratio, potassium, CK–MB, lactate and troponin I. Results: There were 21 subjects enrolled (98.5% male, mean age 40.3 years, SD 6.8, range 29–55; mean body mass index 28.4, SD 3.5, range 21.1–36.8). Eleven had the single continuous exposure and 10 had the 2 shorter exposures. Repeated measure ANOVA showed no significant change from baseline at the 4 time points or between exposure types for BUN/creatinine ratio (mean value 14.8 μg/L, SD 3.7, range 6.6–23, p = 0.40), serum potassium (mean value 4.0 mmol/L, SD 0.4, range 3.0–5.1, p = 0.26), or serum CK–MB (mean baseline value 2.45 μg/L, SD 2.89, range 0–20.9, p = 0.32). A significant decrease in serum lactate occurred from baseline at the 8-hour time point (p = 0.005; baseline mean 1.87 mmol/L 95% CI 1.39–2.35; immediate postexposure mean 1.35 mmol/L 95% CI 1.04–1.65; 8 hour mean 1.06 mmol/L 95% CI 0.92–1.2; 24 hour 1.22 mmol/L 95% CI 1.1–1.4). All troponins were < 0.2 μg/L. Conclusion:There were no worrisome changes in the measured serum biomarkers. There was a significant decrease in serum lactate after exposure. This data does not support a causal relationship between ECD DS exposure and worsening physiology. Keywords: TASER, electronic control devices, serum biomarkers

62 COMPLIANCE with infection control guidelines for central venous catheter insertion in the emergency department

Woo MY, Roth V, Beland C, Cahill M, Worthington JR. University of Ottawa, Ottawa, ON

Introduction: Central venous catheter (CVC) related bloodstream infections are a major cause of morbidity and mortality. This study determined the compliance rate with infection control guidelines for CVC insertion in the ED. Methods: A prospective observational cohort study was performed between November 2006 and March 2007 in a tertiary ED referral centre. Ethics approval was obtained. A standardized data collection form was completed by nurses assisting with the procedure. The primary outcome was the rate of compliance with current best practice infection control guidelines for CVC insertion. Analysis included descriptive statistics and χ2 tests. Results:There were 33 CVC insertions performed on 29 patients. CVC insertions were performed by staff physicians (30.3%) or residents (69.7%) from emergency medicine (45.5%), intensive care (45.5%) or medicine (9.0%). Of insertions, 42.4% were emergent and 54.5% of insertions were urgent. Ultrasound was used 42.4% of the time. The femoral vein, internal jugular vein and the subclavian vein were used in 45.5%, 33.3% and 21.2% of the cases, respectively. The femoral vein was more likely to be used in emergent insertions (71.4%; p = 0.01). All recommended infection control measures were used only 29.0% of the time, but only 15.4% of the time by emergency medicine personnel (p = 0.05). The most common breaches were failure to cover the patient with a sterile gown (36%), to maintain a sterile field (15%), to use a sterile drape (12.1%) or to wear a cap (12.1%). Hand hygiene was performed before the procedure by 58% of operators and 57% of assistants. A mask was not worn in 6% of cases. There were no statistical differences in compliance with infection control between staff and residents, urgency of the procedure or time of day. Conclusion: There is poor compliance with infection control guidelines for CVC insertion in the ED. Further educational strategies are required to improve compliance with guidelines. Keywords: infection control, central venous catheter, compliance

63 IS the combination of vasopressin and epinephrine superior to epinephrine alone in the treatment of cardiac arrest? A systematic review

