Introduction: The chaotic environment of the emergency department has a deleterious effect on clinical judgement. The diagnosis of abdominal pathology is difficult to differentiate. There are also many diagnoses that could be considered abdominal in nature, exacerbating the task of diagnosing these patients. We propose a novel machine-learning method, Hierarchical Structured Models (HSMs), to provide an adjunct to clinician judgement, that provides a ranking of the probabilities of a patient having each of 39 abdominal pathologies, using only variables at the triage stage of emergency department care, and compare its performance to several machine-learning methods. Methods: This was a retrospective analysis of 25,861 patients that presented with one of 39 ICD-9 abdominal pathologies. 90% of the data was used to build and fine-tune the model, and 10% was used for testing. Predictors included age, gender, triage vitals and presenting complaint. All variables were solely collected from the Emergency Department Information System (EDIS). A decision tree structure was built using hierarchical clustering algorithms, and then a support vector machine (SVM) was fit at each node. To optimize the parameters for each node, a grid-search method was used to maximize ten-fold classification accuracy. The output of the decision tree was the probability of a particular presentation having each of the 39 diagnoses. This output was translated to a ranking of the relative likelihood of each of the diagnoses as a suggestion system for the treating physician. The accuracy of the system on the test set was compared to conventional machine-learning methods: pair-wise SVMs, gradient boosted models (GBM), neural networks (NN) and k-nearest neighbours (KNN). Results: The HSM ranked the correct diagnosis first 51.0% of the time, and ranked the correct diagnosis within the top three ranks 67.6% of the time. The most accurate model was GBMs (52.3%), and the least was neural networks (50.4%). Conclusion: The HSM approach using only variables available electronically at triage successfully ranked the correct diagnosis 51.0% of the time, and within the top three 67.6% of the time. Future research will focus on the inclusion of clinically lab results and radiology reports that are available electronically to improve HSM accuracy, and supplement physician diagnosis.

Introduction: Many emergency departments (EDs) have begun publishing wait times. This study seeks to develop an understanding of patients’ needs with respect to publishing ED wait times, which, to our knowledge, has not been described in the literature. Methods: We conducted a two-stage mixed methods study at a dual campus tertiary care academic center. First, we held focus group discussions comprising of 7 patient advocacy hospital committee members. Themes generated from focus group discussions were then utilized to create a patient survey. Focus groups were analyzed using content theme analysis. Hospital sites for survey administration were randomized and pre-assigned shifts were established to ensure a balance of weekdays, weekends, days, evenings, and overnights. All adult patients (age >18) in the waiting room were eligible, but excluded if they were directly referred to a specialty service or did not speak French or English. Survey data was analyzed using descriptive statistics. Results: We found 9 dominant focus group themes: definition of wait time, wait time posting, lack of communication, education in waiting room, patient expectations, utilization of the ED, patient behavior, physical comfort, and patient empowerment. Of the 240 patient questionnaires administered, 81.3% (195) wanted to know ED wait times before arrival to hospital and 90.8% (217) wanted ED wait times posted in the ED waiting room. The most popular choice for publishing wait times outside the ED was a website (46.7%) whereas, within the ED, patients were not particular about the specific display modality as long as times were displayed (39.6%). Overall, 76.7% (184) stated their satisfaction with the ED would be improved if wait times were posted. Conclusion: ED patients we surveyed strongly supported both the idea of having access to wait time information prior to arrival, as well as physical display of wait times in the waiting room.

Introduction: Paramedics are sometimes called for crisis management and relief of symptoms or for patients receiving palliative care. To address the mismatch between the system protocols and resources, and patient’s goals of care, a new protocol, new medications, and an 8-hour training program Learning Essentials Approach to Palliative Care (LEAP) were implemented in our provincial EMS system. Methods: Prior to attending their training session paramedics received an invitation to complete an online survey regarding their comfort, confidence, and attitudes toward delivering palliative care. Comfort and confidence questions were scored on a 4-point Likert scale, while attitudes toward specific aspects of care were scored on a 7-point Likert scale. Descriptive statistics were calculated. Identifiers will permit linkage of these responses to a repeat survey post-implementation. Results: 188 (58%) paramedics completed the survey of the 325 who opened the link. 134 (68%) were male with a mean age of 38.5 years. 95 (50%) were primary care paramedics. The average experience as a paramedic was 12.7 years, with an estimated mean number of palliative calls per year of 9.6 each. On a 4 point scale, most (156, 83%) were comfortable with providing care to someone with palliative goals, and 130 (69.1%) were comfortable providing care without transport. Only 82 (43.6%) were confident they had the tools to deliver this care, and 76 (40.4%) were confident they could do so without transport to hospital. On a 7 point scale, paramedics disagreed with the statement “caring for dying persons is not a worthwhile experience for me”, median 7 (IQR 5-7). Paramedics also disagreed with the statement “Dying persons make me feel uneasy”, median 5 (IQR 4-6). Conclusion: Prior to the implementation of the new protocol, medications, and training, most paramedics were comfortable with the concept of providing care with palliative goals and felt that caring for dying persons is a worthwhile experience, but they were not confident that they have the tools and resources to do so. This suggests paramedics would be open to system improvements to meet an unmet healthcare need for crisis management of patients with palliative goals of care.

