2008 CAEP/ACMU Scientific Abstracts - Oral Presentations: 25-43

2008 Scientific Abstracts

CJEM 2008;10(3):255-295

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

25 AN international survey for the management of high risk transient ischemic attacks in the emergency department

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

Introduction: Up to 10% of patients with transient ischemic attack (TIA) suffer a stroke or die within 1 week of their initial TIA. Our study objectives were to assess for patients with TIA: 1) current emergency department (ED) practice; 2) need for a clinical decision rule (CDR) to identify those at high risk of an impending stroke or death; and 3) the required sensitivity for such a CDR. Methods: We administered a mail survey with 31 questions to a random sample of members of 3 national emergency physician associations, ACEM (Australia), CAEP (Canada) and ACEP (USA), using a modified Dillman technique. A prenotification letter and up to 3 surveys were sent. Descriptive statistics were calculated. Results: Overall, 801 responses from 1493 surveys were received with individual country response rates of 55%, 60% and 46% for Australia, Canada and the United States, respectively. Of physicians, 76% were male with a mean age of 41.5 years of age, with 12 years of clinical practice and 75% of physicians had formal emergency medicine credentials. Ninety percent of respondents have 24-hour on-site access to CT scanners. Ninety-one percent of physicians indicated that they always or most often investigate TIA patients with a CT Head and 95% routinely order an ECG. Twenty-nine percent reported that most or all of their patients are admitted. Thirty-five percent of ED physicians reported having a dedicated stroke prevention clinic to follow their patients; however, only 17% estimated that most patients are seen within 7 days from referral. A CDR to identify highrisk TIA patients would be used by 97% of ED physicians surveyed. The median required sensitivity of a CDR for stroke or death within 7 days was 97% (IQR 99%, 95%) (98% for both Canada and US, and 97% for Australia). Conclusion: Most TIA patients are treated as outpatients, which is often neither expedited nor in a dedicated stroke prevention clinic. ED physicians support the need for a highly sensitive CDR to identify TIA patients at a high risk for impending stroke or death within 7 days of TIA diagnosis. Keywords: transient ischemic attack, survey, clinical prediction guide

26 DEFINING normal jugular venous pressure on bedside ultrasound: an anatomic survey

Socransky S, Wiss R, Robins R, Yeung C. Northern Ontario School of Medicine, Thunder Bay, ON

Introduction: Elevated jugular venous pressure (JVP) may be useful diagnostically in dyspneic patients. Unfortunately, physical examination for JVP (E-JVP) is unreliable. JVP measurement with ultrasound (U-JVP) is easy, but normal values have not been defined. The study objective was to define normal U-JVP. Methods: This was a prospective anatomic survey conducted on a convenience sample of voluntary ED patients over 35 years old. Exclusion criteria were presenting complaints, physical exam findings, past medical history or medications which would possibly lead to an alteration of JVP. With the head of bed at 45 degrees and the legs parallel to the ground, the point at which the diameter of the internal jugular vein (IJV) began to decrease on ultrasound (the taper) was marked. UJVP was measured in two ways on all participants: 1) the vertical height (cm) of the taper above the sternal angle plus 5 cm; and 2) the IJVs path from the clavicle to the angle of the jaw was divided into four quadrants. The quadrant in which the taper was located was recorded. The measurements were conducted by three medical students trained by the principal investigator. A subset of participants had U-JVP measured by 2 students to determine interrater reliability.Results: U-JVP could be determined in all 38 male and 39 female participants. Mean U-JVP was 6.35 cm (95% CI 6.00–6.59). In 76 patients (98.8%), the taper was located in the first quadrant. Fifteen patients had their U-JVP measured by two students. κ = 1.0 for method (1). κ = 0.87 for method (2). Conclusion: Normal UJVP is 6.35 cm consistent with published normal E-JVP. U-JVP can be determined in all patients. Interrater reliability is excellent. The top of the IJV column is located less than 25% of the way from the clavicle to the angle of the jaw in the vast majority. Keywords: bedside ultrasound, jugular venous pressure, shock

27 PREVALENCE of chest x-ray findings requiring intervention in emergency department patients with possible acute coronary syndrome

