Introduction: Dizziness is among the most common presenting complaints in the emergency department (ED). Although the vast majority of these cases are the result of a benign, self-limiting process, many patients undergo computed tomography (CT) of the head. The objective of this study was to define the yield of and diagnostic accuracy of CT in dizziness in addition to defining high-risk clinical features predictive of an abnormal CT. Methods: At a tertiary care ED we performed a medical records review from Jan 2015-2018 including adult patients with a triage complaint of dizziness (vertigo, unsteady, lightheaded), excluding those with symptoms >14days, recent trauma, GCS < 15, hypotensive, or syncope/loss of consciousness. Five trained reviewers used a standardized data collection sheet to extract data. Our outcome was a central cause defined as: cerebrovascular accident (CVA), brain tumor (BT) or intracranial haemorrhage (ICH) diagnosed on CT or magnetic resonance imaging. Univariate analysis/logistic regression were performed and odds ratios reported. A sample size of 796 was calculated based on an expected prevalence of 5% with an 80% power and 95% confidence interval to detect an odds ratio greater than 2. Results: 2310 patients were recruited, 800 (35%) underwent CT head, 471(59%) female and a mean age of 62.8 years (+/−17.5 years). The top three diagnoses for patients undergoing CT were peripheral vertigo/benign positional vertigo (153 – 19%), vertigo not-otherwise-specified (137 – 17%) and dizziness not-otherwise-specified (137 – 17%). The number of CT scans considered abnormal was 30 (3.7%). The top three diagnoses for patients with an abnormal CT were CVA (22 – 75%), BT (9 – 26%) and ICH (6-17%). High risk clinical findings associated (p < 0.001) with an abnormal head CT were dysmetria, objective motor neurological signs, positive Rhomberg, ataxia and inability to walk 3 steps. Objective motor neurological signs (OR 8.4 [95% CI 3.27-21.72]) and ataxia (OR 3.4 [95% CI 1.62-7.41]) were both independently associated with an abnormal CT. Patients without any high risk findings on exam had a 0.7%(3/381 – 2 CVA,1 Tumour) probability of an abnormal CT. Sensitivity of CT for a central cause of dizziness was 71.43%(95%CI 55.4-84.3%), specificity 100%(95%CI 99.5-100%). Conclusion: Current rate of imaging in dizziness is high and inefficient. CT should be the first imaging test in those with high-risk clinical features, but a normal result does not rule out a central cause.

Introduction: In order to better characterize procedural skills curricula in Canada, a national survey was conducted. The objectives of the survey were: (i) to characterize procedural skills education currently employed in pre-clerkship and clerkship curricula; (ii) to determine what skills physician-educators think medical students should know upon graduation; and (iii) to identify physician-educator perceptions regarding the development of pre-clerkship procedural curriculum. Methods: A web-based survey was distributed to 201 clinician-educators across Canada's 17 medical schools. Respondents were directed to an individualized survey based on their self-identified roles at their institution. Respondents were asked demographic questions, what procedural skills are being taught and in what setting at their institution, and their opinions on the value of a pre-clerkship procedural curriculum. Results: From the 17 school's surveyed, 12 schools responded, with 8 schools responding “yes” that they had a clerkship procedural curriculum. For a pre-clerkship procedural curriculum, only 4 schools responded “yes”. The 5 of the top 10 procedurals skills identified that medical students should know upon graduation, in order, are: IV Access, Airway Management/Ventilator Management, Local anesthesia/field block, Casting, Spontaneous Vaginal Delivery. On a Likert scale, clinician-educators strongly supported a pre-clerkship procedural curriculum (median = 4.00/5.00, mode = 5.00/5.00), and they believed it would decrease anxiety (median = 4.00/5.00), increase confidence (median = 4.00/5.00), and increase technical ability (median = 3.00/5.00) in incoming clerks. Conclusion: Across Canada, the state of undergraduate medical education procedural skills education is inconsistent. With the identification of the Top 10 procedural skills medical students should know upon graduation, the learning objectives of a formal curriculum can be developed. With overwhelming support from physician-educators, a formal pre-clerkship procedural curriculum is poised to redefine the landscape of procedural care for a whole new generation of physicians.