Introduction: The goal of this study was to determine if emergency department (ED) surge and end of shift assessment of patients affect the extent of diagnostic tests, therapeutic interventions performed and accuracy of diagnosis prior to referral of patients to Internal Medicine as well as the impact on patient outcomes. Methods: This study was a health records review of consecutive patients referred to the internal medicine service with an ED diagnosis of heart failure, COPD or sepsis, at two tertiary care EDs. We developed a scoring system in consultation with senior emergency and internal medicine physicians to uniformly assess the treatments and investigations performed for patients diagnosed in the ED with heart failure, COPD or sepsis. These scores were then correlated with surge levels and time of day at patient assessment and disposition. Rate of admission and diagnosis disagreements were also assessed. Results: We included 308 patients (101 with heart failure, 101 with COPD, 106 with sepsis). Comparing middle of shift to end of shift, the overall weighted mean scores were 92.2% vs. 91.7% for investigations and 73.5% vs. 70.0% for treatments. Comparing low to high surge times, the overall weighted mean scores were 89.9% vs. 92.6% for investigations and 68.6% vs. 71.7% for treatments. Evaluating each condition separately for investigations and treatments according to time of shift or surge conditions, there were no consistent differences in scores. We found overall high admission rates (93.1 % for heart failure, 91.1% for COPD, 96.2% for sepsis patients), and low rates of diagnosis disagreement (4.0 % heart failure, 10.9% COPD, 8.5% sepsis). Conclusion: We found that surge levels and end of shift did not impact the extent of investigations and treatments provided to patients diagnosed in the emergency department with heart failure, COPD or sepsis and referred to internal medicine. Admission rates for the patients referred were above 90% and there were very few diagnosis disagreements or diversion to alternate service by internal medicine. We believe this supports the emergency physician's ability to adapt to time and surge constraints, particularly in the context of commonly encountered conditions.

Introduction: Currently the top 5% of complex patients consume 84% of Ontario’s Hospital and Home Care costs. There is a critical need for a dynamic, person-centred care planning process for medically complex patients with real time dialogue between ED/acute care and community care providers at care transitions. A care pathway was developed in the Sunnybrook Health Science Centre’s Emergency Department using quality improvement methodology and team. The purpose of this study is to evaluate the impact of the emergency room huddle for complex care patients on emergency doctors’ perceptions of patient safety and ED efficiency measures such as department flow and delays. Methods: Intervention - Medically complex patients with frequent ED use are now automatically flagged upon registration in the Emergency Department (ED) and an ED Care Coordination team is notified by secure email: GEM nurse, ED CCAC Care Coordinator, SW, OT/PT. The GEM nurse initiates a comprehensive patient assessment in the Emergency Department right after triage and the CCAC Care Coordinator initiates a teleconference with the patient’s family physician and community Care Coordinator with the patient’s consent. Usual physician assessment is preceded and followed by an inter-professional huddle (including the EM doctor, GEM nurse, CCAC nurse and SW, OT, PT) to ensure patient's needs, goals and team recommendations are clear. Emergency doctors who have participated in an inter-professional huddle for complex care patients are contacted via a semi-structured interview and Qualtrics surveys evaluating perceptions of patient safety and ED efficiency measures such as department flow and delays. Results: Qualitative analysis of the results will be conducted and results updated at a later date. Conclusion: Safety is enhanced through better communication between ED providers, patients, their family physicians and community care providers. It is essential that the inter-professional huddle is recognized by emergency physicians as an important element of patient safety and care. An evaluation of ED doctor’s perception of the huddle will help us understand enablers and barriers to the process and inspire further quality improvements to enhance patient care.

