2004 CAEP/ACMU Scientific Abstracts - Poster Presentations (#93 to #111)

CAEP Abstracts

CJEM 2004;6(3):173-212

abstracts:  001-016 | 017-032 | 033-052 | 053-075 | 076-092 | 093-111 | 112-132

093 Validation of the Gothenburg futility criteria for out-of-hospital cardiac arrest presenting with pulseless electrical activity.

Abu-Laban RB, Christenson JM, Innes GD, McKnight RD, van Beek CA, MacPhail IA, Woods R. Division of Emergency Medicine, University of British Columbia, Vancouver, BC

INTRODUCTION: Pulseless electrical activity (PEA) has a poor prognosis and accounts for a significant proportion of presenting cardiac arrest rhythms. Engdahl et al recently evaluated prognostic factors in 1069 PEA cardiac arrests in Gothenburg, Sweden and concluded that no patients over 80 years of age with unwitnessed collapse survived to hospital discharge (Resuscitation 2001;51:17-25). The objective of this study was to evaluate these "Gothenburg Futility Criteria". METHODS: Using an explicit approach, we reviewed data from a prospective cardiac arrest study conducted from 2/12/1998 to 9/30/1999 in a Canadian advanced life support system. The study database contains Utstein-style information on 1583 out-of-hospital cardiac arrest patients. Cases with a presenting rhythm of PEA were described using descriptive statistics and survivors to hospital discharge were evaluated with reference to the Gothenburg criteria. RESULTS: The presenting rhythm was PEA in 363 cardiac arrests (22.9%). Patients with PEA had a mean age of 69 years, and 66.9% were male. The collapse was witnessed in 61.5% of cases; 22.8% received bystander ventilation; and 21.9% received bystander chest compression. Return of spontaneous circulation occurred in 42.7% of cases and 7 patients survived to hospital discharge (1.9%: 95%CI 0.8% to 3.9%). Thirty-six of the PEA cardiac arrests met the Gothenburg criteria of unwitnessed collapse and age over 80 years (9.9%: 95% CI 7.0% to 13.5%), and none of these patients survived to hospital discharge (0%: 95% CI 0% to 9.7%). CONCLUSIONS: We found the Gothenburg Futility Criteria for out-of-hospital cardiac arrest presenting with PEA to be valid. Our findings are limited by wide confidence intervals, but present a compelling argument for further evaluation and, if justified, inclusion of these criteria in EMS "do not resuscitate" guidelines.

Key words: cardiac arrest, resuscitation, pulseless electrical activity

094 Relationship between rapidity of fibrinolytic administration and probability of pulse return in patients given tissue plasminogen activator during pulseless electrical activity cardiac arrest.

Abu-Laban RB, Christenson JM, McKnight RD, van Beek CA, MacPhail IA, Woods R. Division of Emergency Medicine, University of British Columbia, Vancouver, BC

INTRODUCTION: Reports suggest fibrinolysis may aid resuscitation from cardiac arrest, however the only randomized controlled trial of this therapy was negative (the "tPA in PEA Study", Abu-Laban et al, NEJM 2002; 346: 1522-8). It has been suggested that earlier administration of tissue plasminogen activator (tPA) in this study could have led to positive results. We sought to test the hypothesis that the rapidity of drug administration is associated with the probability of return of spontaneous circulation (ROSC) in patients given tPA during cardiac arrest. METHODS: Using an explicit approach, we reviewed data from the tPA in PEA Study database. This database contains Utstein-style information on 117 cardiac arrest patients randomized to receive tPA (100mg IV over 15 minutes) while CPR was ongoing. The relationships between the time from call received to tPA infusion start and ROSC, and time from paramedic arrival to tPA infusion start and ROSC, were evaluated for all cases with documented time data using non-parametric statistics. For secondary purposes, a logistic regression was fit using known survival predictors and time to tPA infusion as covariates. RESULTS: Twenty-five of 117 patients given tPA achieved ROSC (21.4%). The minimum/median/maximum time from call received to infusion start in those achieving ROSC was 25/30/61 minutes vs 16/31/62 minutes for those not achieving ROSC (n = 86, p = 0.82). The minimum/median/maximum time from paramedic arrival to infusion start in those achieving ROSC was 12/23/54 minutes vs 10/24/46 minutes for those not achieving ROSC (n = 112, p = 0.45). The logistic regression showed no evidence of an effect on ROSC related to rapidity of tPA administration (n = 105). CONCLUSIONS: We found no evidence that that the rapidity of drug administration is associated with a higher probability of ROSC in patients given tPA during cardiac arrest resuscitation. It remains undetermined what factors, if any, are predictive of ROSC after fibrinolysis during cardiac arrest.