Hess EP, Perry JJ, Ladouceur P, Stiell IG. University of Ottawa, Ottawa, ON

Introduction: Chest x-rays (CXRs) are routinely performed in patients with chest pain to rule out a non-cardiac etiology. We sought to determine the prevalence of CXR findings requiring intervention in patients with chest pain and possible acute coronary syndrome (ACS). Methods: This prospective study conducted in an academic tertiary care emergency department enrolled consecutive patients ≥ 25 years of age with a primary complaint of chest pain and possible ACS over 5 months. Exclusion criteria were: acute ST-segment elevation, hemodynamic instability, cocaine use, terminal illness, or pregnancy. Thirty-six clinical variables were collected on a standardized data collection form prior to radiography. All CXRs were interpreted by attending radiologists blinded to the data collection forms. Two investigators independently classified CXRs as normal, abnormal requiring intervention or abnormal not requiring intervention based on review of the radiology report and the medical record. Analyses included descriptive statistics and inter-rater reliability (κ). Results: There were 541 patients enrolled. Patient characteristics included: mean age 59.6 years, 40.3% female, 50.5% hypertension, 19.0% diabetes, 3.0% congestive heart failure and 35.5% ischemic heart disease. Of patients, 40.7% reported shortness of breath, 9.6% had an abnormal lung examination (crackles, rales or wheezes) and 6.8% pitting edema; 11.6% experienced MI, revascularization or death within 30 days. CXRs were obtained in 83.4% of patients, 2.7% (1.5–4.7) of which had an abnormality requiring intervention (κ for CXR classification = 0.78, 95% CI 0.61–0.96).Conclusion: The prevalence of CXR findings requiring intervention was low. These data suggest that a clinical decision rule for CXR in patients with possible ACS is feasible and has potential to decrease costs, radiation exposure, and length of stay in our busy EDs. Keywords: chest radiograph, acute coronary syndromes, clinical prediction guide

28 PERCEIVED barriers to emergency medicine ultrasonography by emergency physicians

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

Introduction: Emergency Medicine Ultrasonography (EMUS) has been identified as a part of current EM practice, but it has not been universally adopted. We set out to identify potential barriers to the use of EMUS by emergency physicians. Methods: As part of the ETUDE study, we conducted a web-based survey of a sample of the Canadian Association of Emergency Physicians membership database using a modified Dillman technique. The 24-item instrument assessed demographics, current and potential EMUS use, and barriers to EMUS use. Barriers were identified from the literature and a pilot study. Ethics approval was obtained. Pediatric emergency physicians, residents, and medical students were excluded from analysis. Descriptive and chi-square statistics were used to analyze the data. Results: The response rate was 36.4% (296/814). 72.5% were male with a median age of 40. 51.9% of respondents graduated after 1995. Certification of respondents was CCFP(EM) (52.3%) and FRCP(EM) (29.3%). Three items were not considered barriers to EMUS use: consultant attitudes lack of impact on patient management, and ease of getting a formal ultrasound. Respondents practicing in inner city/urban/suburban settings were more likely to endorse heavy clinical workload as a barrier to using EMUS vs. small town/rural/remote EDs (p < 0.05). Female respondents felt that not enough training was a barrier (p < 0.05). Respondents born before 1965 compared with those born after 1965 felt that easy access to CT scans was a barrier (p < 0.05). Part-time physicians reported more barriers compared to full-time physicians (p < 0.05) including clinical workload, difficulty recording scans, inadequate access to formal training, supervision and review of findings with credentialed staff, and inability to complete 200 scans in a timely manner. Conclusion: This is the first national survey to characterize the barriers to adopting EMUS into practice. Leaders in EMUS will need to address these issues in order to enhance care in this domain. Keywords: bedside ultrasound, implementation barriers, survey

29 SOCIOECONOMIC status influences bystander CPR and survival rates for out-of-hospital cardiac arrest victims

Lui A, Vaillancourt C, De Maio VJ, Wells GA, Stiell IG. University of Ottawa, Ottawa, ON