Sillberg VAH, Perry JJ, Stiell IG, Wells GA. University of Ottawa, Ottawa, ON

Introduction: No evidence supports vasopressin (VASO) over epinephrine (EPI) in cardiac arrest; however, animal and some clinical studies support the concurrent use of VASO and EPI together. This systematic review compares the efficacy of VASO and EPI used together versus repeated doses of EPI alone in cardiac arrest. Methods:For this systematic review, we searched MEDLINE, EMBASE and Cochrane Central Register of Controlled Trials. We included randomized controlled trials (RCTs) where VASO and EPI were administered concurrently to cardiac arrest patients within the half life of VASO (6 min). Two reviewers assessed studies for eligibility, data extraction and quality. κ statistics measured interrater agreement. Appropriateness of studies for meta-analysis was assessed. The primary outcome was return of spontaneous circulation (ROSC) and the secondary outcome was survival to hospital discharge. Relative risk was obtained of each outcome. Results: From 235 titles identified, we reviewed 29 abstracts. Twenty-three were excluded (wrong outcomes, not RCTs, not both EPI and VASO, or unobtainable data). Three cardiac arrest studies were included. Study 1 randomized VASO versus EPI then subsequent EPI. Study 2 randomized 2 doses of VASO versus EPI. Study 3 randomized VASO versus placebo, administered a mean 4.6 minutes following initial EPI. κ for included studies and quality were both 1.0. All studies favored combination treatment for ROSC, but only study 2 was statistically significant: RR 1.1, 95% CI 0.82–1.46, RR 1.42, 95% CI 1.14–1.77, and RR 1.02, 95% CI 0.74–1.42, respectively. Studies 1 and 2 reported survival to discharge, with RRs of 0.88, 95% CI 0.26–2.92, and 3.69, 95% CI 1.52–8.95, respectively. The methods for the 3 studies were too dissimilar to allow pooling of results. Conclusion:This systematic review of the combination of VASO and EPI found trends towards better ROSC but equivocal effects on survival. There is a need for RCTs to evaluate the simultaneous use of VASO and EPI in cardiac arrest. Keywords: vasopressors, cardiac arrest, systematic review

64 RESOURCE utilization by the very elderly in the emergency department

Unger B, Afilalo M, Lang E, Segal E, Colacone A, Xue X, Soucy N. Sir Mortimer B Davis–Jewish General Hospital, McGill University, Montréal, QC

Introduction: Several tools for measuring patient load in emergency departments have defined age as a key determinant of resource utilization (RU). However, the elderly are not a homogenous group and those over the age of 85 might represent a subgroup of patients that require even greater resource utilization. The objective of this study is to compare RU in the younger seniors (YS 65–84) and the very elderly (VE greater and equal to 85). Methods: The data from an Oracle database from April 2003 to December 2007 of a university teaching, tertiary adult emergency department was analyzed using Excel. The age cohorts for this analysis were patients from 65 to 84 years old (YS) and those above 85 years old (VE). Indices of RU included length of stay, admission rate, consultation requests, imaging and the need for isolation. Results: Of the 304 599 visits analyzed retrospectively, the VE accounted for 7% of all visits while the YS accounted for 22%. RU by the VE is consistently higher than that of YS in the following ways: triaged as a code 1; RR of 1.81 (with 95% CI 1.66–1.96), LOS 18.6 verus 12.8; p < 0.001, accounting for 14% and 33% of all patient hours, and the rates of hospitalizations was 37% and 25% of each cohort. The RU is also reflected by the patients with a LOS over 48 hours: RR of 2.12 (with 95% CI 2.06–2.18); LOS over 60 hours: RR of 2.31 (with 95% CI 2.23–2.39), requiring consultation; RR of 1.32 (with 95% CI 1.30–1.34), requiring advanced imaging; RR of 1.20 (with 95% CI 1.17–1.23), preventive isolation; RR of 1.98 (with 95% CI 1.90- 2.06) and the admission rate of 37% versus 25%; p < 0.001, with hospitalization from the ED; RR of 1.49 (with 95% CI 1.47–1.51).Conclusion: We report that the very elderly above the age of 85 year old are a clear distinct group within the elderly cohort in patient load and resource utilization in an adult tertiary emergency department. This could be reflected in increased resource allocation and physician billing agreements. Keywords: resource utilization, ED crowding, very elderly

65 IMPROVING care of subacute patients in the emergency department: the Kaizen approach

Penciner R, Ennis A, Paraghamian G, Richard H, Rutledge T, Woollard S. North York General Hospital, Emergency Services Program, Toronto, ON

