Introduction: Chest pain and symptoms of acute coronary syndrome are responsible for a large proportion of ED visits and acute hospitalizations. However, only about 15% of patients presenting to the ED with high-risk symptoms do, in fact, have an acute coronary syndrome. The objective of this study is to derive a 2-hour high-sensitivity Troponin T (hsTnT) testing algorithm with outcome based-cutoffs to rapidly rule out acute myocardial infarction (AMI) in a large proportion of ED chest pain patients. Methods: Patients included consecutive ED patients with a chief complaint of cardiac chest pain who had an hsTnT assay performed at ED arrival and 2 hours after ED arrival. Administrative databases were queried to identify troponin results and major adverse cardiac outcomes (MACE) including death, MI, and revascularization. Test characteristics of iterative combinations of initial troponin level and absolute change in troponin level were quantified in order to identify the testing algorithm that identified the greatest proportion of patients eligible for early discharge while maintaining a target sensitivity of 98.5% for the primary outcome of 7-day AMI. Results: 755 eligible patients had hsTnT assays performed at ED arrival and at 2 hours. 91 patients (12.1%) had a 7-day AMI while 108 (14.0%) had 7-day MACE. An initial hsTnT level of less than 14 ng/L, in combination with a 2-hour absolute change of less than 10ng/L had a sensitivity of 98.9% (95% CI 94.0,99.8) and an NPV of 99.8% (95% CI 98.7, 100.0) for 7-day AMI. This identified 58.5% of all patients as being suitable for early discharge. Sensitivity and NPV for 7-day MACE were 90.0% (95% CI 83.3, 94.2) and 97.3% (95% CI 95.3,98.4) respectively. Sex-specific differences in test characteristics were not clinically important. Rule-in hsTnT cutoffs were also evaluated, with specificities ranging from 85-95%, although cutoffs with higher specificity had less ability to rapidly rule-in AMI, leaving more patients with indeterminate results after 2 hours. Conclusion: A hsTnT algorithm can safely and accurately rule out AMI in 58.5% of ED chest pain patients within 2 hours of ED arrival. The lower sensitivity of this algorithm for MACE compared to AMI speaks to the importance of clinical assessment and ECG findings in identifying patients at risk for acute coronary syndromes.

Introduction: Background: A single-agent approach to children’s moderate to severe pain is often inadequate. To date, no studies have evaluated the combined use of oral morphine and ibuprofen for optimal pain management of children presenting to an Emergency Department (ED) for musculoskeletal (MSK) trauma. Objective: To assess the efficacy of a combination of oral morphine and ibuprofen for pain management in children with MSK trauma in the ED. Methods: A double-blind, placebo-controlled, multi-centered, three-arm, randomized clinical trial of 500 patients was conducted at three pediatric tertiary care EDs. Patients 6 to 17 years of age, who presented to the ED with a MSK trauma, and a score >30 mm on the 100 mm Visual Analogue Scale were eligible to participate. Patients were randomized (in a 2:1:1 ratio) to receive (orally): (a) morphine (0.2mg/kg) + ibuprofen (10mg/kg) (Group MOR + IBU) or (b) morphine (0.2 mg/kg) + placebo (Group MOR) or (c) ibuprofen (10mg/kg) + placebo (Group IBU). Primary outcome was pain intensity score under 30 mm (mild pain) at 60-minutes (T-60) after treatment administration. Results: A total of 456 patients were included in analyses: 177 (MOR + IBU), 188 (MOR), 91 (IBU). Mean age was 11.9 + 2.7 years, with a majority of boys (55.3%) and soft tissue injuries (62%). There were no differences in baseline characteristics in the three groups. Baseline mean pain score was 60.9 + 16.2 mm. Only 30% (MOR + IBU), 29% (MOR) and 30% (IBU) of patients reached a pain score under 30 mm at T-60 (p=0.83). Mean pain scores at T-60 were 42.3 + 23.2 mm (MOR + IBU), 43.8 + 23.1 mm (MOR) and 42.3 + 23.3 mm (IBU) (p=0.83). No severe adverse events were observed in any of the groups, at any of the study measurement points. Conclusion: Combination of morphine with ibuprofen did not provide any additional pain relief for children with MSK injuries, in the ED. None of the study medication provided optimal pain management, as the majority of children did not reach the WHO definition of mild pain. Alternative analgesic combinations should be investigated to optimize pain relief of children who present to the ED with MSK injuries.