Introduction: 9-1-1 telecommunicators receive minimal education on agonal breathing, often resulting in unrecognized out-of-hospital cardiac arrest (OHCA). We successfully piloted an educational intervention that significantly improved telecommunicators’ OHCA recognition and bystander CPR rates in Ottawa. We sought to better understand the operations of Canadian 9-1-1 communications centers (CC) in preparation for a multi-centre study of this intervention. Methods: We conducted a National survey of all Canadian CCs. Survey domains included information on organizational structure, dispatch system used, education curriculum, and performance monitoring. It was peer-reviewed, translated in French, pilot-tested, and distributed electronically using a modified Dillman method. We designated respondents in each CC before distribution and used targeted follow-up and small incentives to increase response rate. Respondents also described functioning of neighboring CCs if known. Results: We received information from 51/51 provincial and 1/25 territorial CCs, representing 99.7% of the Canadian population. CCs largely utilize the Medical Dispatch Priority System (MPDS) platform (93%), many are Province/Ministry regulated (50%) and most require a High School diploma as minimum entry level education (78%). Telecommunicators receive initial in-class training (median 1.3 months, IQR 0.3-1.9; range 0.1-2.2), often followed by a preceptorship (84.4%) (median 1.0 months, IQR 0.7-1.7; range 0.4-6.0). Educational curriculum includes information on agonal breathing in 41% of CC, without audio examples in 34%. Among responding CCs, over 39,000 suspected OHCA 9-1-1 calls are received annually. Few CCs maintain local performance statistics on OHCA recognition (25%), bystander CPR rates (25%) or survival rates (50%). Most (97%) expressed interest in future research collaborations. Conclusion: Most Canadian telecommunicators receive no or minimal education in recognizing agonal breathing. Further training and improved OHCA monitoring may assist recognition and enhance outcomes.

Innovation Concept: Competence by Design (CBD) was implemented nationally for Emergency Medicine (EM) residents beginning training in 2018. One challenge is the need to introduce residents to Entrustable Professional Activities (EPAs) that are assessed across numerous clinical rotations. The Royal College's resources detail these requirements, but do not map them to specific rotations or present them in a succinct format. This is problematic as trainees are less likely to succeed when expectations are unclear. We identified a need to create practical resources that residents can use at the bedside. Methods: We followed an intervention mapping framework to design two practical, user-friendly, low-cost, aesthetically pleasing resources that could be used by residents and observers at the bedside to facilitate competency-based assessment. Curriculum, Tool or Material: First, we designed a set of rotation- and stage-specific EPA reference cards for the use of residents and observers at the bedside. These cards list EPAs and clinical presentations likely to be encountered during various stages of training and on certain rotations. Second, we developed a curriculum board to organize the EPA reference cards by stage based upon our program's curriculum map. The curriculum board allows residents to view the program's curriculum map and the EPAs associated with each clinical rotation at a glance. It also contains hooks to hang and store extra cards in an organized manner. Conclusion: We believe that these practical and inexpensive tools facilitated our residency program's transition to competency-based EPA assessments. Anecdotally, the residents are using the cards and completing the suggested rotation-specific EPAs. We hope that the reference cards and curriculum board will be successfully incorporated into other residency programs to facilitate the introduction of their EPA-based CBD assessment system.

Introduction: In 2016, a team at McMaster began developing GridlockED, an educational (or “serious”) board game designed to teach medical learners about patient flow in the emergency department. As serious board games are a relatively new phenomenon in medical education, there is little data on how marketed games are actually used once received by end-users. In this study our goal was to better understand the demographics and game usage for purchasers of the GridlockED board game, which will inform the further improvement or expansion of the game. Methods: Individuals who expressed interest in purchasing gridlockED via our online storefront were sent an anonymous online survey via Google Form. The survey collected demographic and qualitative data with a focus on the respondent's role in medicine, how they have used GridlockED, who they have played GridlockED with, and what changes or additions to GridlockED they would like to see. We also asked about changes for a potential mass-market version of the game targeted towards non-medical individuals. Individuals who did not purchase the game were asked about their barriers to purchase. We received an exemption for this study from our institutional review board. Results: 42 responses (out of 300 individuals on our mailing list, 14% response rate) were collected. Responding purchasers were from 16 different roles in healthcare and 11 different countries. The top 5 roles were: EM trainee, Community EM MD, Academic EM MD, Physicians from other specialties, and EM program director. The majority of respondents were Canadian (38%), with America (21%), New Zealand (10%), and Turkey (7%) the only other countries to have more than 2 respondents. 50% reported having played the game, with the most common use cases being for fun (76%), for teaching trainees (33%) or training with colleagues (19%). For those who did not purchase, price was the largest barrier (81%). 50% of respondents expressed interest in a disaster scenario expansion pack, with 33% interested in set lesson plans. Conclusion: GridlockED attracted interest from a wide range of medical professionals, both in terms or role and location. Users mainly reported using the game for fun, with fewer users using the game for teaching/training purposes. The main barrier to purchase was the game's price.