Key words: fibrinolysis, cardiac arrest, pulseless electrical activity

095 Chest pain emergency department observation unit: a Canadian experience.

Mathieu B, Vadeboncoeur A, Sirzyk J. Emergency Medicine Department, Maisonneuve-Rosemont Hospital, University of Montreal, Montreal, PQ

INTRODUCTION: Chest pain represents both a common complaint in the Emergency Department (ED) and a high-risk challenge for the ED physician. Chest Pain Observation Units (CPU) could allow for better standardization of care, shorter lengths of stay, optimal and cost effective management. We demonstrated that, when applied to a Canadian urban setting, CPU can provide safe and effective management of patients presenting to an ED with chest pain. METHODS: Our protocol consisted of serial EKGs, troponins 9 hours apart, plus either treadmill or MIBI stress test. Inclusion criteria were: (1) age above 18 years, (2) chest pain of low or intermediate risk of coronary artery disease and (3) normal EKG or non-specific ST-T changes. Data were collected prospectively from December 1999 to September 2002. Measured outcomes were: length of stay (LOS), frequency of cardiology consultations, admission rate and final diagnosis. We conducted a 30-day telephone follow-up for safety appraisal and patient satisfaction survey. We used retrospective data for comparison. Final diagnoses were reviewed. RESULTS: A total of 859 patients where evaluated in our CPU protocol. This represents 24.6% of non-ambulatory chest pain stretcher patients. LOS was reduced from 17,1 +/- 10,9 hours before the protocol to 15,8 +/- 1,4 hours; 169 patients (19,6%) were referred for a cardiology consultation, among which 98 were admitted (11%). Troponins were positive in 5,3% of the 859 patients. During the 30-day follow-up period, three patients sustained a myocardial infarction (0,3%) and one had bypass surgery. No patient died from cardiac causes. Patient satisfaction rated from very good to excellent. CONCLUSIONS: Our CPU protocol allows safe and effective management of low to intermediate risk chest pain patients presenting to our ED.

Key words: chest pain, diagnosis, chest pain unit

096 Systematic review of the treatment of post-cardiac arrest patients with mild hypothermia.

Cheung KW, Magee KD, Green RS. Department of Emergency Medicine, Dalhousie University, Halifax, NS

INTRODUCTION: Animal studies and prospective randomized controlled trials in humans suggest that induced mild hypothermia (32-34°C) may be beneficial in resuscitated post-cardiac arrest patients with anoxic brain injury. In this systematic review, we will quantify the extent of this benefit, along with complication rates, in patients treated with mild hypothermia compared to patients treated with normothermia. METHODS: We conducted a comprehensive search of the Cochrane Controlled Trials Register (3rd Quarter, 2003), MEDLINE (January 1966 to December 2003), EMBASE (1980 to December 2003), CINAHL (1982 to December 2003), and Web of Science (1989 to December 2003). There were no language or publication restrictions. Using search terms including cardiac arrest, heart arrest, cardiopulmonary resuscitation, and hypothermia, 1,346 citations were retrieved. Titles and abstracts were screened by one reviewer and 34 articles were selected to be pulled for review. Two reviewers will independently review these 34 articles to assess eligibility for inclusion and methodological quality using both the Cochrane and the Jadad approach. Discrepancies will be adjudicated by a third investigator. Authors of all included studies will be contacted to determine if unpublished studies which meet the inclusion criteria are available and to verify and clarify any missing data. Data will be independently extracted by two reviewers and entered into the Review Manager 4.1 software. For dichotomous variables, individual and pooled statistics will be calculated as odds ratios with 95% confidence intervals (95% CI); a random effects model will be used. For continuous outcomes individual and pooled statistics will be calculated as weighted mean differences or standardized mean differences and 95% CIs using a random effects model. If significant heterogeneity exists (p < 0.1), a sensitivity analyses will be conducted based on methodological quality and on study size. RESULTS: Results will be available for presentation by March 2004.

Key words: cardiac arrest, hypothermia

097 Emergency Department presentations of Atrial Fibrillation (AF) in the United States.

Diner BM, Ratcliff JJ, Pitts SR, Rowe BH. Department Emergency Medicine, Emory University, Atlanta, GA

OBJECTIVES: Atrial fibrillation (AF) is the most common arrhythmia seen in the emergency department (ED); however, the burden of acute AF on EDs is infrequently studied. We describe the epidemiology of AF visits in US EDs using a large administrative database. METHODS: The National Hospital Ambulatory Care Survey (NHAMCS) is a probability sample of US ED visits based on a stratified cluster survey of visits to approximately 486 hospital EDs. Diagnostic categories are recorded using ICD-9 coding and represent the primary diagnostic code. Using the 1997-2000 NHAMCS data, we identified AF visits with ICD 9 codes, 427.3, 427.31, and 427.32. Descriptive statistics, crude and adjusted presentation rates, as well as therapeutic interventions are described. RESULTS: Of the 410 million ED visits in the US between 1997-2000, 1.1 million (95% CI: 0.97, 1.3) were for AF. There are 290,000 (95% CI: 240, 330) AF visits in US EDs annually (incidence: 107/100,000; 95% CI: 89, 122). The overall admission rate was 64 % (95% CI: 56, 71) with 17% (95% CI: 12, 24) to the ICU. Patients over age 65 accounted for 68% of AF admissions. The most common presenting symptom for AF was chest pain, followed by increased heart rate, shortness of breath, palpitations and irregular heart beat. Calcium channel blockers were used more often than beta blockers (42%, 95% CI = 35, 50 vs. 7.8%, 95% CI = 4, 14) to control rate. Only 1% of all AF were electrically cardioverted and 1.5% received procainamide. Overall, 7.6% received heparin of which 72% were > 65 (OR: 1.24, 95% CI = .54, 3.17), 8.4% received warfarin of which 70% were > 65 (OR: 1.13, 95% CI = .22, 4.4) and 11.6% received aspirin of which 72% were > 65 (OR: 1.31, 95% CI = .54, 3.17). CONCLUSIONS: AF is a common ED presentation in the US. ED electrical and chemical cardioversions are rare and admission is common. Anticoagulation and antiplatelet agents are not commonly used in AF patients. Treatment changes may improve resource allocation and improve ED care.