Introduction: Kaizan (Japanese for continuous improvement) is a methodology used in the workplace for applying rapid process improvement. A Kaizen was held to address the care of patients triaged to the subacute area of the emergency department (ED). The purpose of this study was to evaluate if the implementation of a new process in the subacute area decreases length of stay for these patients. Methods: Members of North York General Hospital ED participated in a 1-week Kaizen event to address care of patients triaged to the subacute area. The process involved analyzing all aspects of care for these patients from registration to discharge. As a result, a rapid assessment zone was created utilizing 5 existing rooms in the subacute zone of the ED and an adjacent waiting area. Patients were placed in rooms only for assessments or specific treatments by the physician and/or nurses. A treatment area with reclining chairs was established for patients requiring ongoing treatments. Utilizing the ED information system, we prospectively evaluated all patients triaged to the subacute area for 120 days after implementation comparing this retrospectively to patients triaged to the subacute area in the 120 days prior. Primary endpoints included length of stay (LOS) of patients, time to be seen by a physician (time to MD) and patients leaving without being seen by a physician (LWBS). Results: Prior to the implementation, 5203 patients were seen in the subacute area with an average LOS of 9.3 hours, compared with 7102 patients (36.5% increase) after implementation with an average LOS of 6.1 hours (34.4% decrease). The time to MD was 2.9 hours before and 2.6 hours (10.3% decrease) after with 1333 patients LWBS before and 1237 patients LWBS (7.2% decrease) after implementation of the rapid assessment zone. Conclusion: Implementing a rapid assessment zone in the subacute area of the ED using Kaizen methodology has significantly decreased LOS of these patients. Keywords: rapid assessment zone, quality improvement, fast track

Winner of the CAEP Resident Research Abstract Award

66 CEREBRAL infarct volumes on diffusion weighted imaging are related with cardiac biomarkers, stroke subtypes and functional outcome in patients with acute ischemic stroke.

Bhagra A, Bellolio MF, Suravaram S, Enduri S, Gilmore RM, Decker WW, Stead LG. Mayo Clinic College of Medicine, Rochester, MN Objective: To determine whether cerebral infarct volumes on diffusion weighted imaging are related with cardiac biomarkers, stroke subtypes and functional outcomes in patients with acute ischemic stroke (AIS). Methods: This was a cohort study done in a tertiary care academic center with patients presenting with AIS within 24 hours of symptoms and brain MRI within index hospitalization (December 2001 to March 2004). We measure total volume of the infarct by an algorithmic approach incorporating and adding effective mapped out areas on all MR slices depicting the infarct. Stroke subtype was categorized with TOAST. Serum troponin was measured at admission. Stroke severity was determined using the National Institutes of Health Stroke Scale (NIHSS). Functional outcome was assessed with modified Rankin score (mRs) at hospital discharge (good 0–3 and poor outcome 4–6). Results: Of 217 patients, the mean age was 70.8 (SD 13.9) years and 116 (53.5%) were males. A total of 58 had large vessel disease, 61 cardioembolic, 30 small vessel, 11 other causes, 16 multiple causes, 32 no cause, 9 insufficient information. The mean NIHSS was 5.8 (median 4, SD 5.2). Patients with large vessel disease or cardioembolic cause had a significantly larger MRI volume than patients with small vessel (both p < 0.001), or no cause identified (p = 0.029 and 0.064). Patients with an elevated troponin were more likely to have a larger volume of infarct. In a regression model the association between elevated troponin (yes/no) and MRI volume yielded an OR 1.2 (95% CI 1.03–1.5, p = 0.025). After adjusting for age and NIHSS, larger MRI volume was a predictor of poor Rankin (OR 1.3, 95% CI 1.1–1.6, p = 0.013). This represents a 30% increase in risk of poor outcome per a doubling in MRI volume. Conclusion: Stroke subtype predicts ultimate volume of infarct on MRI. An elevated troponin appears to be associated with a larger volume of infarct. Larger infarct volume on initial MRI scan is associated with a poor functional outcome at hospital discharge. Keywords: cerebral infarct volumes, acute ischemic stroke, troponin

Winner of the CAEP Resident Research Abstract Award

67 OUTCOME of children treated with prochlorperazine for migraines in a pediatric emergency department

Doyon-Trottier E, Bailey B, Dauphin-Pierre S, Gravel J. CHU Sainte-Justine, University of Montréal, Montréal, QC