Introduction: Acute upper gastrointestinal bleeding (UGIB) is a common presentation to emergency departments (ED). Of these patients, 35-45% receive a blood transfusion. Guidelines for blood transfusion in UGIB have been well established, and recommend a hemoglobin (Hb) level below 70 g/L as the transfusion target in a stable patient. There is no consensus on a transfusion threshold for unstable UGIB. There is limited data regarding physician practices in the ED. The aim of our study is to determine the appropriateness, by expert consensus, of blood transfusions in UGIB in a tertiary care hospital ED. Methods: We retrospectively reviewed patients presenting with UGIB to the University of Alberta Hospital ED in 2016. These patients were then screened for blood transfusions. Data were obtained from the patient records. Chart derived data were verified with records obtained from the blood bank. For each patient, the history, vitals, Glasgow Blatchford Score (GBS), relevant labs, and record of blood transfusions were collected and organized into a case summary. Each patient summary was presented individually to a panel of three expert clinicians (2 Gastroenterology, 1 Emergency Medicine), who then decided on the appropriateness of each blood transfusion by consensus. Results: Blood transfusions (data available 395/400) were given to 51% (202/395) of patients presenting with UGIB. Of these, 86% (174/202) were judged to be appropriate. Of the 395 patients, 34% (135/395) had a Hb of <70 g/L. Of these, 93% (126/135) were transfused, and all of these were considered appropriate. 18% (70/395) had a Hb between 71-80. 74% (52/70) of these patients were given blood, and 79% (41/52) were considered appropriate. 13% (50/395) of the patients had a Hb between 81-90, with 28% (14/50) receiving a transfusion. Of these, 36% (5/14) were deemed to be appropriate. 35% (140/395) of patients had a Hb of >90. 7% (10/140) of these received blood. 20% (2/10) were considered appropriate. Conclusion: The panel of expert clinicians judged 86% of the blood transfusions to be appropriate. All transfusions under the recommended guideline of 70 g/L were considered appropriate. In addition, the majority of transfusions above a Hb of 70 g/L were considered appropriate, but 37% were not. Further studies evaluating the feasibility of current guideline recommendations in an ED setting are required. Educational interventions should be created to reduce inappropriate blood transfusions above a Hb 70 g/L.

Introduction: Maintaining competence in high-acuity low-occurrence (HALO) procedures is often difficult due to their infrequent occurrence. While simulation is a valuable tool to hone skills, providing effective simulation-based education (SBE) to learners outside academic centers can be challenging. Utilizing a mobile tele-simulation unit (MTU) with expert instruction from a geographically separated mentor could prove a valuable approach to overcoming barriers in this setting. However, to maximize benefit and buy-in, the training modules developed for this unique delivery method must align with the needs of those practicing in rural settings. Objectives: - To evaluate the procedural skills training needs of emergency medicine (EM) physicians in rural Newfoundland and Labrador (NL) - To inform the development of simulation modules designed for use in a MTU Methods: A web-based needs assessment was distributed to physicians registered with the NL Medical Association, working in rural locations, and having EM listed as their primary specialty. Participants evaluated their comfort, performance frequency and desire to have further training for 12 HALO procedures. Two EM physicians selected these from a broader list of core procedural skill competencies for CCFP-EM residents at Memorial University. Participants were also able to suggest other procedures that might benefit from SBE. Results: The data collection occurred for 8 weeks with a 68% response rate (N = 22). No respondents had formal EM training outside of exposure in family medicine residency. 60% had 10+ years practicing EM. Chest tube insertion (100%), difficult intubation (92.3%) and surgical airway (92.3%) were the procedures that most respondents felt required more SBE. In practice, they most often performed bag-valve ventilation, splint application and procedural sedation (>10 per year). Additional procedures felt to require SBE were central venous line placement and trauma assessment. Opportunities to participate in SBE were limited (66.7%-less than annually). Despite this, most participants agreed SBE would be a significant benefit if accessible (93.3%). The greatest barriers to SBE included lack of equipment, rural location, and time necessary for travel to larger centres. Conclusion: The provision of medical care in rural settings can be particularly challenging when HALO procedures must be performed. Unique delivery methods of SBE targeted to the needs of rural practitioners may help bridge gaps in knowledge and technical skills.