Key words: atrial fibrillation

098 A comparison of methodologic approaches to quantify return of spontaneous circulation (ROSC) in cardiac arrest research including ROSC survival analysis.

Abu-Laban RB, Shuster M, MacPhail IA, Innes GD, Christenson JM, McKnight RD, van Beek CA, Woods R. Division of Emergency Medicine, University of British Columbia, Vancouver, BC

INTRODUCTION: Return of spontaneous circulation (ROSC) is a standard outcome in cardiac arrest research. Although defined as any return of a palpable pulse, even if transient, many researchers apply a time duration when categorizing ROSC. ROSC is often the primary outcome in preliminary studies, however its validity as a treatment effect predictor is unproven and the strength of its correlation with meaningful survival is unknown. Our objective was to determine the impact various common ROSC definitions have on effect size estimates using a common data set, and compare these to a novel approach for ROSC quantification: "ROSC survival analysis". METHODS: Using an explicit approach, effect sizes were determined for various pre-specified common ROSC definitions from a prospective database of 233 cardiac arrests treated with either tissue plasminogen activator (tPA) or placebo in a recently-published trial (Abu-Laban et al, NEJM 2002;346:1522-8). The treatment arms were compared using chi-square or Wilcoxon rank sum tests as appropriate. Kaplan-Meyer curves, censoring survivors to discharge at the point of death of all non-survivors to discharge, were used for the ROSC survival analysis. RESULTS: For tPA vs placebo respectively, the following effect size estimates were found: ROSC of any duration 21.4% vs 23.3% (p = 0.85); ROSC on emergency department arrival 16.2% vs 8.6% (p = 0.12); ROSC > = 20 min 12.8% vs 6.9% (p = 0.13); ROSC > = 1 hr 6.8% vs 6.0% (p = 0.99); ROSC > = 4hr 6.0% vs 5.2% (p = 0.99); median maximum duration of ROSC 28 min vs 11 min (p = 0.45). When plotted, the ROSC survival analysis showed diverging curves favoring tPA (p = 0.18). CONCLUSIONS: Different ROSC definitions produce strikingly different effect size estimates using a common data set. Further research evaluating the predictive and correlational performance of various ROSC methodologies, including those such as ROSC survival analysis that avoid arbitrary categorization and resulting information attenuation, should occur.

Key words: cardiac arrest, outcomes, resuscitation

099 Heparin versus placebo for acute coronary syndromes.

Magee KD, Rowe BR, Campbell S, Moher D. Department of Emergency Medicine, Dalhousie University, Halifax, NS

INTRODUCTION: Despite weak evidence for its use, unfractionated heparin (UFH) is considered an accepted treatment for unstable angina and non-ST segment elevation myocardial infarction (Non-STEMI). We conducted a systematic review to determine the effect of heparin compared to placebo for the treatment of patients with acute coronary syndromes (ACS). METHODS: We searched the Cochrane Controlled Trials Register, MEDLINE, EMBASE and CINAHL for randomized controlled trials of UFH or low molecular weight heparin (LMWH) versus placebo in people with unstable angina or non-STEMI and reference lists of included articles. Authors of all include studies and pharmaceutical industry representatives were contacted to determine if unpublished studies that met the inclusion criteria were available. All subjects were required to receive standard ASA therapy. Two reviewers independently assessed quality of studies and extracted data. RESULTS: Eight studies (3110 participants) were included in this review. We found no evidence for difference in overall mortality between the groups treated with heparins and placebo (RR = 1.01; 95% CI: 0.43, 2.38). Heparins reduced the occurrence of MI (RR = 0.45; 95% CI: 0.30, 0.67). We found no evidence for difference in occurrence of recurrent angina (RR = 0.81; 95% CI: 0.60, 1.09), revascularization procedures (RR = 0.93; 95% CI: 0.76, 1.15), major bleeds (RR = 2.05; 95% CI: 0.91, 4.60), or thrombocytopenia (RR = 0.20; 95% CI: 0.01, 4.24). More patients treated with heparins experienced minor bleeds (RR = 6.80; 95% CI: 1.23, 37.49). Patients treated with LMWH experienced a lower incidence of MI, recurrent angina and required fewer revascularization procedures. CONCLUSIONS: Compared to placebo, heparins when added to standard ASA therapy reduced the incidence of MI in ACS; however, overall mortality was unchanged. Given the advantages of LMWHs, the use of these agents should be the initial therapy of choice early in the treatment of unstable angina and non-STEMI who meet enrolment criteria.