Introduction: While lower socioeconomic status is associated with lower level of education and increased incidence of cardiovascular diseases, the impact of socioeconomic status on out-of-hospital cardiac arrest outcomes is unclear. We used residential property values as a proxy for socioeconomic status to determine if there was an association with 1) bystander CPR rates; and 2) survival to hospital discharge for out-of hospital cardiac arrest. Methods: We performed a secondary data analysis of cardiac arrest cases prospectively collected as part of the Ontario Prehospital Advanced Life Support study which was conducted in 20 cities with ALS and BLS-D paramedics. We measured patient and system characteristics for cardiac arrests of cardiac origin, not witnessed by EMS, occurring in a single residential dwelling. We obtained property values from the Municipal Property Assessment Corporation. Analyses included descriptive statistics with 95% CIs and stepwise logistic regression. Results:3600 of 7707 (46.7%) cardiac arrest cases met our inclusion criteria between Jan. 1, 1995, and Dec. 31, 1999. Patient characteristics were: mean age 69.2 years, male 67.8%, witnessed 44.7%, bystander CPR 13.2%, VF/VT 33.8%, time to vehicle stop 5:36 min:sec, return of spontaneous circulation 12.7% and survival 2.7%. Median property value was $184 000 (range $25 500 – $2 494 000). Patient and system characteristics were similar across property values. For each $100 000 increment in property value, the likelihood of receiving bystander CPR increased (OR 1.07, 95% CI 1.01–1.14; p = 0.03) and survival decreased (OR 0.77, 95% CI 0.61–0.97; p = 0.03).Conclusion: This is the first study to show an association between socioeconomic status and bystander CPR and survival rates for outof- hospital cardiac arrest victims. Future studies should determine the nature and quality of the bystander CPR being performed in residential dwellings. Our findings might be used to target CPR training among lower socioeconomic groups. Keywords: socioeconomic status, cardiopulmonary resuscitation, advanced life support

30 CAEP Critical Care Committee Sepsis Position Statement

Green RS, Djogovic D, Gray S, Howes D, Brindley P, Stenstrom R, Patterson E, Easton D, Davidow J. Dalhousie University, Halifax, NS

Introduction: Optimal management of severe sepsis in the ED has evolved rapidly. The purpose of this guideline is to review key management principles for Canadian emergency physicians, utilizing an evidence based grading system. Methods: Key areas in the management of septic patents were determined by members of the CAEP Critical Care Committee (C4). Members of C4 were assigned a question to be answered after literature review, based on the Oxford grading system. After completing, each section underwent a secondary review by another member of C4. A tertiary review was conducted by C4 as a group, and modifications were determined by consensus. Grading was based on peer reviewed publications only, and where evidence was insufficient to address an important topic, a practice point was provided based on group opinion. Results: The project was initiated in 2005, and completed in December 2007. Key areas which area reviewed include the definition of sepsis, the use of invasive procedures, fluid resuscitation, vasopressor/ inotrope use, the importance of cultures in the ED, antimicrobial therapy, and source control. Other areas reviewed included the use of corticosteroids, activated protein C, transfusions and mechanical ventilation. Conclusion:Early sepsis management in the ED is paramount for optimal patient outcomes. The CAEP Critical Care Committee Sepsis Position Statement provides a framework to improve the ED care of this patient population. Keywords: sepsis, clinical practice guideline, CAEP

31 A universal termination of resuscitation clinical prediction rule for both advanced and basic life support providers

Morrison LJ, Allan KS, Zhan C, Kiss A, Verbeek PR. Pre-hospital and Transport Medicine Research Program, Sunnybrook Health Sciences Centre, Division of Emergency Medicine, Department of Medicine and the Department of Health Administration, University of Toronto, Toronto, ON

Introduction: Different termination of resuscitation (TOR) rules have been retrospectively derived for advanced life support (ALS) providers and prospectively validated for basic life support (BLS) providers. To determine whether either of these rules could stand as a universal pre-hospital TOR rule, we validated each on a prospectively collected cohort of out of hospital cardiac arrest (OHCA) patients attended by both levels of providers. Methods: We performed a secondary cohort analysis of data prospectively collected for the Resuscitation Outcomes Consortium Epistry trial from Apr. 1, 2006, to Mar. 31, 2007. The BLS and ALS TOR rules were applied and the diagnostic test characteristics as well as the predicted transport rate for each rule were calculated. Results: Both TOR rules were applied against 2410 patients with presumed cardiac etiology OHCA. The ALS TOR rule recommended TOR for 744 patients and none of these patients survived. The ALS TOR rule had a specificity of 100% for recommending transport of potential survivors and had a positive predictive value of 100% for death when TOR was recommended. Implementation of the ALS TOR rule would result in a transport rate of 69%. The BLS TOR rule recommended TOR for 1306 patients and none survived. The BLS TOR rule had a specificity of 100% for recommending transport of potential survivors and had a positive predictive value of 100% for death when TOR was recommended. Implementation of the BLS rule would result in a transport rate of 46%. Conclusion: Both TOR rules have high specificity and positive predictive values when TOR was recommended. Implementing the BLS TOR rule for all providers would result in a lower overall transport rate without missing any survivors. These findings may be useful for EMS systems that wish to implement TOR procedures for OHCA patients attended by ALS or BLS providers. Keywords: termination of resuscitation, advanced life support, basic life support, clinical prediction guide