Introduction: Previous systematic reviews suggest early mobilization in the intensive care unit (ICU) population is feasible, safe, and may improve outcomes. Only one review investigated mobilization specifically in trauma ICU patients and failed to identify any relevant articles. The objective of the present systematic review was to conduct an up-to-date search of the literature to assess the effect of early mobilization in adult trauma ICU patients on mortality, length of stay (LOS) and duration of mechanical ventilation. Methods: We performed a systematic search of four electronic databases (Ovid MEDLINE, Embase, CINAHL, Cochrane Library) and the grey literature. To be included, studies must have compared early mobilization to delayed or no mobilization among trauma patients admitted to the ICU. Meta-analysis was performed to determine the effect of early mobilization on mortality, hospital LOS, ICU LOS, and duration of mechanical ventilation. Results: The search yielded 2,975 records from the 4 databases and 7 records from grey literature and bibliographic searches; of these, 9 articles met all eligibility criteria and were included in the analysis. There were 7 studies performed in the United States, 1 study from China and 1 study from Norway. Study populations included neurotrauma (3 studies), blunt abdominal trauma (2 studies), mixed injury types (2 studies) and burns (1 study). Cohorts ranged in size from 15 to 1,132 patients (median, 63) and varied in inclusion criteria. Most studies used some form of stepwise progressive mobility protocol. Two studies used simple ambulation as the mobilization measure, and 1 study employed upright sitting as their only intervention. Time to commencement of the intervention was variable across studies, and only 2 studies specified the timing of mobilization initiation. We did not detect a difference in mortality with early mobilization, although the pooled risk ratio (RR) was reduced (RR 0.90, 95% CI 0.74 to 1.09). Hospital LOS and ICU LOS were decreased with early mobilization, though this difference did not reach significance. Duration of mechanical ventilation was significantly shorter in the early mobilization group (mean difference −1.18. 95% CI −2.17 to −0.19). Conclusion: Our review identified few studies that examined mobilization of critically ill trauma patients in the ICU. On meta-analysis, early mobilization was found to reduce duration of mechanical ventilation, but the effects on mortality and LOS were not significant.

Background: There is growing concern about emergency physicians overuse of computed tomography (CT). In an attempt to ensure appropriate ordering many hospitals implement strict protocols for ordering of CT scans in the emergency department (ED) that include approval of all scans by a board-certified radiologist, and a reduced access to CT overnight. Aim Statement: The aim of this study is to review the impact of RAD ED – direct access to CT ordering by ED physicians, 24hr CT technologist and third-party reporting on CT scans overnight. Our objectives were to assess the effect on; 1) ED length of stay, 2) number of CT scans ordered and 3) admission rates. Measures & Design: We conducted a prospective pilot before & after study at a single tertiary-care emergency department between February 1st, 2018 and July 31st, 2018. Inclusion criteria were adult patients presenting to the emergency department and undergoing CT for any of the following: face, neck, spine, upper and lower extremities, chest, abdomen and pelvis. Exclusion criteria were those undergoing CT head for stroke or trauma. Evaluation/Results: A total of 924 patients met our criteria, 352 before and 568 after implementation. Comparison of the patient populations demonstrate very similar characteristics in both groups; (49% male, average age 56 years, CTAS 2(40%) and 3(47%). Results demonstrate that an additional 216 scans were performed in post-implementation group. This equates to an increase of 61%. ED length of stay averaged 5.6 hours pre-implementation and 4.7 hours post-implementation. This corresponds to a significant reduction in length of stay of approximately 0.9 hours (p < 0.01). Collection is currently ongoing for factors that we will adjust for a multivariate analysis, including admission rates. Discussion/Impact: RAD ED led to a significant increase in CT ordering and decrease in ED length of stay. We believe that this project provides important information to clinicians and patients with regards to overall CT utilization, ED wait times, follow up visits for CT scanning and admission rates. It is also important for administrators to help decide if these new rules are leading to improved efficiency, and to help estimate their financial impact.