Key words: acute coronary syndrome, myocardial infarction, heparin, anticoagulation

100 What are the most important unanswered questions for pediatric cardiac arrest?

Stiell IG, Nesbitt L, Wells GA, Campbell S, Nadkarni V, Berg R, Osmond M, Donoghue A, De Maio V, for the NIH Pediatric Cardiac Arrest Study Group. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

INTRODUCTION: Pediatric cardiac arrest survival is very poor and resuscitation efforts are hampered by a paucity of evidence from robust clinical trials evaluating therapeutic options. We surveyed experts to determine research priorities for planning such trials in pediatric cardiac arrest. METHODS: Our international, NIH-funded, resuscitation group conducted this prospective mail survey using a modified version of Dillman's Total Design Method. We selected 241 experts from professional organizations in 15 countries. Recipients rated the research priority, using a 5-point Likert scale, for each of 47 potential study questions presented in 4 domains (EMS Systems, Gas Exchange, Circulation, Post Resuscitation). We distributed the survey 4 times, twice each by email and by regular mail. The results were tabulated with simple descriptive statistics. RESULTS: The 109 returns (45.2% completion) were from U.S.A. 50.5%, Canada 38.5%, Europe 2.7%, Australia 4.6%. Respondents were 14.6% EMS professionals and 85.4% physicians: emergency medicine 45.0%, pediatrics 42.2%, internal medicine 4.6%, anesthesia 3.7%, cardiology 1.8%. The following had the 10 highest proportions of priority 1 and 2 responses regarding pediatric cardiac arrest survival:

75.0% What is the incremental value of prehospital CPR, defib- rillation, and ALS?

73.9% What is the effectiveness of bag-valve-mask vs. intuba- tion?

73.4% Does epinephrine improve outcome?

69.4% Does post resuscitation hypothermia improve outcome?

65.8% Does chest compression before defibrillation improve out- come?

65.2% Does vasopressin improve outcome?

61.7% What is the efficacy of CPR-only for the 1st 3 minutes?

59.6% Do AEDs improve outcomes?

58.5% Is end-tidal CO2 effective for ETT tube placement?

58.3% Is compression-only CPR as effective?

 CONCLUSIONS: This is the first international survey of research priorities for this under-studied area and the results will considerably assist the planning of large, multicentre clinical trials for pediatric cardiac arrest.

Key words: cardiac arrest, pediatric

101 What are the characteristics and outcomes of non-transported pediatric patients?

Kahale J, Osmond MH, Stiell IG, Nesbitt L, Maloney JP, Trickett J. Department of Emergency Medicine, University of Ottawa, Ottawa, ON

INTRODUCTION: We recently showed that 28% of children attended by paramedics are not transported to hospital by ambulance. This study aims to determine the characteristics of this non-transported population. METHODS: This was a prospective cohort study in a single city with a 2-tiered EMS system. Enrolled were all children <16 years seen but not transported by paramedics to hospital over a 5 month period. We collected data from Ambulance Call Reports(ACRs), phone interviews and hospital charts. We used descriptive statistics. RESULTS: Over 5 months there were 345 non-transported pediatric patients. Mean age = 6 (range = 0-15); 58.3% male; Dispatch Priority: urgent 68.1%, prompt 30.4%; Pick-up location: residence 58.6%, street 21.7%, public place 10.4%; Primary problem: trauma 50.7% (normal exam 36.6%, laceration 22.9%, contusion 13.7%) medical 45.2% (respiratory 30.7%, GI/vomiting 10.9, seizure 9.0%) Procedures performed: BLS 15.1%, ALS 3.5% Reason stated by paramedics for not transporting patient: parent will take child to MD 27.8%, parent will monitor child's condition 25.8% Phone interview: 106 parents (30.7%) participated. 76.4% believed there was a true emergency at the time of the 911 call. 75.5% stated that paramedics did not recommend that the child be brought to the hospital by ambulance. Hospital Charts: We confirmed that 51 of the 345 children were seen at an ED within 48 hours of the 911 call. Triage at the ED: Semi-urgent 47.8%, Urgent 39.1% and Emergent 6.5%; Outcomes: 68.6% discharged without follow-up, 15.8% discharged with follow-up, 7.8% admitted to the hospital. Average length of stay for those admitted was 2 days. No deaths were reported. CONCLUSIONS: This is the largest study of non-transported children following EMS attendance. The majority of patients did not require immediate or urgent medical care and had good short-term outcomes. We recommend that EMS services review their handling of non-transported children to ensure good outcomes.

Key words: emergency medical services, pediatric

102 Prospective validation of the MET-AP clinical decision support system for pediatric emergency department triage of acute abdominal pain.