Winner of the CAEP Resident Research Abstract Award

32 UTILITY of impedance cardiography in the early diagnosis and management of emergency department shock patients

Kashyap R, Campbell R, Sztajnkrycer M, Stead LG, Smith V, White R, Afessa B, Gajic O, Hess E. Mayo Clinic College of Medicine, Rochester, MN

Introduction: Impedance cardiography (ICG) is a noninvasive technology that measures hemodynamic parameters. We sought to determine the utility of ICG in the early diagnosis and treatment of emergency department (ED) patients with shock. Methods: This study was conducted at a tertiary care academic emergency department (90 000 visits/yr). ED patients with shock were prospectively enrolled over 2 years. Inclusion criteria were: SBP < 90 mm Hg or a drop > 40 mm Hg from baseline, shock index (HR/SBP) > 1, or lactate > 3 without evidence of liver disease. Patients < 16 years of age and trauma patients were excluded. Thirty-five clinical and ICG variables were collected on a standardized data collection form. Treating physicians were interviewed to determine the predominant shock type (distributive, hypovolemic, cardiogenic/obstructive) and initial treatment plan before and after viewing the ICG data. Critical care experts blinded to the collected data determined the reference standard shock type and therapeutic plan by review of the entire electronic medical record. Results: Seventy-five patients were enrolled. Patients mean age was 67.7 years, 56% male, mean BMI 26.9. According to the reference standard shock type 52% of patients had distributive, 39% hypovolemic, 4% cardiogenic/obstructive and 5% no shock. ICG resulted in a change in the emergency physicians diagnosis in 20% of cases and intervention in at least one of the 4 categories (volume resuscitation, antibiotic, inotropic support and vasopressors) in 17.3%. Pre-ICG and post-ICG agreements of ED physicians with expert were 58.7%; κ = 0.30 (95% CI 0.12–0.47) and 66.7%; κ = 0.46 (95% CI 0.29–0.62). ICG did not significantly improve the accuracy of the ED Physician diagnosis (p = 0.083). Conclusion: 1) A nonsignificant trend toward increased agreement in the diagnosis of shock type and management was observed. 2) ICG is a simple, noninvasive tool that may improve risk stratification of ED patients presenting with shock.3) Further studies are needed to determine ICG role more clearly. Keywords: impedance cardiography, hemodynamics, shock

33 PATIENTS who are discharged from the ED with referral to outpatient specialty clinics, but do not complete the referral: demographics, reasons for nonattendance, and barriers to care

Vergel de Dios J, Hanneman K, Friedman SM. Department of Family and Community Medicine, University of Toronto, Toronto, ON

Introduction: To characterize patients who are referred from the ED to specialty clinics but do not complete the referral, and to identify reasons for failure to follow-up. Methods: Retrospective cohort study over three months of patients who were discharged from the ED of a teaching hospital with referral to internal medicine, cardiology or neurology clinics but did not complete follow-up. A standardized telephone interview assessed demographics, barriers to care and reasons for not completing the consultation. Results: Of 171 ED referrals, 42 (24.6%) were not completed. Interviews were completed for 71.4% (n = 30). Nonattenders were functional in English (80%) and educated (73.1% were high school graduates, 60.7% university graduate); 93% of interviewees could get to hospital by themselves or have someone take them. Only 42.9% (n = 12) understood why the emergency physician (EP) requested consultation, and 42.9% (n = 12) described EP instructions as poor or fair. Most followed up with another physician (72.4%, n = 21), generally family doctor (51.7%, n = 15) or another specialist (17.2% n = 5). Primary reasons for noncompletion of consult were patient choice (46.7%), physical or social barriers (13.3%), communication failure (20%) and consultants’ refusal of the consultation (20%). All clinic refusals were from one internal medicine clinic, representing 42% (8/19) of EP referrals to that clinic. No interviewed patient (6/6) who was declined consultation was aware of his/her status.Conclusion: Clinic follow-up requested by the EP is often not completed. Approximately one-half of consult noncompletion was attributable to communication and system failures, rather than patient choice.Keywords: outpatient referral, continuity of care, ED follow-up