Introduction: We examined our local sepsis patient population, and specifically our most vulnerable patients - those presenting to the emergency department (ED) in septic shock - for variables predictive of survival to hospital discharge. We applied the familiar ED paradigm of, “Door to,” to calculate the impact of time to antibiotics against patient survival to hospital discharge. Methods: Retrospective chart review of patients aged > = 18 years, presenting to tertiary care ED between 01 Nov 2014 and 31 Oct 2015. Patients determined to have sepsis if A) > = 2 SIRS criteria and ED suspicion of infection (ED acquisition of blood/urine cultures or antibiotic administration) and/or B) received ED or Hospital discharge diagnosis of sepsis (ICD-10 diagnostic codes A4xx and R65). Patients sub-classified with septic shock if A) triage SBP < = 90mmHg, B) triage MAP < = 65mmHg or C) serum lactate > = 4mmol/L. “Door Time” was defined as the earliest time recorded for the patient encounter, either the time the patient registered in the Emergency Department, or the triage time. A generalized linear model was performed with a binomial distribution using survival to discharge as the response variable. Age, sex, ED arrival method, time to antibiotics, ED serum lactate and ED serum glucose level were the predictor variables. Results: 13506 patient encounters met inclusion criteria (10980 unique patients). Linear regression of time to antibiotics against survival to hospital discharge failed to achieve statistical significance. Linear regression of the secondary outcome variables achieved statistical significance for age and serum lactate level. Per the model, as age increased by 1 year, the odds of dying prior to hospital discharge increased by 3.8% and as serum lactate increased by 1 mmol/L, odds of dying prior to hospital discharge increased by 11.1%. Conclusion: We found no association between time to antibiotic treatment and mortality. Causal relationships require randomized controlled trials, and this analysis contributes to clinical equipoise.

Introduction: Falls are a common presentation to the emergency department among geriatric patients. The incidence of intracranial bleeding following a fall is unclear and approach to ordering a CT head scan is not standardized. The aim of this systematic review and meta-analysis was to establish the incidence of intracranial bleeding after a fall in geriatric patients. Methods: The systematic review was registered in PROSPERO. Two authors independently searched Medline and EMBASE (OVID interface) from conception till 20th June 2018. The search combined multiple MESH terms and text words for [falls], [elderly] and [brain injury]. The search was repeated in Google Scholar and recent conference abstracts were reviewed. Studies were included if > 80% of the included patients were > 65 years who presented to the emergency department after a fall on level ground. We excluded studies enrolling select populations (for example trauma team activation, neurosurgical patients or only anticoagulated patients). There were no language restrictions. The random effects model was used to perform a meta-analysis on the incidence of intracranial bleeding in geriatric patients after a fall on level ground. Results: From the 7,043 titles and abstracts, 175 full articles were reviewed and 7 studies, including 6758 patients, were included in the analysis. 2/7 studies were prospective. The studies varied in their inclusion criteria with 3/7 studies only including patients with normal neurological testing. Most retrospective studies included patients if they had a CT head scan. Neither prospective study imaged all patients but both followed the patients for a delayed diagnosis of intracranial bleeding. Risk of bias was moderate or high for the majority of studies. The random effects pooled incidence of intracranial bleeding was 5.2% (95% CI 2.8 – 8.2%), I2 96%. Conclusion: Around 1 in 20 geriatric patients who present to the emergency department after a fall have intracranial bleeding. This point estimate can be used to calculate sample size requirements for future studies on intracranial bleeding in this population.