Farion KJ, Michalowski W, Rubin S, Wilk S, Correll R. Childrens Hospital of Eastern Ontario, Ottawa, ON

INTRODUCTION: Mobile Emergency Triage (MET) is a Clinical Decision Support System (CDSS) developed to assist clinicians assessing pediatric acute abdominal pain (AP). It recommends one of three triage dispositions: 1) Discharge for benign causes; 2) Consult surgery for possible appendicitis; and 3) Observe/investigate for other serious pathology. Running on a PDA, the clinician enters values for up to 13 clinical attributes and can invoke a triage recommendation at any time. The main objective was to determine the proportion of times the MET triage recommendation agrees with the final diagnosis category, compared to the triage prediction of treating residents (RES) and physicians (PHYS). METHODS: A prospective cohort trial of children (1-16 y) at a tertiary Pediatric ED (PED) with acute AP (<10 d) was conducted. The enrolling RES or PHYS collected patient attributes on a PDA and entered their independent prediction of the appropriate triage category after initial assessment. An alternate observer provided secondary assessments when possible. Patients were contacted at 10-14 days with chart review to determine the correct triage recommendation corresponding to the final diagnosis. Personnel were blinded to the MET recommendation until a final category was assigned. RESULTS: 500 patient enrollments and 332 follow-ups of 640 are completed, with expected conclusion in early 2004. Sample includes 12% appendicitis, 74% benign and 12% other pathology. Accuracy of MET (PHYS) is 68% versus PHYS prediction 68% (n = 259). MET (RES) and RES accuracies are both 61% (n = 210). Consult versus non-consult accuracy of MET (PHYS) is 88% (Sens 70%, Spec 91%) versus PHYS 91% (Sens 67%, Spec 94%). CONCLUSIONS: A CDSS for PED triage of acute AP has been successfully implemented. The accuracy of MET used early in the patient s assessment appears to be similar to specialty trained PED physicians. Further refinement with the prospective data is expected to improve the MET accuracy. Future research will assess MET in non-PED settings.

Key words: abdominal pain, diagnosis, pediatric, decision support

103 The effect of an electronic clinical practice guideline on the management of acute otitis media in the emergency department.

Bullard MJ, Choi D, Holroyd BR, Meurer D, Rowe BH. Division of Emergency Medicine, University of Alberta, Edmonton, AB

INTRODUCTION: Acute otitis media (AOM) is a common pediatric emergency presentation. An interactive electronic prescription-writing tool and discharge instructions were developed for AOM based on the provincial clinical practice guideline (CPG). We evaluated whether EPs who used the electronic CPG (eCPG) would comply with the guidelines more frequently than those who managed from memory. METHODS: Design: Using the hospital database and an AOM eCPG database we compared those AOM patients treated with and without accessing the eCPG. A standardized audit form was used for the chart review. Setting: University tertiary referral center. Subjects: Otitis media patients < 16 years of age. Control cases were randomly selected from a list of patients treated at the same pediatric ED between January 2002 and April 2003. Patients were excluded if they were admitted or left without being seen. Data Analysis: Groups were compared using chi-square, ANOVA and K-W tests. RESULTS: There were 56 non-CPG patients and 11 eCPG patients who met inclusion criteria over 3 months. Of these, 16 patients (15 non-CPG and 1 eCPG) were treated conservatively or given prescriptions to fill in 48 hours, after follow up with the Family Physician or no improvement. Based on the guidelines, those physicians using the eCPG were more likely to correctly select amoxicillin as their first choice (p = 0.06; OR = 6.0, 95% CI: 0.67, 134); have a clearly documented, guideline appropriate dose (p = 0.01; OR = 13.1, 95% CI: 1.43, 303); and also have a documented, guideline appropriate dosing duration (p = 0.001; OR = 23.3, 95% CI: 2.35, 565). CONCLUSION: For patients with AOM who met guideline criteria for receiving antibiotics, physicians using the eCPG were significantly more likely to select the appropriate antibiotic agent, dose, and duration compared with those who did not employ the eCPG. Use of eCPGs in the ED may reduce medical error and improve quality of care.

Key words: otitis, clinical practice guideline

104 Sedation and analgesia for incarcerated inguinal hernia in the paediatric emergency department.

Goldman RD, Wales P, Balasubramanian S, Mace S. Division of Paediatric Emergency Medicine and the Division of General Surgery, The Hospital for Sick Children, Department of Paediatrics and Surgery, University of Toronto, Toronto, ON