34 PATIENTS who leave prior to the completion of acute care: further evidence of overcrowding in emergency departments

Rowe BH, Yiannakoulias N, Bullard MJ, Holroyd BR, Voaklander DC, Rosychuk RJ. University of Alberta, Edmonton, AB

Introduction: Patients who leave without being seen (LWBS) or against medical advice (LAMA) constitute potential problems for emergency departments (ED) owing to patient dissatisfaction and potential morbidity. Few jurisdictions can accurately evaluate the pattern of LWBS/LAMA for a large population; this study examines this ED phenomenon over a 6-year period in 1 Canadian province. Methods: All patients presenting to Alberta EDs were eligible for inclusion. Data were derived from a sample of ED patients treated in 17 health regions over 6 years (1998/99–2004/05) with a disposition code of LWBS or LAMA. Data were extracted from the Ambulatory Care Classification System (ACCS) database, a computerized abstract system coded similarly across all regions. All data elements are recorded by medical record nosologists and LWBS can be differentiated from LAMA by the absence of a treating ED physician code. Descriptive statistics and crude presentation rates are reported. Results: Between 1999/2000 and 2004/05, ED visits increased from 1.49 to 1.8 million. Over the study period LWBS rates were orders of magnitude higher (30/1000 compared to LAMA (~ 4/1000). The annual rates of LAMA and LWBS both increased; the average increase in rates was 10.8% and 12.6% per year, respectively (p = 0.64). A larger proportion of LAMA patients were classified as CTAS one-half than LWBS cases (p < 0.001). Death at follow-up following LWBS and LAMA did not differ (p = 0.581); however, LAMA patients were more likely to be admitted to an ED (p < 0.001) and were hospitalized more rapidly following ED discharge than LWBS (p = 0.029) patients. Conclusion: These results indicate that LWBS and LAMA cases across a large population are increasing annually. LWBS cases are orders of magnitude higher than LAMA and can be considered a proxy marker for ED overcrowding. LAMA is a less common problem overall; however, readmissions for LAMA are higher than LWBS cases. Interventions to reduce LWBS and LAMA in EDs appear warranted. Keywords: left without being seen, left against medical advice, ED crowding

35 STATE of the nation: a profile of Canadian EM clerkships 2007

Frank JR, Penciner R, Upadhye S, Nuth J, Lee AC. University of Ottawa, Ottawa, ON

Introduction: It is a marker of the growth of emergency medicine (EM) as a specialty that EM content is rated as essential in the curricula of Canadian medical schools. However, no national profile of EM teaching in MD programs exists. We set out to characterize the current state of EM undergraduate medical education in Canada.Methods: We conducted a cross-sectional online survey of EM clerkship directors (CDs) or their equivalent at Canada’s 17 medical schools. We used Dillman methods with up to 6 contacts. Ethics approval was obtained. We used descriptive statistics to analyze the data. Results: Two schools confirmed they had no EM clerkship or equivalent. Of 15 schools, 12 responded (80%). Of these 12, 10 (83.3%) schools had a mandatory clinical rotation in EM, with 60% of these occurring in third year of training. EM rotations varied in length from 4 weeks (50%) to 2 weeks (40%). The number of teaching sites per school varied from 1 to 6. Clerkship features included clinical shifts (100%), seminars (90%), procedure skills (70%), lectures, simulations, ACLS (all 60%), web modules, EMS (40%), student presentations (30%) and journal club (20%). Student assessment methods varied, but included written tests (22%), daily encounter cards (22%), clinical evaluations (20%), OSCEs (11%), observed procedures (9%), simulations (6%) and assignments (3%). Most EM clerkship faculty were not directly compensated for clinical teaching (80%), and were evaluated only at the end of the clerkship (41%) or at the end of each teaching activity (36%). Senior residents taught in 9/10 clerkships. Clerkship budgets varied from $10 000 to $350 000. Ten of 12 EM undergraduate directors had salary support and dedicated a mean 1.6 days/week (range 0.25–2.5). Schools took a mean of 42 elective students per year (range 4–150). Conclusion: This is the first national profile of the state of EM teaching in medical schools in Canada. In 2007, EM teaching was extremely heterogeneous. National standards for quality EM UGME are needed. Keywords: undergraduate medical education, clerkships, undergraduate curriculum

36 METHODOLOGIC quality of the American Board of Emergency Medicine Life-Long Self Assessment reading lists

Upadhye S, Carpenter C, Diner B, Fan J. Division of Emergency Medicine, McMaster University, Hamilton, ON