Introduction: The objective was to study the operational impact of an intervention comprised of simultaneous process improvements to triage, patient inflow, and physician scheduling patterns on emergency department (ED) patient flow. The intervention did not require any increase in ED resources or expenditures. Methods: A 36-month pre-/post-intervention retrospective chart review at an urban community emergency department from January 2010 to December 2012. The ED process improvements started on June 6, 2011 and involved streamlining triage, parallel processing, flexible nurse-patient ratios, flexible exam spaces, and flexible physician scheduling. The main outcomes were ED length-of-stay (LOS). Secondary outcomes included time to physician-initial-assessment (PIA), left-without-being-seen (LWBS) rates, and left-against-medical-advice (LAMA) rates. Segmented regression of interrupted time series analysis was performed on Canadian Triage and Acuity Scale (CTAS) 2 to 5 patients to quantify the immediate impact of the intervention on the outcome levels, and whether there were changes in the trend between pre-intervention and post-intervention segments. Results: 251,899 patients attended the ED during the study period. Daily patient volumes increased 17.3% during the post-intervention period. Post-intervention, for CTAS 2-5 patients, there was a reduction in average LOS by 0.64 hours (p<0.001), and 90th-percentile LOS by 0.81 hours (p=0.024). When separated by acuity and disposition, there were reductions in LOS for non-admitted CTAS 2 (-0.58 hours, p <0.001), 3 (-0.75 hours, p <0.001), 4 (-0.32 hours, p=0.002), and 5 (-0.28 hours, p=0.008) patients. For secondary outcomes, there was a decrease in overall average PIA by 43.81 minutes (p<0.001), and 90th-percentile PIA by 91.39 minutes (p<0.001). LWBS and LAMA rates decreased by 35.2% (p<0.001) and 61.9% (p<0.001), respectively. Conclusion: A series of process improvements meant to optimize flow in the ED without the addition of resources was associated with clinically significant reductions in LOS, PIA, LWBS and LAMA rates for non-resuscitative patients.

Introduction: Patients seen primarily for hypertension are common in the emergency department. The outcomes of these patients have not been described at a population level. In this study we describe the characteristics and outcomes of the patients making these visits, as well as changes over time. Methods: This retrospective cohort study used linked health databases from the province of Ontario, Canada, to assess emergency department visits made between April 1, 2002 and March 31, 2012 with a primary diagnosis of hypertension. We determined the annual number of visits as well as the age and sex standardized rates. We examined visit disposition and assessed mortality outcomes and potential hypertensive complications at 7, 30, 90, 365 days and 2 years subsequent to the ED visit. Results: There were 206,147 qualifying ED visits from 180 sites. Visits increased by 64% between 2002 and 2012, from 15793 to 25950 annual visits, respectively. The age- and sex-standardized rate increased from 170/100,000 persons to 228/100,000 persons over the same time period, a 34% increase. Eight percent of visits ended in hospitalization, but this proportion decreased from 9.9% to 7.1% over the study period. Mortality was very low, at less than 1% within 90 days, 2.5% within 1 year, and 4.1% within 2 years. Among subsequent hospitalizations for potential hypertensive complications, stroke was the most frequent admitting diagnosis, but the frequency was still <1% within 1 year. Together hospitalizations for stroke, heart failure, acute myocardial infarction, atrial fibrillation, renal failure, hypertensive encephalopathy and dissection were <1% at 30 days. Conclusion: The number of visits made primarily for hypertension has increased dramatically over the last decade. While some of the increase is due to aging of the population, other forces are contributing to the increase. Subsequent mortality and complication rates are low and have declined. With current practice patterns, the feared complications of hypertension are extremely infrequent.