Introduction: In order to achieve the best possible outcomes for patients requiring resuscitation (PRRs) in the emergency department (ED), health care providers (HCPs) must provide an efficient, multi-disciplinary and coordinated response. A quality improvement (QI) project was undertaken to improve HCP response to PRRs at two tertiary care hospital EDs in Toronto. Methods: We conducted a before-and-after mixed-method survey to evaluate the perception of the adequacy of HCP response and clarity of HCP role when responding to PRRs. The results were compared using the Chi-square test. Qualitative responses to the first survey were also used to inform the development of the QI project. Through interviews of key stakeholders and with continuous input from front-line ED HCPs, a multi-disciplinary team modified the ED resuscitation protocol. This included standardized pre-hospital communication form with paramedics, ED-wide overhead announcement of ‘Code Resus’, dedicated HCPs assigned to respond to PRRs, and specific duties assigned to each responder. Change initiatives were reinforced through education and posters in the ED. Six months after implementation, a second survey was conducted to evaluate the sustained effects of the intervention. Results: Baseline measures indicated that 16 of 52 (30.8%) nurses surveyed believed their role was often or always apparent to themselves and others when they attended to a PRR (on a 5-point rating scale). This proportion increased to 35 of 55 (63.6%) nurses in the post-implementation survey (p < 0.001). Regarding adequacy of the number of HCPs responding to PRRs, 17 of 39 (43.6%) physicians and 23 of 53 (43.4%) nurses surveyed thought the appropriate number of HCPs responded to PRRs; the remainder thought that there were too few or too many HCPs. In the post-implementation survey, 34 of 41 (82.9%) physicians (p < 0.001) and 36 of 56 (64.3%) nurses (p = 0.029) surveyed felt that the appropriate number of HCPs attended to PRRs. Conclusion: Using a quality improvement approach, we identified and quantified perceived deficiencies in HCP response to PRRs in the ED. Through feedback-based modifications of the ED resuscitation protocol and by engaging HCP stakeholders, change initiatives were implemented to improve HCP response. As a result, this project achieved significant and sustained improvements in HCPs’ perceived response to PRRs.

Introduction: Podcasts have become increasingly popular as a medium for free online access medical education (FOAM). However, little research has examined the naturalistic use of podcasts as a tool in undergraduate medical education. This study aims to determine usage conditions, preferences, and level of retention of information from podcasts by medical students at a Canadian University. Methods: Medical students (Years 1 to 3) were instructed to complete an online test assessing their baseline knowledge on the topics of the podcasts and for qualitative data on podcast usage and preferences. Audio podcasts on two topics (adult asthma, and introduction to toxicology) were then distributed to study participants. One week and two weeks after the initial survey students were asked to complete a follow-up survey for knowledge assessment and further podcast usage data. Simple descriptive statistical generated using Microsoft Excel. Paired samples t-tests were utilized to assess knowledge acquisition using Microsoft SPSS version 23. Results: Participants who successfully completed the knowledge assessments demonstrated a significant effect of learning (Asthma, average test score improvement of 30%, p=0.002; Toxicology, average test score improvement of 13%, p=0.004). The majority of participants who stated a preference in podcast length indicated they preferred podcasts of 30 or less minutes (85%). The top three activities participants were engaged in while listening to the podcasts were driving (46%), completing chores (26%), and exercising (23%). A large number of participants who did not complete the study in its entirety cited a lack of time and podcast length to be the top two barriers to completion. Conclusion: This is one of the first studies to examine podcast usage data and preferences in a Canadian undergraduate medical student population. This information may help educators and FOAM producers to optimize educational tools for medical education.

Introduction: Emergency department (ED) transthoracic echocardiography (TTE) is an important application of emergency department bedside ultrasound. Given limited curricular hours and economic constraints, training using ultrasound simulators represents an attractive alternative to using live-human models. Despite increased uptake of ultrasound simulator technology, educators lack evidence informing how best to use this technology. Three educational paradigms will be explored in this study: self-guided theory (learners are able to determine when they have had “enough practice”), desirable difficulties (manipulating practice conditions to create more durable and flexible learning), and the challenge point framework (avoiding cognitive overload). The question we seek to answer is: in novice medical trainees, which practice condition leads to improved learning in a test of retention when assessing the ability to generate and interpret a parasternal long axis (PLAX) and apical four-chamber view (A4CH) of the heart? Methods: Ultrasound-novices will be recruited from rotators in the ED. Participants will be allocated to one of three groups based on a 2x2 orthogonal design: Group A (variable difficulty × self-determined practice); Group B (variable difficulty × fixed practice); Group C (static difficulty × fixed practice). A standardized didactic lecture will be presented to each participant. Practice conditions with respect to difficulty level (easy, medium, hard) and structure of practice (learner-determined or fixed practice) will vary according to assigned groups. All groups will receive standardized feedback. The ability to identify anatomy and pathology will be assessed. At the conclusion of practice, a post-practice skills assessment and survey will be administered. Two to three weeks later, participants will be retested using three case scenarios. Screenshots of the participant-determined “best image” and video of the performance will be taken to be evaluated by two blinded (to group allocation) reviewers. Results: We have currently enrolled 14 participants. We aim to complete enrollment by April 2016. Conclusion: We anticipate that our study will provide evidence to inform the best use of ultrasound simulators for teaching TTE in the ED. It will also provide insight into the ability of three educational theories to predict best learning using a novel educational intervention.