Introduction: The American Board of Emergency Medicine (ABEM) uses an annual reading list of articles as the basis of the annual online examination for its Lifelong Self-Assessment (LLSA) program. This study aimed to examine the methodologic quality of listed articles, using accepted standard hierarchies of evidence.Methods: The Oxford Centre for Evidence-based Medicine Levels of Evidence scales of evidence hierarchy in multiple study categories was adapted to 10-point Likert scales (1 = highest level of evidence, 10 = lowest). Articles in each LLSA year (2004–2007) were rated using these scales. Articles were also rated using the McMaster Online Rating of Evidence (MORE) Likert scales (7 points) for relevance and newsworthiness. Descriptive statistics for each year’s ratings were obtained as well as reliability measures. The Oxford scales were selected on the basis of widest representation of study designs, and the MORE scales based on proven reliability for content assessment.Results: A total of 79 articles were independently reviewed by 3 blinded raters. The interrater reliability for overall ratings across 3 scales was 0.83 (95% CI 0.77–0.81). Average relevance and newsworthiness scores were high over 4 years (6.0 and 5.0, respectively). Average quality ratings were poor over 4 years (7.5 out of 10). A small trend towards higher quality scores were noted in later years. Conclusion: Articles included on the LLSA reading lists from 2004 to 2007 are relevant and newsworthy to the practice of emergency medicine, but represent a low standard of evidence. Readers should be cautious in using such information to guide clinical practice. Certifying bodies should re-examine their article screening and selection processes to ensure that high quality articles are disseminated for recertification purposes and to optimize patient care.Keywords: American Board of Emergency Medicine, critical appraisal, Lifelong Self-Assessment

37 OLD dogs and new tricks — formalizing feedback around the CanMEDS Roles

Bandiera GW, Lendrum D. St. Michael’s Hospital and the University of Toronto, Toronto, ON

Introduction: In-training evaluations often contain halo and leniency bias yet must address all seven CanMEDS competencies. Our previous work has shown that teachers evaluate some roles much more than others using an emergency medicine daily encounter card model. We sought to explain the reasons why this might be true and explore options for improvement. Methods:Semistructured focus groups at 2 large academic teaching hospitals were completed by a third-party moderator until the data was saturated. A dual reviewer grounded theory analysis generated themes.Results: Sixteen faculty participated in 4 focus groups; data saturation was achieved. Teachers felt that the Medical Expert, Communicator and Professional roles were easy to assess, applicable to EM rotations and important, while Health Advocate, Scholar, Collaborator and Manager had less obvious applicability and objectivity. Barriers to assessment included lack of “buy-in” from trainees, time constraints and lack of observable behaviours for each role. Participants did, however, appreciate the structure that the CanMEDS roles provided for feedback sessions. A follow-up workshop demonstrated that education about the Roles can improve understanding and create insight into the practical assessable behaviours associated with each Role. Conclusion: The CanMEDS roles are abstract for many teaching faculty in our centre. Teachers felt the roles were useful as a framework but had difficulty providing feedback on objective behaviours for each role. Opportunity exists to improve use of the CanMEDS roles in resident assessment through education about the operational aspects of each role. Keywords:CanMEDS, trainee assessment, focus group

38 PROSPECTIVE validation of the Ottawa Ankle Rule and the Malleolar Zones Algorithm for children with ankle trauma

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

Introduction: The Ottawa Ankle Rule (OAR) and the Malleolar Zone Algorithm (MZA) have been developed to identify children needing a radiography following an ankle trauma. Objective: To determine the criterion validity of the 2 rules for predicting significant ankle fractures in children. Methods: This was a prospective cohort study performed in a university-affiliated pediatric emergency department (ED). Children younger than 16 years suffering from an acute ankle trauma were eligible. A priori, physicians were instructed in how to assess patients according to the rules. The ED physician implemented each rule with the aid of a standardized diagram. Ankle radiography and referral to orthopedics were left at the physicians’ discretion. Final diagnosis was made at follow-up at the orthopedic clinic. A phone follow-up was made for those who needed no orthopedics follow-up. All fractures other than a nondisplaced Salter I fracture of the fibula were considered significant. The primary outcome was the presence of a significant fracture as confirmed in the orthopedic follow-up. The criterion validity was calculated for each rule. Results: Radiography was performed in 245 (90%) of the 272 participants. All patients with no radiograph were reached by phone. Eighty patients had an ankle fracture including 33 with a Salter I fracture of the fibula and 47 with a significant fracture. The sensitivity and specificity for a significant fracture were 1.00 (95% CI 0.94–1.00) and 0.27 (95% CI 0.24–0.30) for the OAR, and 0.87 (95% CI 0.84–1.00) and 0.24 (95% CI 0.21–0.27) for the MZA. Conclusion: The OAR identified all children with a significant fracture, whereas the MZA showed a lower sensitivity. The use of the OAR would decrease by 10% the number of requested ankle radiography in our setting whereas the use of the MZA would not decrease it. Keywords: Ottawa Ankle Rule, Malleolar Zone Algorithm, clinical prediction guide