Introduction: Low back pain (LBP) may be having a significant impact on emergency departments (ED) around the world. Two analyses conducted in the USA and Australia suggest that LBP is one of the leading causes of emergency department visits. However, in the peer-reviewed literature, there has been limited focus on the prevalence and management of back pain in the ED setting. Furthermore, the applicability of the available research to our local ED setting is unclear. Methods: This project includes two studies to investigate the prevalence of LBP in the ED: 1. a comprehensive systematic review of the published literature to gather a comprehensive and global perspective about the prevalence of LBP in the ED setting, and 2. a retrospective cross sectional analysis using six years of data from our local ED, the Charles V. Keating Emergency and Trauma Centre, Halifax, Nova Scotia. Results: Searches from multiple databases including PubMed (392 citations), resulted in 3024 citations, of which 20 studies were found to have prevalence data for LBP. Studies were reported between 2001–2015 and used mixed methods of data collection, including electronic databases, surveys and patient charts. Ranges for prevalence estimates were 1.9% to 17% of patient visits. Results indicated there are many gaps in the literature, for example research in rural EDs and in Canada. In our primary study, we have identified a sample of 10 000 patients presenting with LBP to our local ED. Analysis of this data will be completed prior to the CAEP conference. Conclusion: This project is the first systematic review; comprehensive search strategy to examine the prevalence of LBP in the ED. It is also the first project to assess the prevalence of LBP in a Canadian ED. Results from this study will inform healthcare providers, as well as administrative and policy decision-makers, of the global and local impact of LBP in the ED, and will identify opportunities for further research to enhance care pathways of patients suffering from LBP.

Introduction: Patients with acute exacerbations of heart failure (HF) or chronic obstructive pulmonary disease (COPD) may be at high risk for preventable adverse events (AEs). Preventable AEs are ED care-associated complications due to medical error. Our objective was to identify and characterize preventable AEs among ED patients over 50 presenting with dyspnea from an acute exacerbation of HF or COPD; who were subsequently admitted or discharged. Methods: We conducted a multicentre health records review from six academic centers in Ontario and Alberta. We analysed health records for all prospectively enrolled patients who experienced flagged outcomes: relapse to ED within 14 days requiring admission; admission to a monitored unit (AMU), cardiac care unit(CCU), or intensive care unit(ICU); intubation(ETI); non-invasive ventilation(NIV); diagnosis of acute myocardial infarction(AMI); or death within 30 days. Using a validated approach, an ED physician analyzed case summaries for flagged outcomes that were associated with ED care, designated as AEs. Preventable AEs had contributing errors in diagnosis, management, procedure, medications or unsafe disposition decisions. We analyzed these data using thematic coding and descriptive statistics. Results: Of 2,515 patients enrolled (1,100 HF and 1,415 COPD), 210 patients experienced flagged outcomes, 47.1% of which were female, 64.3% had HF and the remaining COPD. The majority (86.2%) of flagged outcomes were related to underlying disease, but 13.8% of cases met criteria for AE and all were deemed preventable. Of the identified AEs, 72.4% returned to the ED and required admission to hospital; 17.2% were admitted to ICU, CCU, or AMU; 6.9% of patients died; 3.4% were intubated; 3.4% had a diagnosis of AMI and 0% required NIV. We found 75.8% of preventable AEs resulted from a management error (eg. not prescribing steroids on discharge for moderate COPD exacerbation); 31.0% from an unsafe disposition decision and 10.3% of AEs resulted from diagnostic error. Conclusion: Patients with acute exacerbations of HF and COPD are at high risk of preventable AEs directly related to care provided in the ED. Management and disposition decisions were a concerning source of error and should compel and focus future quality improvement efforts.