Introduction: In an RCT, prochlorperazine has been demonstrated to relieve pain in children with migraines. However, the International Headache Society Clinical Trial Subcommittee also suggests evaluating other outcomes that are meaningful to patients, such as the need for further rescue medication or symptoms recurrence. Objective: To evaluate the rate of treatment failure associated with intravenous prochlorperazine used in the treatment of children with severe migraines in a pediatric emergency department. Methods:This study was a retrospective chart review of patients less than 18 years of age that visited the ED of a tertiary care pediatric hospital between November 2005 and June 2007. All patients diagnosed with a migraine by the emergency physicians were included in the study. All charts were evaluated by a data abstractor blinded to the study hypothesis using a standardized datasheet. Interrater agreement was measured between the data abstractor and the primary investigator after formal training. Prochlorperazine treatment failure was defined as either an administration of further rescue therapy, a hospitalization or a return visit to the ED within 48 hours for either symptoms recurrence or side effects from the medication. Results:There were 286 episodes of migraine in patients who presented to the ED during the 20 month period. Prochlorperazine was administered in 99 episodes of migraine at a mean dose of 0.14 mg/kg (SD 0.02) all received diphenhydramine in adjunction in order to prevent akathisia. A total of 15 (15%) of these patients had a treatment failure according to our definition: 8 patients received further rescue therapy after the administration of prochlorperazine, 6 patients were hospitalized, including 3 who received further rescue therapy, and 4 patients returned to the ED within 48 hours owing to symptom recurrence. Conclusion: There appears to be a treatment failure rate of 15% with the use of prochlorperazine for severe migraines in children seen in a pediatric ED. Keywords: prochlorperazine, pediatric migraine, treatment failure

68 PRACTICE variation over 10 years in the treatment of migraines in children seen in a pediatric emergency department

Doyon-Trottier E, Bailey B, Dauphin-Pierre S, Dusseault M, Doray JP, Gravel J. CHU Sainte-Justine, University of Montréal, Montréal, QC

Introduction: Several important recommendations on the management of migraine in children were published in the past 10 years. The integration of this new knowledge into clinical practice in a pediatric emergency department (ED) is not known. Objective: To analyze changes that occurred in the management of migraines in children treated in a pediatric ED from 1996/97 to 2006/07.Methods: This study was a comparative retrospective chart review of children diagnosed with migraine in the ED of a tertiary care pediatric hospital. Patients that had a final diagnosis of migraine were eligible. The 2 study periods were from April 1996 to March 1997 and from April 2006 to March 2007. All identified charts were evaluated by a data abstractor blinded to the study hypothesis using a standardized datasheet. Interrater agreement was measured between the data abstractor and 1 of the investigators. Medications used at home, in the ED and upon discharge were evaluated. Results: There were 144 visits to the ED for migraine in the first study period compared to 182 visits in the second one. The absence of treatment at home before the ED visit was more frequent in 1996/97 compared to 2006/07: 69/144 (48%) v. 45/182 (25%) D; 23% (95% CI 12–33), respectively. More patients were not treated in the ED in 1996–97 compared to 10 years later: 101/144 (70%) v. 91/182 (50%), D; 20% (95% CI 9–30). Patient pain was less documented in the charts in 1996/97 compared to 2006/07: 2/144 (1%) v. 131/182 (72%), D; –71% (95% CI –77 to –63). Finally, more patients were sent home without a prescription in 1996–97 compared to 10 years later: 67/144 (46%) v. 29/182 (16%), D; 30% (95% CI 20–40). The type of medications used at home, in the ED and upon discharge also differed.Conclusion: Many changes occurred in the management of migraine headaches from 1996–97 to 2006–07 in our pediatric ED. It appears that changes in management reflect the evolution of recommendations published in the literature as well as the protocol for migraine treatment proposed in our ED. Keywords: pediatric migraine, practice variation, medical records review

69 MARKERS of overcrowding in a pediatric emergency department

Stang A, McGillivray D, Bhatt M, Colacone A, Soucy N, Léger R, Afilalo M. Montreal Children’s Hospital, McGill University Health Centre, Montréal, QC