39 LONG-TERM outcomes of a cluster randomized knowledge transfer trial in the ED

Stiell IG, Clement C, Grimshaw J, Brison R, Rowe BH, Schull MJ, Lee JS, Brehaut J, McKnight RD, Dreyer J, Eisenhauer M, MacPhail I, Rutledge T, Letovsky E, Shah A, Clarke A, Ross S, Perry JJ, Wells GA; for the CCC Study Group. University of Ottawa, Ottawa, ON

Introduction: We previously conducted a randomized trial in which implementation of the Canadian C-Spine Rule (CCR) by physicians led to significantly lower ED use of c-spine imaging (CSI) without missed injuries. We sought to evaluate the long-term impact of the CCR and conducted a prospective 12-month follow-up at the same EDs. Methods: The original trial had a matched–pair cluster design that compared outcomes during two 12-month before and after periods at 6 intervention and 6 control EDs, stratified by teaching or community hospital. We randomly allocated sites to intervention or control groups and enrolled alert, stable adult trauma patients. During the intervention-site after-period, active strategies were employed to implement the CCR. During the follow-up period, all intervention strategies were removed and sites were unaware that we continued to collect data. We used univariate analyses, appropriate for a cluster design. Results: The 11 947 patients were similar, comparing after (n = 6144) to follow-up (n = 5803) periods, and control (n = 4339) to intervention (n = 7608) sites, with mean age 38.7 years, female 50.1%, MVC 69.7%, important c-spine injury 0.9%. At the 6 intervention sites, CSI rates dropped from before to after periods (61.7% v. 53.8%; p < 0.01, RR 12.8%) and remained low in the follow- up period (53.1%). At the 6 control sites, CSI rates increased from before to after (53.8% v. 60.5%; p < 0.05, RI 12.5%) and remained higher in the follow-up period (61.7%). There was variation in CSI rates among intervention (40.9%–61.4%; p < 0.05) and control sites (50.2%–67.3%; p < 0.05). There were no missed c-spine injuries.Conclusion: This trial demonstrated that actively implementing the CCR led to significantly lower CSI use by physicians without missed injuries and that this effect was sustained for at least 12 more months without interventions. Widespread implementation of the CCR could reduce health care costs and improve patient flow in busy EDs over a long term period. Keywords: knowledge translation, implementation, long-term outcomes

Winner of the CAEP Resident Research Abstract Award

40 OTTAWA Ankle Rules to exclude fractures in the pediatric population: a systematic review and meta-analysis

Dowling SK, Spooner C, Liang Y, Dryden DM, Friesen C, Klassen TP, Wright RB. University of Calgary, Calgary, AB; Alberta Research Centre for Child Health Evidence, Department of Pediatrics, University of Alberta, Edmonton, AB

Introduction: Although the role of the Ottawa Ankle Rules (OAR) is routinely accepted in adults, the accuracy of the OAR in children is less well established. The primary objective of this report was to conduct a quantitative systematic review to determine the diagnostic accuracy of the OAR to exclude fractures of the ankle and midfoot in children. Methods: We searched electronic databases (i.e. MEDLINE, EMBASE and CINAHL) for published and gray literature and hand searched reference lists. Two reviewers independently applied inclusion/exclusion criteria (children [< 18 yr] presenting to the ED with an acute blunt ankle or midfoot injury, x-ray or proxy measure to confirm/exclude a fracture and sufficient data to create a 2 × 2 table). Data were extracted by one reviewer using a standard form and verified by a second reviewer. The data was then pooled using an approximation of the inverse variance approach; 95% confidence intervals were calculated using the exact method. Results:We included 12 studies, which enrolled 3226 patients with 552 fractures (prevalence 17.1%). The pooled sensitivity was 98.2% (95% CI 96.7–99.1) and specificity was 29.5% (95% CI 27.8–31.3). The pooled negative and positive predictive value were 98.8% (95% CI 97.7–99.4) and 22.3% (95% CI 20.7–24.1), respectively. Of the 10 missed fractures, only 4 were characterized; 1 was a Salter-Harris (SH) I fracture, 1 was a SH-IV fracture and 2o were insignificant (either SH-I or avulsion fractures < 3 mm). Pooled estimate for x-ray reduction rates was 24.8% (95% CI 23.3–26.3). There was evidence of funnel plot asymmetry (p = 0.002), suggesting that there may be missing studies with smaller sample sizes but more significant results. Conclusion: Based on the findings from our study, the OAR appears to be a reliable tool to exclude fractures in children presenting with ankle and midfoot injuries. The usage of the OAR in children would significantly decrease radiography usage with a very low likelihood of missing a fracture. Keywords: Ottawa Ankle Rule, pediatric ankle injury, systematic review