Introduction: While a majority of patients presenting to the emergency department (ED) are discharged home without need for inpatient hospitalization, many require outpatient follow-up. Currently, outpatient referrals from our ED are made via a complex and error-prone series of manual steps which have the potential to be unreliable and negatively impact quality of care. We sought to perform a current state analysis of our outpatient referral processes across the hospital’s specialties. Methods: We conducted a retrospective health records review at a tertiary academic centre (>160,000 ED visits/year) from January 1 to January 7, 2015. All consecutive outpatient consultation requests triggered by an ED physician were identified and included for chart review. All cases were subsequently followed up to 11 months. A single reviewer extracted data on demographics, actual referral attendance rates, incomplete referrals, return ED visits, and time intervals. The top 3 and bottom 3 performing services were identified for further analysis of their outpatient referral mechanisms and processes. We present descriptive statistics. Results: A total of 251 outpatient referrals to a broad range of specialty services were identified during the study period. 216 (86.1%) of patients attended the intended appointment, while 35 (13.9%) of referrals were incomplete at 11 months post index ED visit. The overall median time to successful outpatient follow-up appointments was 8.5 days [IQR=3.8-24.2]. 8 (3.2%) patients had a return ED visit for a related complaint prior to being seen at their outpatient appointment. The top 3 performers were Ophthalmology [Median=1.0 day, IQR=0.0-1.0, Incomplete=2.8%], Plastic Surgery [Median=5.0 days, IQR=2.8-6.0, Incomplete=7.7%], and Orthopedics [Median=8.0 days, IQR=7.0-10.0, Incomplete=0.0%]. The bottom 3 performers were Dermatology [Median=52days, IQR=41.5-92.5, Incomplete=25.0%], Neurology [Median=40.0 days, IQR= 2.5-43.5, Incomplete=56.3%], and Urology [Median=14.0 days, IQR=10.5-48.0, Incomplete=33.3%]. Conclusion: We found a tremendous range of variability in both the waiting times and actual reliability of outpatient referral processes from the ED. Future phases of this project will focus on examining specific processes of the top and bottom performing specialties in order to improve and standardize all outpatient referrals.

Introduction: Short-term risk of arrhythmia or death among emergency department (ED) syncope patients with atrial fibrillation/flutter (AFF) has not been reported in the literature. Our objectives were to assess the incidence and the independent risk of 30-day arrhythmia or death for syncope patients with AFF after ED disposition. Methods: We conducted a prospective study at 6 Canadian academic EDs to include adults with syncope. We collected demographic, clinical and ECG characteristics while our outcome assessments were completed by medical records review and by telephone follow-up of patients after 30 days. Primary outcome was arrhythmia or death within 30-days after ED disposition and secondary outcomes included non-arrhythmic cardiac and non-cardiac outcomes. We performed descriptive and logistic regression analyses. Results: We enrolled 4,266 patients: mean age 53.4 years, 55.4% females, and 8.5% with AFF. After excluding those with outcomes in the ED, lost to follow-up and those with other non-sinus rhythms, 3,417 patients in the sinus and 280 patients in the AFF groups were analyzed. The incidence of arrhythmia or death was significantly higher in the AFF group (Relative Risk 5.1; 95% CI 3.1-8.4; p<0.0001) but there were no significant differences in secondary outcomes between the groups. The unadjusted odds ratio for 30-days arrhythmia or deaths among ED syncope patients with AFF was 5.4 (95% CI 3.2- 9.2). After adjusting for important baseline risk factors by multivariable analysis, the odds ratio for arrhythmia or death in patients with AFF was 1.5 (95% CI 0.8-2.7). Conclusion: The risk of AFF for 30-day arrhythmia or death among syncope patients after ED disposition is higher but is attenuated when adjusted for important patient characteristics. Future research should assess long-term outcomes among syncope patients with AFF to guide follow-up after ED discharge.