Introduction: Adverse drug events (ADEs), harmful and unintended consequences of medications, account for 1.7M emergency department (ED) visits in Canada each year. Up to 30% are due to unintentional re-prescribing of culprit drugs, partly due to lack of accessible, succinct, and comprehensible ADE information at the time of prescribing. Through a systematic review and workshops with physicians and pharmacists, we designed new ADE documentation fields. Our objective was to pilot-test the fields to anticipate and address problems prior to their integration into an electronic medical record (EMR). Methods: We seek to introduce structured ADE documentation into an EMR and PharmaNet, BC’s medication-dispensing database, to generate patient-level alerts when attempts to re-prescribe culprit drugs are made. We conducted this qualitative study in the EDs and on the wards of two BC hospitals. The ADE fields collect information about the culprit drug, its effect on the patient, treatment and outcome. We recruited a convenience sample of pharmacists, and distributed paper forms with the ADE fields to them before data collection shifts. We recorded how pharmacists evaluated patients for ADEs and completed the forms. We collected completed forms, and conducted semi-structured interviews for feedback. We analyzed data for common themes using inductive reasoning and constant comparison methods. Results: We observed 6 pharmacists documenting 24 ADEs. The field design was perceived as simple, clear, with sufficient detail to capture ADE information. Users identified fields to be omitted (e.g., excess details of culprit drug), modified (e.g., reporting options), or needing clarification (e.g., treatment details). Users were uncertain about what to report when the differential diagnosis included an ADE, but diagnostic uncertainty remained. Thus, ADE fields should enable communication about suspected events and potential alternative diagnoses. Pharmacists required follow-up in some cases to complete their determination (e.g., C. difficile toxin assay), emphasizing the need to be able to modify an ADE report. Conclusion: Paper-based pilot testing uncovered barriers to ADE documentation, and allowed us to plan for modifications and required linkages between electronic systems. In order to be functional, electronic ADE documentation must be dynamic, representing a departure from previous reporting platforms.

Introduction: Hypoglycemia is a common treatment consequence in diabetes mellitus (DM) and the second most common cause of Emergency Department (ED) visits for adverse drug events. Prior studies have examined the rates of ED visits and inpatient hospitalizations for hypoglycemia. These represent only a small proportion of severe hypoglycemic events, as many do not present to hospital. To date, there have been no Canadian population-based studies examining the rates of EMS assist-requiring hypoglycemia in DM patients in the pre-hospital setting. The objective of this study was to determine the prevalence and describe the EMS assist-requiring hypoglycemia in DM patients in Southwestern Ontario. Methods: A population-based retrospective cohort study was conducted on all EMS calls for diabetic emergency from 2008-2014 in Southwestern Ontario, Canada. Data was extracted from the electronic ambulance call records for 11 EMS services in the region. Results: There were 9,265 EMS calls for a diabetic emergency (mean age 59 ± 20 years, 57% male, 82% DM). For 223 calls (2.4%) patients were younger than 19 years of age. The mean blood glucose level on presentation was 2.49 ± 1.02 mmol/L and 2,116 (24%) call subjects had initial GCS score less than 9. Treatment (intravenous glucose or IM glucagon) was given in 7,126 (77%) calls. There were 3,884 (51 %) hypoglycemia episodes with documented insulin use and 1,436 (19 %) documented oral hypoglycemia agents use. Between 2008 and 2014, rates of calls increased by 7.4% (p<0.0001). Prevalence of hypoglycemia calls during the study period was estimated at 189 per 10,000 diabetes patients per year. In 2,297 (24.8%) instances, the patient refused transport to the ED. Conclusion: The rates of EMS assist-requiring hypoglycemia are almost double the rates of hospitalization/ED visits for acute DM complications in our region. Many life threatening episodes of hypoglycemia may go unreported and subsequently not followed by the patient's primary health care provider. Further assessment and proper education following those episodes may help decrease the rate of severe hypoglycemia.