Introduction: Research on the possible effects of emergency department (ED) overcrowding in childrens hospitals is hampered by the lack of a standardized definition or measure of overcrowding in this setting. The objective of this study was to identify possible markers of overcrowding according to expert opinion and to use statistical methods to extract the more relevant measures.Methods: A survey of all pediatric ED directors (12) and fellowship program directors (10) across Canada was conducted to elicit expert opinion on relevant markers. The list of markers was reduced to those specific to the ED for which data could be collected using a computerized patient tracking system. Data representing 2190 consecutive shifts and 138 361 patient visits between April 2005 and March 2007 were collected retrospectively from one tertiary care pediatric ED. An analytic procedure (Principal Component Analysis [PCA]) was then used to reduce the data and determine the linear combination of markers that best represented the complete data set. Results: The survey had a response rate of 95% and provided measures of patient volume (35%), ED operational processes (46%) and delays in transferring admitted patients to inpatient beds (12%). After discarding redundant markers (correlation coefficient > 0.80), data collected on 29 markers was retained for the PCA. The results of the PCA indicated that the largest portion of variation in the data (25%) was accounted for by markers describing patient flow through the ED. The maximum remaining variability (12%) was explained by patient volume measures. Measures of admission delays only accounted for a relatively small proportion of variability (4%). Conclusion: The results suggest that for pediatric EDs, markers of ED operational processes and patient volume may be more relevant than measures reflecting delays in transferring patients to inpatient beds. This study provides a foundation for further research on measures of overcrowding specific to the pediatric setting. Keywords: ED crowding, pediatric emergency department, survey research

70 A qualitative study of barriers and supports to implementation of metered-dose inhaler (MDI)/spacer use in Canadian pediatric emergency departments

Scott S, Osmond MH, O’Leary K, Grimshaw J, Graham I, Klassen TD. Department of Pediatrics, University of Alberta, Edmonton, AB; Department of Pediatrics, University of Ottawa, Ottawa, ON

Introduction: We aim to determine the barriers and supports to implementing MDI/spacer instead of nebulizers for the provision of β-agonist medications to treat mild-to-moderate acute asthma, and to identify factors associated with early and late adoption of MDI/spacers in Canadian pediatric EDs. Methods: A case study design was used. Sites were classified as late adopters, adopting or early adopters (3 sites each) according to their stage of implementation based on the results of a recent survey. Data at each site were collected using focus group interviews with physicians, nurses and respiratory therapists, and individual interviews with both patient care and medical directors. Data collection and analysis proceeded concurrently using the constant comparative approach, with initial coding being based on the Ottawa Model of Research Use categories of elements known to influence the uptake of innovations. Results: Key facilitators to adoption of MDI/spacer were the presence of a research champion, nurse buy in, and a well-coordinated roll-out of the practice change. Frequently mentioned barriers to MDI/spacer adoption included a perceived cost increase associated with MDI/spacer use, lack of a research champion, perceived increase in workload and little perceived benefit compared to the use of a nebulizer. Late adopter sites were more likely to lack someone willing to champion the adoption of MDI/spacer use and were more likely to cite the higher costs of MDI/spacer use as a reason for not adopting this practice. Early adopter sites were more likely to have participated in research involving MDI/spacer use and more likely to have the MDI/spacer included in a protocol/pathway. Conclusion: Lack of leadership in the form of a research champion along with perceived increased cost associated with MDI/spacer use were the biggest barriers to the adoption of the MDI/spacer. Future interventions intended to increase MDI/spacer use in pediatric EDs will need to address these barriers. Keywords: metered dose inhalers, implementation barriers, qualitative research

71 DISTRIBUTION of chief complaints using the Canadian emergency department information systems chief complaint list

Unger B, Afilalo M, Lang E, Segal E, Colacone A, Xue X, Soucy N. Sir Mortimer B Davis–Jewish General Hospital, McGill University, Montréal, QC

Introduction: The Canadian Emergency Department Information System (CEDIS) chief complaint (CC) codes were introduced in 2002 and linked to the updated Canadian Emergency Department Triage and Acuity Scale (CTAS) in 2005. The CC list was developed with the aim of balancing granularity (detail) with practicality, however, its performance in the emergency department (ED) has not been examined. This study’s objective was to examine the distribution of CCs, which could assist in guiding further refinement of the list. Methods: Retrospective analysis of the ED administrative database of CC codes was performed. Data was entered by triage nurses during the triage process. The study population consisted of adult patients presenting to a tertiary care, level 3 trauma hospital ED. Results: The CCs of 111 143 visits from April 2006 until December 2007 were analyzed. Of the 170 CCs in the list, 168 CCs were used at least once. The CCs Abdominal Pain, Chest pain with cardiac features, Shortness of breath and Lower extremity pain were used in 24% of the visits. There were 156 CCs with less than 2% of the visits and 142 CCs with less than 1%. Overall, 33% (n = 57) of CCs covered together less than 1% of all the visits. Furthermore, 39 different CCs (23% of all CCs) were used in 80% of the visits.Conclusion: Our results indicate that the majority of ED patients visits could be classified using a reduced number of CCs since about one-third of the CCs remained largely unused. Further analysis of this list in multiple settings (including pediatric and trauma) should be performed to determine if the list can be modified to simplify its usage while maintaining the quality of the information gathered. Keywords: chief complaint list, CEDIS, administrative database