INTRODUCTION: Ten percent of children with inguinal hernia will experience incarceration and a reduction procedure will be required. The procedure involves pain and discomfort in most cases. No recommendations on the level of sedation or analgesia for children exist. METHODS: We conducted a formal mail survey among paediatric emergency physicians (131) and paediatric surgeons (45) in eight centers across Canada. The survey consisted of multiple choice questions to determine the use of sedation or analgesia for incarcerated hernia reduction and the length of the trials. Up to 3 surveys were sent. Data was entered into Microsoft Excel software and descriptive statistics were done with the SPSS program. RESULTS: A total of 118 (67%) of the physicians responded. Eighty-four (71%) stated that they would perform the first reduction trial with no sedation or analgesia. If another trial was needed, 74 (62%) would have given sedation or analgesia, mostly through the intravenous route. Fentanyl and Midazolam were the most frequently prescribed medications. The first attempt at reduction would last 5 ± 5 minutes (range 1-30); waiting between trials was suggested to be on average 14 ± 10 minutes long (range 1-45) and the number of attempts the responders would carry out was on average 2 ± 1 trials (range 0-3). CONCLUSIONS: Paediatric emergency physicians and paediatric surgeons do not use a unified protocol for reduction of incarcerated hernia in the paediatric emergency department and significant practice variation exists. Effort should be made to create and disseminate a protocol for sedation and/or analgesia during this urgent procedure.

Key words: hernia, procedural sedation

105 Status migrainosus in children: a 3 year experience with prochlorperazine.

Cummins McManus B, Bailey B. Section of Emergency Medicine, Department of Pediatrics, Hôpital Ste-Justine, Montreal, PQ

BACKGROUND: There is limited evidence suggesting that prochlorperazine is effective and safe in the treatment of status migrainosus in children. Evidence in adults suggests that the prevalence of akathisia is high. OBJECTIVE: To describe the use of prochlorperazine in the treatment of status migrainosus in children. METHODS: From Jan/00 to Dec/02 all patients presenting to a pediatric emergency department (PED) diagnosed with status migrainosus and treated with prochlorperazine were followed prospectively 24 hours after discharge with a standardized telephone questionnaire. Our protocol recommends IV prochlorperazine 0.15 mg/kg (max 10 mg) with IV diphenhydramine 0.5 mg/kg (max 25 mg). The subjective part of an akathisia scale (0 to 9) was used: a score of 2 or more was considered positive. Parents were asked to rate the pain associated with the migraine on a 0 to 10 scale on presentation to the PED, at discharge, and if a relapse occurred after discharge. RESULTS: Over a three-year period 51 patients (29 female/22 male) aged 12 ± 3 years were treated with prochlorperazine. The pain associated with the migraine on presentation was 8.4 ± 1.3 while at discharge it was 1.5 ± 1.8. Out of 51 patients, 18 (35%) had a relapse 14 ± 12 hours after discharge with reported pain level at 4.3 ± 2.4. Three parents reported their child as having abnormal movements described as shaking or tic. Six patients were reported as being restless, nine as being unable to stand or sit still, and five as being unable to keep their legs still. On the akathisia scale, 13 patients had a score of 1 or more, but only six (12%) patients could be said to have had akathisia with a score of 2 or more (range 2-6). CONCLUSIONS: Prochlorperazine appears to be effective in the treatment of status migrainosus in children. Even with the use of diphenhydramine as prophylaxis, it is associated with some extrapyramidal side effects. The prevalence of akathisia appears to be lower than reported in adults; however more objective and subjective assessment is needed.

Key words: migraine, prochlorperazine, pediatric

106 Variability in investigation and management of transient ischemic attacks among Canadian emergency physicians.

Harris DR, Christenson J, Johnston D, Connolly H, Innes G, Grafstein E, Stenstrom R. Department of Emergency Medicine, St. Pauls Hospital and University of British Columbia, Vancouver, BC

INTRODUCTION: A transient ischemic attack (TIA) confers a high short-term risk of debilitating stroke. This study assessed Canadian emergency physicians' patterns of practice in the management of TIA, and their attitudes toward a clinical prediction rule to aid in the management of patients with TIA. METHODS: An electronic and standard mail survey was administered to 467 randomly selected members of the Canadian Association of Emergency Physicians between November 2002 and March 2003. Three common clinical scenarios were presented and respondents were asked to choose between various management options. Opinions toward a clinical prediction rule were obtained. RESULTS: The response rate was 60.1%. 80.7% were male with a mean age of 40 (SD 7.3). For the patient with atrial fibrillation and recurrent hemispheric TIA (scenario 1), 71.0% of respondents would order a computed tomography (CT) scan at the index visit, 21.3% a carotid ultrasound and 28.3% an echocardiogram. 63.5% of respondents would consult a specialist prior to discharge and 39.7% would elect to admit the patient. For the patient with a single hemispheric TIA (scenario 2), 60.9% would order a CT and 37.7% a carotid ultrasound. 35.9% would consult a specialist prior to discharge whereas only 12.5% would admit. For the patient with transient monocular blindness (scenario 3), 46.2% would order a CT in the emergency department and 22.3% a carotid ultrasound; most would discharge (80.4%) for an outpatient workup. 35.5% of respondents felt uncomfortable with their current management of TIA and 91.3% felt management could be improved at their hospital. 92.5% felt a clinical prediction rule for TIA would be useful or very useful. CONCLUSIONS: This survey clearly shows the high practice variability in the investigation and management of patients with TIA. This may reflect variability in the clinical accuracy of estimated short-term stroke risk in each of the scenarios. A clinical prediction rule may aid in the management of these patients.