Introduction: Chest pain is one of the most common reasons for emergency department (ED) visits in developed countries. Patients discharged after ED assessment remain at risk for adverse cardiac events. Although a volume-outcome relationship has been shown for myocardial infarction, it is uncertain whether a similar relationship exists with ED chest pain volume. Accordingly, we aimed to determine whether ED chest pain volume influences outcomes of patients presenting to the ED with chest pain who were discharged home. Methods: This was a retrospective cohort study using population-based data from Ontario, Canada. Patients who were discharged home from an ED in Ontario with a primary diagnosis of chest pain from April 1, 2004 to March 31, 2010 were included. High-risk patients were defined as the presence of diabetes or pre-existing cardiovascular disease, while low-risk patients were defined as the absence of these conditions. ED volume was categorized as low, medium, or high, based on tertiles of annual chest pain patient volume. The primary outcome of this study was all-cause mortality one year after the index ED visit. Mantel-Haenszel Chi-Square was used to compare crude outcome rates. Results: There were 56,767 high-risk patients. The average age was 66 years and 53% were male. All-cause mortality rates were 6.8%, 6.3%, and 6.0% (p=0.028), and rates of hospitalization for acute coronary syndrome were 5.8%, 4.6%, and 4.0% (p<0.001) among low, medium, and high volume EDs respectively. There were 216,527 low-risk patients. The average age was 64 years and 42% were male. All-cause mortality rates were 2.0%, 1.9%, and 1.6% (p<0.001), and rates of hospitalization for acute coronary syndrome were 1.5%, 1.4%, and 1.0% (p<0.001) among low, medium, and high volume EDs respectively. Conclusion: Higher volume EDs were associated with decreased rates of all-cause mortality and admission for acute coronary syndrome among chest pain patients who were discharged home. Future research should study the reasons for this finding and attempt to improve outcomes in lower volume EDs.

Introduction: Acute Aortic Dissection (AAD) is life threatening, requiring early diagnosis. Although previous literature suggest interarm BP differential is an independent predictor of AAD, up to 20% of a healthy population can have a significant differential. Our objectives were to assess the rate of bilateral BP measurement in acute non-traumatic truncal pain patients, and the association of BP differential with non-traumatic AAD. Methods: This is a historical matched case control study: participants were adults >18 years old presenting to two tertiary care EDs with a triage diagnosis of truncal (i.e. chest, abdominal, flank, back) pain. Cases were selected based on an ED or in-hospital diagnosis of non-traumatic AAD confirmed by CT or Echo. Controls were from a single calendar year matched in a 1:1.5 ratio by sex and age within 5 years. ED and referral consult BP measurements were used. Exclusion criteria: clear diagnosis on basic investigation (i.e. UTI, pneumonia, pneumothorax, acute fracture) or pain >14 days/no pain. Sample size of 126 cases and 183 controls was calculated based on 20% exposure in controls (80% power and alpha of 5%), to detect an OR >2. P-values were calculated using chi square analysis. Results: A total of 294 (119 cases, 175 controls) patients were included (mean 66+/-14.5yrs, 59.5% male). Cases (199 potential: 119 included; 80 excluded). Controls (8239 potential: 305 reviewed; 175 included; 130 excluded). Bilateral BP was measured in 70.6% of cases (n=84, mean difference=15.5mmHg) versus 31.3% of controls (n=55, mean difference=10.9mmHg). Among included controls, most common diagnoses were: Unspecified Chest (36.0%) or Abdominal (9.7%) Pain, ACS (12.6%), Muscular Back Pain (5.1%), and Renal Colic (4.0%). BP differential >10mmHg was found in 58.8% of cases and 40.7% of controls (P=0.10). A BP differential >20mmHg was found in 31.3% of cases and 22.2% of controls (P=0.37). BP differential >20mmHg did not significantly increase the odds of AAD (OR 2.0 (95%CI 0.82-4.90), p<0.129). Conclusion: Interarm BP differential is not routinely measured in ED patients with acute non-traumatic truncal pain, and there is no significant difference in the presence or magnitude of differentials in patients with or without AAD. Therefore, physicians should not rely on BP differentials to aid in their diagnosis or exclusion of AAD.