Introduction: Quality improvement and patient safety (QIPS) are increasingly recognized as integral to the provision and advancement of emergency medicine (EM) care. In 2015, QIPS were added to the Canadian Medical Education Directives for Specialists (CanMEDS) framework. However, the level of QIPS education and support that Canadian EM residents receive is unknown. In order to better plan national QIPS efforts aimed at enabling EM residents to improve their local care settings, we sought to assess the current state of QIPS education and support in Canadian EM residency programs. Methods: This was a descriptive, cross-sectional electronic survey that was disseminated to all current Canadian EM residents from both Royal College (RC) and Family Medicine - EM training streams. Residents were recruited either directly or through their program's administrative assistant. The survey consisted of multiple-choice, Likert and free-text entry questions. Themes included a) familiarity with QIPS; b) local opportunities for QIPS projects and mentorship; and c) desire for further QIPS education and involvement. The survey was open for a five-week period, with formal reminders after the first and third weeks. Descriptive statistics are reported. Results: 189 (35%) of 535 current EM residents completed the survey, representing all 17 medical schools. 77% of respondents were from the RC stream. 54.7% of respondents reported being “somewhat” or “very” familiar with QIPS. 47.2% of respondents reported “not knowing” or “not having readily available” QIPS projects to participate in their local environment, and 51.5% had equivalent responses with respect to QIPS mentorship opportunities. Only 17.5% of respondents reported that QIPS methodologies were already formally taught in their residency program, and 66.9% indicated a desire for increased QIPS teaching. The majority of respondents were “slightly” (35.9%), “moderately” (23.2%) or “very” (11.3%) interested in becoming involved with QIPS training and initiatives. Conclusion: Responding Canadian EM residents are interested in obtaining greater QIPS education as well as project and mentorship opportunities, but many perceive that they do not have adequate access to these at the current time. As the importance of QIPS increases in the EM community, supporting residents with more robust educational infrastructures may be necessary. Future efforts may include the standardizing of QIPS postgraduate curricula and improving access to QIPS opportunities across the country.

Introduction: Intranasal dexmedetomidine (IND) is an emerging agent for procedural distress in children. However, studies to date have been limited by small samples and imprecise estimates of effect size. We sought to summarize the evidence on the effectiveness of IND for procedures associated with distress in children. Methods: We performed electronic searches of MEDLINE (1946-2018), EMBASE (1980-2018), Google Scholar (2018), CINAHL (1981-2018), Cochrane Central Register of Controlled Trials (2018), 6 clinical trials registries and conference proceedings (2010-2018). Title searches, data abstraction, and risk of bias assessments were performed in duplicate. We included all published and unpublished, randomized and quasi-randomized trials of IND for procedures in children younger than 19 years of age without language restriction. The methodological quality of studies was evaluated using the Cochrane Collaboration's Risk of Bias tool. The primary outcome was the proportion of participants that were deemed to be adequately sedated for the procedure. Results: Of 661 studies, 18 met inclusion criteria. Trials involved 2128 participants, age 1 month - 14 years (836, 39.3% females), who received IND 1 - 4 mcg/kg either by drops (n = 12), atomizer (n = 4), or both (n = 2). 12 trials were eligible for meta-analysis. 13 trials used validated instruments to assess sedation. All studies except one were associated with low or moderate risk of bias. For painful procedures (IV insertion; laceration repair; dental extraction), the pooled OR (95% CI) for adequate sedation and need for additional analgesia was non-significant [1.19 (0.53, 2.65)] and [2.16 (0.62, 7.49)], respectively (n = 5). For non-painful procedures (diagnostic imaging), the corresponding pooled OR (95% CI) favored IND [3.04 (1.58, 5.82)] and [4.44 (2.11, 9.35)], respectively (n = 7). Time to onset and duration of sedation ranged from 13-31 minutes and 41-91.5 minutes, respectively. For adverse effects, the pooled OR (95% CI) was not significantly different between IND and comparators [0.58 (0.22, 1.55] and there were no serious adverse events. Conclusion: IND at doses 1 to 4 mcg/kg are safe and adequately sedate children undergoing non-painful procedures, although the ease of administration must be weighed against the risk of prolonged sedation. Additional trials with larger sample sizes and greater methodologic rigor are needed for painful emergency department procedures such as laceration repair and IV insertion.