Introduction: Deep venous thrombosis (DVT) is a common diagnosis in the Emergency Department (ED). Despite evidence that Rivaroxaban is non-inferior to the low molecular weight heparin (LMWH) bridge to Warfarin approach for anticoagulation, there is still variability in physician practice. A collaborative ED-Hematology quality improvement initiative, that included a treatment guideline and increased access to a thrombosis clinic, was introduced to guide anticoagulation. Methods: A retrospective chart review of ED patients with DVT one-year pre (April 1, 2013-March 31, 2014) and one-year post (April 1, 2014-March 31, 2015) implementation of an outpatient DVT treatment guideline was conducted. Primary outcomes were percentage of patients discharged from the ED on Rivaroxaban or LMWH/Warfarin. Secondary outcomes included mean ED length of stay (ED LOS), mean number of return ED visits per patient and percentage of thrombosis clinic referrals. Balance measures included percentage of return ED visits with pulmonary embolism (PE) within one month and percentage of return ED visits with bleeding (major bleeding or clinically relevant non-major bleeding) due to anticoagulation use. Clinical and administrative data was extracted with 15% independently reviewed for inter-rater reliability. Results: 95 patients met inclusion criteria (52 patients pre and 43 post guideline implementation). The prescribing of Rivaroxaban increased from 9.6% (5/52) to 62.7% (27/43). Mean ED LOS for the Rivaroxaban group was 7.5 hours (95% CI, 5.8-9.2) versus 10.0 hours in the Warfarin group (95% CI, 8.5-11.4) [p=0.04]. The mean return ED visits for the Rivaroxaban group was 0.2 (95% CI, 0-0.3) versus 3.9 in the Warfarin group (95% CI, 3.2-4.6) [p<0.001]. The thrombosis clinic referrals increased from 29.5% (13/44) to 86.0% (37/43). There was one PE diagnosed in the Warfarin group within one month of treatment and zero in the Rivaroxaban group. There were 7.9% (5/63) return visits for bleeding in the warfarin group and 3.1% (1/32) in the Rivaroxaban group. Conclusion: By implementing an outpatient DVT treatment guideline at our academic center, we increased the prescribing of Rivaroxaban. This significantly decreased both the ED LOS and return ED visits in the Rivaroxaban group. There was also a threefold increase in referrals to a thrombosis clinic. This was all achieved without increasing patient harm.

Introduction / Innovation Concept: Orbital compartment syndrome (OCS) is a vision threatening ocular emergency that occurs when there is a sudden rise in orbital pressure resulting in damage to intraocular structures. Lateral canthotomy and cantholysis (LCC) is a simple procedure used to decompress the orbit. Emergency physicians should be comfortable evaluating and diagnosing OCS, and performing a LCC to decrease the risk of vision loss in the event that consultation and intervention by an ophthalmologist is not possible in a timely manner. Developing this skill is challenging as this procedure is seldom performed, therefore resources need to be available. Current training videos are an excellent learning tool but are limited by several factors, such as not capturing from the perspective of the physician performing the procedure. Point-of-view (POV) cameras show the physician’s perspective, which is more conducive to training as it mimics the experience for trainees. We report our novel technique of recording a LCC using a head-mounted POV camera as a resource for emergency physician trainees in learning this procedure. Methods: We used a head mounted POV GoPro Hero 4 Silver camera (GoPro, San Mateo, CA, U.S.A.) with a modified 5.4mm f/2.5 aftermarket lens with a 60° field of view (Peau Productions Inc, San Diego, CA, U.S.A.). This lens was pre-focused to a working distance of 17 inches, set to 1080P on narrow recording at 48 frames per second, and had spot metering and the low light functions turned on. The camera functions were controlled remotely by an assistant with the use of GoPro App on a tablet computer to ensure proper framing of the camera. Curriculum, Tool, or Material: Our novel use of a POV camera for recording LCC is an efficient, cost effective tool useful for medical education at an academic institution as well as a valuable resource for emergency room clinicians. The POV recording system can be a training device in an emergency setting for performing a LCC or other procedures that emergency physicians may seldom encounter. Conclusion: Point-of-view cameras have great potential in assisting the education at the post-graduate level within residency training programs. Video recording from the physician’s perspective simulates the experience for trainees and could leave them feeling more confident in their ability to perform the procedure.