Key words: transient ischemic attack, stroke, clinical practice guideline

107 Prevention of stroke after transient ischemic attack by Canadian emergency physicians.

Harris DR, Christenson J, Johnston D, Connolly H, Innes G, Grafstein E, Stenstrom R. Department of Emergency Medicine, St. Pauls Hospital and University of British Columbia, Vancouver, BC

INTRODUCTION: Recent studies show that the risk of stroke after transient ischemic attack (TIA) is significant. We hypothesized that there exists much variability in clinical management of patients with TIA among Canadian emergency physicians despite published practice guidelines, and that most emergency physicians underestimate short-term stroke risk after TIA. METHODS: An electronic and standard mail survey was administered to 467 randomly selected members of the Canadian Association of Emergency Physicians between November 2002 and March 2003. Three common clinical scenarios were presented and respondents were asked to choose between various treatment options. Estimates of short-term stroke risk were obtained from each respondent. RESULTS: The response rate was 60.1%. 80.7% were male and 67.2% in full-time practice. For the patient with atrial fibrillation and recurrent hemispheric TIA, 85.7% would prescribe warfarin, 58.1% heparin and 55.6% acetylsalicylic acid (ASA) while 27.3% would add an angiotensin-converting enzyme inhibitor (ACE-I) and 19.3% an HMG-CoA reductase inhibitor (statin). For the patient with a single hemispheric TIA, 87.8% would prescribe ASA, 11.7% would add clopidogrel, 24.6% would prescribe an ACE-I and 14.1% a statin. For the patient with transient monocular blindness, 85.8% would prescribe ASA, 11.3% ASA and persantine (AggrenoxTM), and 8.9% clopidogrel alone, while 11.8% would add an ACE-I and 11.8%, a statin. In scenario 1, 36.6% felt the risk of stroke within 90 days was <10%, whereas 37.8% felt the risk was >30%. In scenario 2, 72.5% felt the 90-day risk of stroke was <10% and in scenario 3, 79.5% felt the 90-day risk was <10%. CONCLUSIONS: Considerable variability in practice patterns exist among Canadian emergency physicians treating patients with TIA and most physicians underestimated the short-term risk of stroke. This study highlights the need for further research and education into treatment of patients with TIA.

Key words: transient ischemic attack, stroke

108 Student patient encounters during a mandatory emergency medicine rotation.

Leger M, Magee KD, Ackoryd S. Department of Emergency Medicine, Dalhousie University, Halifax, NS

INTRODUCTION: Since the recognition of Emergency Medicine (EM) as a free standing specialty in 1980, many medical schools have sought ways to incorporate EM as a core component of clerkship. In 1999, Dalhousie University implemented a unique mandatory EM rotation whereby students were longitudinally scheduled to complete 18 to 20 shifts throughout their third year. To our knowledge, this study represents the first attempt to quantify the volume and breadth of the patients encountered by medical students during a mandatory EM rotation. METHODS: Throughout the 2001-02 academic year, third year clerks prospectively kept logs of their clinical experience in the Emergency Department (ED). A computer database was developed to analyze this experience. The patients' presenting complaints were grouped into one of the following thirteen categories:

1. Altered level of consciousness

2. Altered Mood

3. Head, ears eyes, nose and throat (HEENT)

4. Chest pain

5. Shortness of breath (SOB)

6. Abdominal pain

7. Nausea, vomiting, diarrhea and fever

8. Perineal disorder

9. Musculoskeletal

10. Lacerations, abrasions, bites and burns

11. Rash, Cysts, abscess and cellulites

12. Other

13. No information given

RESULTS: A preliminary convenience sample of 28 medical students representing 3065 individual patient encounters has been entered into the database. Students completed an average of 16.1 ED shifts during their clerkship and saw 7.2 patients per shift. The average age of patients seen by clinical clerks was 33.2 years with a range of 8 days to 100 years. The male to female ratio of patients was 1:1.1. Patients with musculoskeletal complaints were the most frequently encountered (25.6%) by medical students, followed by HEENT (13.8%), SOB (11.4%), lacerations, abrasions, bites and burns (10.0%) and abdominal pain (9.3%). CONCLUSION: A mandatory rotation in EM affords medical students the opportunity to experience a diverse range of undifferentiated patient encounters in a primary care setting while be supervised directly by a staff physician.