Introduction: Patients with chronic non-cancer pain (CNCP) and opioid-use disorders make up a category of patients who present a challenge to emergency department (ED) providers and healthcare administrators. Their conditions predispose them to frequent ED utilization. This problem has been compounded by a worsening opioid epidemic that has rendered clinicians apprehensive about how they approach pain care. A systematic review has not yet been performed to inform the management of CNCP patients in the ED. As such, the purpose of this project was to identify and describe the effectiveness of interventions to reduce ED visits for high-utilizers with CNCP. Methods: Included participants were high-utilizers presenting with CNCP. All study designs were eligible for inclusion if they examined an intervention aimed at reducing ED utilization. The outcomes of interest were the number of ED visits as well as the amount and type of opioids prescribed in the ED and after discharge. We searched Medline, EMBASE, CINAHL, CENTRAL, SCOPUS, Web of Science, and the grey literature from inception to June 16, 2018. Two independent investigators assessed articles for inclusion following PRISMA guidelines. Risk of bias will be assessed using the Cochrane ROBINS-I and RoB 2 tools for non-randomized and randomized trials, respectively. Results: Following review, 14 of the 5,018 identified articles were included for analysis. These articles assessed a total of 1,670 patients from both urban and rural settings. Interventions included pain protocols or policies (n = 5), individualized care plans (n = 5), ED care coordination (n = 2), a chronic pain management pathway (n = 1), and a behavioural health intervention (n = 1). Intervention effects trended towards the reduction of both ED visits and opioid prescriptions. The meta-analysis is in progress. Conclusion: Preliminary results suggest that interventions aimed at high-utilizers with CNCP can reduce ED visits and ED opioid prescription. ED opioid-restriction policies that sought to disincentivize drug-related ED visits were most successful, especially when accompanied by an electronic medical record (EMR) alert to ensure consistent application of the policy by all clinicians and administrators involved in the care of these patients. This review was limited by inconsistencies in the definition of ‘high-utilizer’ and by the lack of high-powered randomized studies.

Introduction: Endotracheal intubation (ETI) is a lifesaving procedure commonly performed by emergency department (ED) physicians that may lead to patient discomfort or adverse events (e.g., unintended extubation) if sedation is inadequate. No ED-based sedation guidelines currently exist, so individual practice varies widely. This study's objective was to describe the self-reported post-ETI sedation practice of Canadian adult ED physicians. Methods: An anonymous, cross-sectional, web-based survey featuring 7 common ED scenarios requiring ETI was distributed to adult ED physician members of the Canadian Association of Emergency Physicians (CAEP). Scenarios included post-cardiac arrest, hypercapnic and hypoxic respiratory failure, status epilepticus, polytrauma, traumatic brain injury, and toxicology. Participants indicated first and second choice of sedative medication following ETI, as well as bolus vs. infusion administration in each scenario. Data was presented by descriptive statistics. Results: 207 (response rate 16.8%) ED physicians responded to the survey. Emergency medicine training of respondents included CCFP-EM (47.0%), FRCPC (35.8%), and CCFP (13.9%). 51.0% of respondents work primarily in academic/teaching hospitals and 40.4% work in community teaching hospitals. On average, responding physicians report providing care for 4.9 ± 6.8 (mean ± SD) intubated adult patients per month for varying durations (39.2% for 1–2 hours, 27.8% for 2–4 hours, and 22.7% for ≤1 hour). Combining all clinical scenarios, propofol was the most frequently used medication for post-ETI sedation (38.0% of all responses) and was the most frequently used agent except for the post-cardiac arrest, polytrauma, and hypercapnic respiratory failure scenarios. Ketamine was used second most frequently (28.2%), with midazolam being third most common (14.5%). Post-ETI sedation was provided by > 98% of physicians in all situations except the post-cardiac arrest (26.1% indicating no sedation) and toxicology (15.5% indicating no sedation) scenarios. Sedation was provided by infusion in 74.6% of cases and bolus in 25.4%. Conclusion: Significant practice variability with respect to post-ETI sedation exists amongst Canadian emergency physicians. Future quality improvement studies should examine sedation provided in real clinical scenarios with a goal of establishing best sedation practices to improve patient safety and quality of care.