Introduction: Recently, intranasal (IN) fentanyl and inhaled nitrous oxide/oxygen (N2O) mixture have been increasingly used for procedural sedation and analgesia (PSA) alone or in combination. There is a lack of data on the efficacy of these combined agents. Methods: The objective was to evaluate the efficacy of IN fentanyl and N2O as PSA for the reduction of mildly-to-moderately displaced fractures and dislocations. We performed a prospective, observational cohort study between September 2014 and October 2015. Patients were recruited at CHU Sainte Justine (Montréal) and Royal Children Hospital (Melbourne, Australia). Patients aged 4 to 18 years were eligible if PSA consisted of IN fentanyl and N2O for the reduction of mildly-to-moderately displaced fractures or dislocations. Patients received at least IN fentanyl 1.5 mcg/kg (100 mcg max) and at least a 50/50% mixture of N2O with oxygen. Primary outcome was the efficacy of PSA measured by the patient assigned Facial Pain Scale-Revised (FPS-R). The Face, Legs, Activity, Cry, Consolability (FLACC) scale was also recorded. Depth of sedation was evaluated using University of Michigan Sedation Scale (UMSS). Adverse events were recorded following criteria of the Consensus Panel on Sedation Research of PERC/PECARN. Additional data concerning satisfaction or discomfort were evaluated via questionnaires, and follow-up telephone calls were made to elicit information on adverse events after discharge. Results: A total of 91 patients aged 9.7 ± 3.0 years were enrolled. There was no difference between the median FPS-R score during the procedure compared to before: Median 2 and 2 (median difference 0 [95% CI 0, 0]), respectively. The FLACC score was higher during the procedure than before: Median 4 and 0 (median difference 2 [95% CI 1, 3]). UMSS was 1 (95% CI 1, 2) during the procedure. 42 (46%) patients had adverse events, all mild: vertigo (20%), nausea (16%)]or vomiting (12%). A total of 85/88 (97%) parents and 82/85 (96%) ED physicians would want the same sedation in another procedure. Conclusion: PSA with IN fentanyl and N2O seems effective in our study, as evaluated by patient assigned FPS-R. Patients were minimally sedated. Adverse events were frequent but mild. Overall, parents and medical staff would want the same agents used in another procedure. Thus, PSA with IN fentanyl and N2O appears to be an attractive option for reduction of mildly displaced fractures or dislocations.

Introduction: Crisis decision-making is an important responsibility of the resuscitation team leader but a difficult process to study. The purpose of this study was to evaluate visual and behavioural differences between team leaders with different objective performance scores using gaze-tracking technology. Methods: Twenty-eight emergency medicine residents in different stages of training completed four simulated resuscitation scenarios. Participants wore gaze-tracking glasses during each station. An outside expert blinded to participant training level assessed performances using a validated assessment tool for simulation scenarios. Several visual endpoints were measured, including time, frequency, order, and latency to observation of task-relevant and task-redundant items. Non-visual endpoints included behaviours such as summarizing, verbalizing concerns, and calling for definitive treatments, among others. Results: Preliminary findings suggest significant differences between high and low performers. High performers check vitals signs faster, and look at patients and vital signs more often than low performers. Low-performing leaders display a more fixed gaze when starting a scenario. Lastly, high performers summarize, verbalize concerns, predict and prepare for future steps, and call for definitive treatment more often than low performers. Conclusion: There are significant differences between high and low-performing resuscitation team leaders in terms of their visual and behavioural patterns. These differences identify potential focus points for competency evaluations, and may direct educational interventions that could facilitate more efficient development of expertise. The potential to study crisis decision-making behaviours and performances using the methods and metrics identified, both in simulated and real-world settings, is substantial.

Introduction: The International Student Association of Emergency Medicine (ISAEM) is a non-profit organization composed of medical students and student groups who believe that everyone deserves high-quality emergency care. Our aim is to promote and foster the concept, philosophy, and art of Emergency Medicine (EM). More specifically, we seek to 1) create an international network of medical students interested in EM, 2) support EM Interest Groups (EMIGs) and medical students in accomplishing their goals, 3) call for the recognition of EM as an independent specialty in countries where it does not exist, 4) help medical students learn, practice, and advance EM in countries where it is already established, and 5) carry out international projects for the benefit of medical students interested in EM. Methods: ISAEM tries to accomplish its goals primarily by connecting interested medical students and EMIGs with each other, as well as with EM professionals and organizations around the world. Additionally, we support medical students and EMIGs financially, offer them extensive benefits through a free membership, represent their local interests through our National Ambassadors, and advocate on their behalf at the local, national, and international level. Results: ISAEM’s membership base is rapidly growing and our organization is currently represented by students in over 20 countries. In areas where the specialty of EM is not yet recognized, such as in Cameroon, ISAEM helped create the first EMIG and assists students with local projects. In countries where EM is new, such as Brazil, ISAEM helps students discover, explore, and advance this specialty. In countries where EM is thriving, like Canada, ISAEM offers students academic and personal opportunities to advance their careers and the specialty of EM internationally. Additionally, with the help of EM leaders worldwide, ISAEM has recently launched the FOAMed (Free Open Access Medical education) Translation Project and the International Observership Program. In the future, we aim to offer students international research, clinical, and mentorship programs, as well as more financial support. Conclusion: ISAEM is the international voice of medical student interested in promoting access to and expertise in emergency medical services worldwide. Through international collaboration, we hope to create an extensive network that will benefit medical students and the specialty of Emergency Medicine for many years to come.