 

109 Effects on a review course on emergency medicine residents' self-assessed confidence.

Blouin D, O'Connor HM. Department of Emergency Medicine, Queen's University, Kingston, ON

OBJECTIVES: To determine whether a review course offered for the first time to Canadian postgraduate year 5 residents (PGY5s) in emergency medicine (EM) to guide their study for certification exams increased or decreased their self-assessed confidence in 1) their knowledge of EM and its application 2) their planned study strategy to master knowledge and its application and 3) their recognition of when knowledge and its application have been mastered. METHODS: Before and after study. 16 PGY5s (40% of all Canadian PGY5s) from various Canadian EM programs who attended the course held in October 2003 prospectively completed a pre- and post-course questionnaire on the above-listed 3 aspects of self-confidence, in 22 clinical domains. Domains reflected the listing of conditions that present to emergency departments as outlined in "The Model of Clinical Practice of EM" published by ABEM Core Content Task Force II. A 5-point Likert-type scale was used for each question. Pre and post-course means were compared overall and for each question in each domain using Wilcoxon Signed-Rank Tests for matched pairs. RESULTS: Analysis included data from all 16 participants. The course significantly increased PGY5s' overall self-confidence (mean difference 0.33, W+ = 136, W- = 0, p = 0.00003), as well as their confidence in knowledge and application of 9 clinical domains, in their study strategy to master knowledge and application of 12 domains, and in their recognition that mastery of knowledge and its application has been achieved for 2 domains. Domains showing increased self-confidence did not directly match domains specifically addressed in the didactic portion of the course suggesting that the course effects extend beyond the selected topics. CONCLUSIONS: A review course increased EM PGY5s' overall self-confidence and confidence in EM knowledge and application and in study strategy beyond the course specifics.

 

110 Performance of emergency medicine postgraduate year 5 residents on visual stimuli stations during a practice exam.

Blouin D, McGraw RC. Department of Emergency Medicine, Queen's University, Kingston, ON

OBJECTIVES: Emergency Medicine (EM) postgraduate year 5 residents (PGY5) from various Canadian residency programs attended a review course to guide their preparation for their upcoming certification exams. Visual stimuli were a component of the course. PGY5s' answers to visual stimuli questions were analyzed to identify deficient areas of knowledge. METHODS: Prospective observational design. 6 sessions of 8 visual stimulus stations were presented during the October 2003 course. Stimuli comprised ECG, X-rays and pictures. 1 or more questions was asked at each station. PGY5 in groups of eight rotated every 3 minutes from 1 stimulus to the next. Each PGY5 anonymously wrote his/her answers to the questions on an index card and put the card in an envelope before leaving the station for the next one. 16 PGY5 (40% of all Canadian PGY5s) attended the course. For analysis purposes, stimuli were classified according to the domains of EM listed in "The Model of Clinical Practice of EM" by ABEM Core Content Task Force II. Each answer was rated as correct or incorrect; two experienced EM attendings provided gold standard answers. RESULTS: Data were available for all 16 residents in all 48 visual stimulus stations. There were 73 questions in total covering 9 domains: 7 in respirology, 6 dermatology, 22 cardio, 5 ophtalmology, 24 trauma, 3 ID, 3 toxicology, 2 MSK, 1 urology. Answers were correct for 70.1% of questions with the following distribution: 60.7% respirology, 36% dermatology, 84.1% cardiology, 76.3% ophthalmology, 66.2% trauma, 79.2% ID, 83.3% toxicology, 59.4% MSK, 50% urology. Only cardiology and trauma domains had enough questions to accept results with a 15% margin of error and 90% confidence (17 questions needed). CONCLUSIONS: 6 months prior to certification exams, PGY5s' trauma knowledge is low whereas cardio knowledge is good. Other domain sample sizes are too small to comment. A larger number of questions is needed in selected domains so that inferences on PGY5s' knowledge and need for defined study focus.

 

111 What's in it for us? Faculty and resident perceptions about formal teaching sessions.

Bandiera G, Morrison L. Division of Emergency Medicine, University of Toronto. Toronto, ON

BACKGROUND: Little is known about factors affecting emergency medicine (EM) faculty attendance at academic teaching sessions. Few reports detail what faculty believes to be the benefits from attending these activities. OBJECTIVE: To determine a) which factors affect EM faculty attendance at academic half day, and b) what faculty perceive to be the strengths and shortcomings of attending these activities. METHODS: A multiple choice and open short answer survey was developed, piloted on five random faculty members and revised. The survey was sent to all academic faculty members in the divisions of emergency medicine and all residents. Responses were tabulated dichotomously (yes/no) for multiple choice answers and analyzed using a two person grounded theory approach for open answers. Differences between full-time and part-time faculty and between faculty and residents were compared using the Chi-square test for significance. RESULTS: Response rates were: faculty 73.8% (48/65), and residents 55.5% (10/18). Significant impediments to attendance at rounds included: clinical, professional and personal responsibilities (75%, 33.3%, 52.1%); location of rounds (31.2%); and time of rounds (27.1%). Perceived benefits of attending were: continuing medical education (CME), social interaction, teaching opportunities, interaction with the residents, comparing one's practice with peers, improving teaching techniques, and enjoyment of the format. There were no statistically significant differences between part-time and full-time faculty in the distribution of responses. CONCLUSIONS: EM faculty benefits from CME, social interaction, and teaching opportunities when attending teaching activities. Faculty feel that they both teach and learn during these sessions. The majority of faculty report competing responsibilities and accessibility as impediments to attending. Efforts to increase faculty attendance should focus on enhancing the social interaction and educational development of faculty rather than appealing to their sense of obligation.