41 PASSIVE versus active distraction for intravenous catheterization in the pediatric emergency department

Goldman RD, Koller D, Wan T, Bever S, Stuart S. PRETx Program, Division of Pediatric Emergency Medicine, BC Childrens Hospital, Department of Pediatrics, University of British Columbia, Vancouver, BC; Child Life Program, The Hospital for Sick Children, Department of Pediatrics, University of Toronto, Toronto, ON

Introduction: Children visiting the ED experience anxiety and pain during procedures. In order to ameliorate the effects of pain and anxiety, many hospitals have Child Life Specialist programs. The objective of this study was to compare the efficacy of active versus passive forms of distraction in decreasing anxiety and pain in children 5–18 years old undergoing IV insertion, the most common procedure in the pediatric ED. Methods: This was a prospective, randomized trial comparing anxiety and pain response among children being active (experimental) or passive (control) during intravenous catheter insertion in the ED and while being supported by a Child Life Specialist. Outcome measures were recorded by a research assistant (RA) and included Observational Scale of Behavioral distress- Revised (OSBD-r) at baseline, immediately before IV insertion( s), immediately after IV insertion attempt(s) and every 2 minutes in between. The RA also recorded the Faces Pain Scale- Revised (FPS-r) for each angiocath insertion attempt. The number of IV attempts and the length of the procedure were also recorded.Results: A total of 79 (89%) children completed the study. Children were 10 (SD 3) years old (range 5–18). Thirty-nine and 40 were randomized to the active and passive groups, respectively. Ten (13%) children that were allocated to the passive group were subsequently offered active forms of distraction to accommodate their perceived needs. There was no significant difference in mean OSBD-r and FPS-r between the groups. More children in the active group had a successful first attempt to insert an IV compared to the active group (89% v. 63%, respectively, p = 0.01) and the mean length of procedure was faster in the active group (1.2 v. 1.8 min, respectively, p = 0.02). Conclusion: Active child participation in the distraction procedure by a Child Life Specialist during angiocath insertion appears to be more effective than passive forms of distraction, particularly as they relate to fewer attempts for successful cannulation. Keywords:distraction, IV catheter insertion, randomized trial

42 DOES the Broselow Tape accurately estimate the weight of Canadian children?

Knight J, Lubell R, Milne WK. South Huron Hospital and University of Western Ontario, London, ON

Introduction: The Broselow Tape (BT) has been used for more than 20 years as a length-based estimate of body weight for children during resuscitation. Increasing childhood obesity has raised concern that the BT no longer accurately estimates weight. The purpose of this study was to validate the BT in Canadian children. Methods: A convenient sample of children attending their community pediatrician’s office between October and December 2007 was used. Their actual weight was compared with their BT weight estimates using the Bland–Altman method. The correlation coefficient and percentage error (PE) was also calculated. Results: A total of 499 children were included in the study. The mean age was 2.58 years (95% CI 2.31–2.85), mean height was 88.42 cm (95% CI 86.15–90.70), mean weight was 14.4 kg (95% CI 13.61–15.20) and mean BT weight was 13.31 kg (95% CI 12.62–14.00). Bland–Altman difference was 1.1 kg (–3.9 to 6.1). The correlation coefficient (r) was 0.9645 (p < 0.001, 95% CI 0.958–0.970). The BT had a PE > 10% one-third of the time and > 20% error 8.6% of the time. Conclusion: The BT was often not accurate and tended to underestimates the weight of Canadian children. Keywords: Broselow Tape, pediatric resuscitation, weight measurements