Introduction: Cultural safety is integral to good clinical care, particularly for Indigenous patients. However, it remains poorly defined in emergency department care (ED). Practitioners at an urban Canadian ED serving a significant Indigenous population sought to engage with the community to define areas for improvement in culturally safe emergency department care. Methods: A participatory action approach was used. A Steering Committee was created, including emergency clinicians and Indigenous health researchers. The Committee collaborated with a local Indigenous health study (Our Health Counts) to aid recruitment. Relevant Indigenous community organizations were identified and engaged via email and personal visits. Recruitment posters were placed in common areas at community sites and the ED. Convenience and snowball sampling was used - potential participants called an ED research coordinator and inclusion criteria were confirmed (self identify as Indigenous, >18 years old, ED visit within the past year). Eligible participants were invited to attend a focus group facilitated by an Aboriginal Elder. Results: 31 individuals called to enroll for a total of 4 potential focus groups. 1 was successfully held: 5 participants were confirmed, 2 attended. Many recruitment challenges were identified, including difficulty maintaining contact/follow-up with a transient population, poster dissemination before recruitment start date, non-Indigenous patients attracted by compensation, and potential participant safety concerns regarding non-Indigenous contact point. Conclusion: Our initiative highlights challenges in engaging vulnerable populations in a large city. Focus groups may be logistically too challenging for this transient population. Other real-time data collection methods, such as phone interviews or surveys may be promising. An Indigenous contact point would likely improve perceived safety. The lack of socio-demographic data collection makes identifying potential participants challenging.

Introduction: La b-card (Boussignac Cardiac Arrest Resuscitation Device) est un dispositif permettant d'assurer une oxygénation passive continue lors des manoeuvres de compressions/décompressions réalisées dans le cadre d'un arrêt cardiaque. Ce dispositif fonctionne par création d'une valve virtuelle induite par l'accélération d'un débit d'oxygéne via des micro-canalicules. Cette valve est censée s'opposer aux flux de gaz entrant et sortant de la cage thoracique lors des compressions/décompressions. Elle permettrait d'obtenir une pression positive intra thoracique lors des compressions, et une pression intra thoracique négative lors des décompressions. L'expérimentation conduite a pour but de mesurer la pression statique créée dans le dispositif par le débit d'oxygène, ainsi que les valeurs de pressions et de flux générés en intra thoracique. Methods: La b-card est almimentée par un débit d'oxygène de 15 L/min, et connectée via différentes interfaces à un poumon test pourvu de capteurs:

-Capteur de pression des voies aériennes (PAW en cm H2O).

-Capteur de débit au niveau des voies aériennes.

-Capteur de pression “intra thoracique” (PIT max et min; et Pression Expiratoire Intra Thoracique).

Les mesures sonrt effectuées sans b-card, puis avec b-card connectée à un masque facial, un masque laryngé, une sonde trachéale.

Results: La pression “statique”, celle de la valve virtuelle, mesurée au niveau de la b-card reste stable à 6 cm d'H2O, sous un débit de 15 L/min. Elle permet une résistance à hauteur de cette valeur aux flux de gaz entrant ou sortant du thorax expérimental en fonction des compressions/décompressions. Les pressions intra thoraciques positives mesurées lors des compressions restent équivalentes autour de 25 à 30 cm H2O, et ce quelle que soit l'interface utilisée. Les pressions intra thoraciques négatives mesurées lors des décompressions restent équivalentes autour de 10 à 15 cm d'H2O, et ce quelle que soit l'interface utilisée. Conclusion: Les pressions intra thoraciques obtenues en associant une oxygénation passive par la b-card à des compressions/décompressions continues permettent d'assurer une ventilation efficace et synchrone. Les pressions mesurées au niveau du dispositif sont constamment inférieures à la pression d'ouverture moyenne oesophagienne, ce qui éviterait toute insufflation gastrique.

Introduction: Minimizing haemorrhage using direct pressure is intuitive and widely taught. In contrast, this study examines the use of indirect-pressure, such as external aortic compression which has been identified as an immediately applicable maneuver to address the leading cause of battlefield mortality: junctional hemorrhage. However, it is currently unclear how to optimize this technique. Methods: This prospective, block-randomized, cross-over simulation study of compression optimization was performed on a model of central vessel compression that recorded weight (lbs) and pressure (mmHg). Forty participants simulated external aortic compression on the ground as well as a stretcher with and without a backboard. Participants were blinded to compression weight and pressure, as well as the purpose of the study, to minimize preparation bias. Manoeuvres were performed in alternating order to control for skill acquisition and fatigue. Scripted instructions were followed to compress with 1 then 2 hands, and to apply “sustainable effort” and then “maximal effort”. Results: The greater the compressor's bodyweight the greater their mean compression (Pearson’s correlation 0.9342). Using one-hand, a mean of 28% participant bodyweight (95%CI, 26% - 30%) could be transmitted at sustainable effort, waist-height, and on a stretcher. A second compressing hand increased rescuer bodyweight transmission by 10-22% regardless of other factors (i.e. presence/absence or a backboard; rescuer position) (p<0.001). Adding a backboard increased transmission of rescuer bodyweight 7%-15% (p<0.001). Lowering the patient from waist-height backboard to the floor increased transmission of rescuer bodyweight 4%-9% (p < 0.001). Kneeling on the model was the most efficient method and transmitted 11% more weight compared to two-handed maximal compression (p <0.001). Conclusion: Efficacy is maximized with larger-mass, two hands, and compression on hard surfaces/backboards. Knee compression is most effective and least fatiguing, thus assisting rescuers of lower weight and lesser strength, where no hard surfaces exist (i.e. no available backboard or trauma on soft ground), or when lengthy compression is required (i.e. remote locations). This study demonstrates the feasibility of indirect pressure as a potential temporizing measure for life-threatening haemorrhage not amenable to direct compression.