72 EMERGENCY medicine residency competency assessment — a national survey

Bishnoi A, Borgundvaag B, Lee S. Mount Sinai Hospital, Faculty of Medicine, University of Toronto, Toronto, ON

Introduction: The Canadian College of Family Physicians offers a 1-year fellowship in emergency medicine (EM) that certifies physicians to practice EM in Canada. Currently, no method exists to confirm that CCFP(EM) residents graduate with the clinical skills and confidence required for the challenges of EM practice. This study aimed to assess the effectiveness of current CCFP(EM) programs by assessing the confidence and self-perceived competence of residents. Methods: In July 2007, 160 surveys were distributed to CCFP(EM) programs using a modified Dillman methodology. The 20-question survey employed 5-point Likert scales to assess self-perception of knowledge and confidence in 2 key educational domains: clinical skills and nonclinical, managerial skills. Survey questions were based on national residency Core Competency Guidelines. The responses were calculated and compared with parametric (Student’s t test) and nonparametric (Mann–Whitney) tests. Results: A representative sample was obtained: 52 of 72 (72.2%) outgoing residents, and 68 of 88 (77.3%) incoming residents. Statistically significant (p < 0.05) improvements was noted in virtually all clinical and nonclinical domains across all CCFP(EM) programs. However, outgoing residents showed no significant improvement in ability to manage paediatric dermatology cases or management of ethical problems. As well, there was a statistically significant difference in residents comfort level with clinical skills as compared to nonclinical, managerial skills. Conclusion:This is the first study to demonstrate that CCFP(EM) residency programs successfully teach virtually all clinical domains outlined in the Core Competency Guidelines. Confidence and perceived competence is lower in paediatric dermatology and ethics. As well, residents are graduating with only an average level of comfort in higher-level, nonclinical managerial skills. These areas constitute important components of EM practice where educational tools can be developed to improve training in these areas.Keywords: competency assessment, postgraduate training, survey research

73 WHAT are the professional development priorities of emergency physicians: a pilot survey

Sherbino J, Upadhye S, Worster A. McMaster University, Hamilton, ON

Introduction: This study sought to determine the professional development priorities (PDPs) of emergency physicians (EPs) using a competency-based framework. Methods: After obtaining institutional ethics approval and informed consent, an electronic survey was distributed to all practicing EPs at 2 university academic centres using a modified Dillman method. Three representative topics addressing each of the CanMEDS Roles (Medical Expert, Communicator, Collaborator, Manager, Health Advocate, Scholar, and Professional) were assessed using a 7-point Likert scale. Results:Of 54 EPs, 37 (68.5%) responded. All were board certified (59.5% CCFP[EM]). The representative topics for Medical Expert, Manager and Scholar were identified by more than 40% of EPs as high, very high or top PDPs. Topics selected as high, very high or top PDPs included keeping up to date with the literature (94.2%), interpretation of diagnostic tests (65.7%), procedural skills (40%), billing strategies (40.0%), new information technology tools (51.4%), bedside teaching of learners (45.7%), and feedback and assessment of learners (47.1%). When asked to select their top 3 PDPs, topics not selected or only selected by a single respondent included communication skills with patients and families, charting, teamwork skills, conflict resolution skills, advocacy skills for patients, health promotion in emergency populations and ability to handle ethical conflicts. Conclusion: To maintain clinical competence, the self-reported PDPs of EPs are focused upon the Medical Expert, Manager and Scholar roles. Faculty development and continuing medical education initiatives may benefit from addressing these needs. However, further research is required to determine why the Communicator, Collaborator, Health Advocate and Professional roles were poorly self-identified. The data does not infer whether these non-PDP topics were rated poorly because of presumed competence in the role or because these roles are not acknowledge by EPs as important for competent clinical practice.Keywords: CME, continuing professional development, needs assessment, CanMEDS

Abstracts: 1-4, 5-24, 25-43, 44-73, 74-103, 104-131