Introduction: Acute vestibular syndrome (AVS - vertigo, nystagmus, head motion intolerance, ataxia, and nausea/vomiting) is a subset of patients presenting with vertigo. They are most often due to benign vestibular neuritis but can be a sign of a vertebrobasilar stroke. The HINTS (head impulse test, nystagmus, positive test of skew) exam has been proposed as an extremely accurate bedside test to rule out stroke in those presenting with AVS. Is the HINTS exam compared to MRI sufficiently sensitive to rule out vertebrobasliar stroke in an adult population presenting to the emergency department with AVS. Methods: We searched in Pubmed, Medline, Embase, the Cochrane database, and relevant conference abstracts from 1968 to December 2018 and performed hand searches. No restrictions for language or study type were imposed. Relevant studies were reviewed and data was extracted by two independent reviewers. Gold standard in ruling out stroke was; Negative late acute (72 hrs–10d) cranial MRI with DWI OR Negative early acute (0–72hrs) cranial MRI plus negative follow-up cranial MRI or clinical follow-up for TIA/stroke of ≥3 months. Included studies were prospective or retrospective with patients presenting with acute vestibular syndrome. Studies combined if low clinical and statstitical heterogeonity. Study quality was assessed using the QUADAS tool. Random effects meta analysis using Revman 5 and SAS9.3 was performed. Results: 6 studies with 715 participants were included( QUADAS 12/14 SD 1.2). Average study length 5.3 years ( STD 3.3 years) . Prevalence of vertebrobasilar stroke ranged 9.3-76% (Mean 39.1% SD 17.1). The most common diagnosis were vertebrobasilar stroke (Mean 34.8% SD 17.1%), peripheral cause (Mean 30.9% SD 16%). Intra cerebral haemorrhage (Mean 2.2%, SD 0.5%). Neurologist/neuro ophthalmologist performed the exam in 5/6 studies. 1 study reported a kappa between emergency medicine physician and neurologist of 0.24-0.41. The HINTS exam had a sensitivity of 96% (CI 95% 0.92-0.98, I2-0%), Specificity 91.4% (CI 95% 64.5-98.4% I2 94%). Positive likelihood ratio 11.9 ( CI 95% 2.9-48.8) and a negative likelihood ratio of 0.04 ( CI 95% 0.01– 0.14). Conclusion: The HINTS exam has excellent diagnostic accuracy for ruling out stroke when performed by a neurologist. The lack of ER proven diagnostic accuracy and high prevalence of serious diagnosis in those presenting with acute vestibular syndrome suggests care should be taken in ruling out central cause of dizziness in this population.

Introduction: Emergency department (ED) return visits are used for quality monitoring. Health information technology (HIT) has historically supported return visit programs in the same hospital or hospital system. The Emergency Department Return Visit Quality Program (EDRVQP) is a novel population level continuous quality improvement (QI) program connecting EDs across Ontario that leverages HIT. We sought to describe the EDRVQP HIT architecture, experience of participants, enabling program factors and barriers. Methods: The Informatics Stack conceptual framework was used to describe the HIT architecture. A literature review of peer-reviewed background literature, and stakeholder organization reports was conducted. Purposive sampling identified key informants. Semi-structured interviews were conducted until saturation. Common themes were identified by inductive qualitative thematic analysis. Results: Twenty-three participants from 15 organizations were interviewed. The EDRVQP architecture description is presented across the Informatics Stack. The levels from most comprehensive to most basic are world, organization, perspectives/roles, goals/functions, workflow/behaviour/adoption, information systems, modules, data/information/knowledge/algorithms, and technology. Enabling factors were a high rate of EHR adoption, provincial legislative mandate for data collection and program membership, use of functional and data standards, local variability, phased deployment, and QI and patient safety culture. Two main barriers were increased case turnaround time and privacy legislation. Conclusion: The Informatics Stack framework provides a robust approach to thoroughly describe the HIT architecture of this population health programs. The EDRVQP is a population health program that illustrates the pragmatic use of continuous QI methodology across a population (provincial) level.