Introduction: In certain circumstances, skin and soft tissue infections are managed with intravenous (IV) antibiotics. In our center, patients initiated on outpatient IV antibiotics are followed up by a home parental therapy program the following day. A significant number of these patients require a repeat visit to the ED because of clinic hours. Probenecid is a drug that can prolong the half-life of certain antibiotics (such as cefazolin) and can therefore avoid a repeat ED visit, reducing health care costs and improve ED capacity. Our goal was to increase probenecid usage in the ED in order to optimize management of skin and soft tissue infections (SSTI) in the ED. The primary outcome was to compare the usage of probenecid in the pre and post-intervention phase. Secondary outcomes were to compare revisit rates between patients receiving cefazolin alone vs cefazolin + probenecid. Methods: Using administrative data merged with Computerized Physician Order Entry (CPOE), we extracted data 90 days pre- and 90 post-intervention (February 11, 2015 to August 11, 2015). The setting for the study is an urban center (4 adult ED’s with an annual census of over 320,000 visits per year). Our CPOE system is fully integrated into the ED patient care. The multi-faceted intervention involved modifying all relevant SSTI order sets in the CPOE system to link any cefazolin order with an order for probenecid. Physicians and nurses were provided with a 1 page summary of probenecid (indications, contra-indications, pharmacology), as well as decision support with the CPOE. Any patients who were receiving outpatient cefazolin therapy were included in the study. Results: Our analysis included 2512 patients (1148 and 1364 patients in the pre/post phases) who received cefazolin in the ED and were discharged during the 180 day period. Baseline variables (gender, age, % admitted) and ED visits were similar in both phases. In the pre-intervention phase 30.2% of patients received probenecid and in the post-intervention phase 43.0%, for a net increase of 12.8% (p=<0.0001). Patients who received probenecid had a 2.2% (11.4% vs 13.6%, p=0.014) lower re-visit rate in the following 72H. Conclusion: We have implemented a CPOE based clinical decision support intervention that demonstrated significant increase in probenecid usage by emergency physician and resulted in a decrease in ED revisits. This intervention would result in health care cost-savings.

Introduction: Fever is one of the most common reasons for presentation to the emergency department (ED). Interestingly, a number of small studies suggest that fever may function as a modulator of the QT interval in healthy individuals and an arrhythmogenic trigger in patients with occult congenital QT abnormalities. The objective of this study was to explore whether presence of fever adversely affects the QT interval, and whether medications known to prolong this interval affect any association found. Methods: We performed a retrospective, single center study identifying patients (age > 18 years) presenting to the ED with fever (temperature > 38.0 °C) between January 1st, 2012 and December 31st, 2013 via electronic chart review. The subset for analysis were those who had an ECG both at time of fever and while afebrile (within 30 days of initial ECG). Temperature measurement was within 30 minutes of ECG. Actively paced patients were excluded. Univariate and multiple regression analysis were used to determine risk factors for QT derangement in patients with fever. Results: 2018 febrile visits occurred during the reviewed period, 181 of these patients went on to be included in the study. 54.1% of study subjects were female, and the average age was 68.9 years old. The etiology of fever was predominately infectious (69.6%), with community acquired pneumonia being the most frequent cause (24.3%). We found the median corrected QT interval to be significantly shorter in febrile as compared to afebrile patients [QTc = 388.7ms, (371.5-407.5) vs 406.7, (386.7-434.4); p < 0.001]. This difference was observed in both sexes. Males were found to be more likely to experience medication induced QTc prolongation [OR 5.35, 95% CI = 1.46 - 19.68; P < 0.05]. Two instances of Torsades de pointes were identified in our study, both occurring in males on QT prolonging medications. Conclusion: In an ED patient population, fever generally shortens the QT interval independent of sex. Prolongation of the QT interval during fever should thus increase clinical suspicion of congenital or acquired QT disorders. Additionally, males appear to be more susceptible to medication-induced derangements in the QT interval and may require more vigilant monitoring when treated with multiple QT prolonging medications.