Introduction: Emergency department (ED) overcrowding is a common and complicated challenge for EDs worldwide. Nurse-initiated protocols, diagnostics and/or treatments implemented by nurses prior to patients being seen by a physician or nurse practitioner, have been suggested as a potential strategy to improve patient flow. Methods: This randomized, pragmatic, controlled evaluation of 5 nurse-initiated protocols occured in a crowded inner-city ED. Six physicians and 44 registered nurses, 3 clinical nurse educators and 3 unit managers were involved in revising 5 patient-complaint focused protocols prior to evaluation. Thirty (30/180) emergency nurses were provided 1 hour of training on inclusion and exclusion criteria, procedure and evaluation methods. Data was abstracted in a manner concealing patient allocation. Primary outcomes evaluated included time to diagnostic test, treatment, consultation or ED length of stay. This evaluation was completed following both the CONSORT and SQUIRE guidelines. Results: Time to acetaminophen for the intervention group (n=11) was 1h:04 min on average (95%CI 30min to 1h:37min) whereas the control group (n=9) was 3h:35min (95%CI 2h:21min to 4h:48min). The average length of stay of a suspected fractured-hip in the intervention group (n=5) was 3h:34min (95%CI 1h:49min to 5h:19min) and 7h:34min for the control group (n=4) was (95%CI 5h:26min to 9h:42min). Time to troponin in the intervention group (n=29) was one quarter (average 48min, 95% CI 32min to 64min) of the time it was in the control group (n=14) (average 3h:16min, 95%CI 1h:53min to 4h:39min; p < 0.001). The vaginal bleeding in pregnancy protocol reduced length of stay by roughly fifty-percent; the intervention group (n=11) had a length of stay of 4h:57min (95%CI 3h:46min to 6h:08min) compared to 8h:33min (95% CI 6h:23min to 10h:44min) for the control (n=7) (p < 0.001). There was no statistical difference in the length of stay for patients who received protocolized diagnostics for abdominal pain. Conclusion: Targeting specific patient groups with carefully written protocols can improve the timeliness of care. A cooperative and collaborative interdisciplinary group are essential to success. Having a system in place to ensure ongoing quality in protocol application and interdisciplinary support has proven more difficult than improving the primary outcomes in this evaluation.

Introduction: Emergency department (ED) crowding is a national challenge. Initiatives to help address this at our ED include the use of a six-bed fast-track unit staffed by advanced-care paramedics (ACPs). Institutional byelaws only allow diagnostic imaging (DI) ordering by physicians (MD). An ACP requesting DI at the time of first assessment would likely improve patient flow. We investigated whether ACPs can safely and cost-effectively request DI for extremity injuries without increasing cost or exposing patients to unnecessary radiation. Methods: A prospective evaluation of a convenience sample of patients presenting with an extremity injury sustained within 48 hours of presentation. At time of initial assessment, the ACP, following specific guidelines, recorded whether or not they believed an x-ray was indicated, and if so, what DI views they felt appropriate. Their opinion was blinded from the physician subsequently assessing the patient. An ACP opinion of the need for DI was compared with the subsequent test ordered by the MD. The MD decision to order DI was considered ‘gold standard’. Opinions were considered “matched” if the MD ordered DI of the same body part that the ACP believed was indicated. Sensitivity, specificity, positive predictive and negative predictive values (PPV, NPV) were calculated. Using data from our ED information system, we estimated the time that would have been saved by allowing ACPs to order DI. Results: Of 199 patients 192 images were ordered and 89 fractures were diagnosed. ACPs and MDs agreed that DI was necessary 94.70% of the time (95% CI: {90.6%, 97.4%}). There were 8 x-rays the ACP did not order that the MD did order, of which one showed a fracture. Twice, the ACP would have ordered an x-ray that the MD did not. In terms of identifying the need for DI, ACPs were 95.8% sensitive and 71.4% specific. The PPV was 98.9% (95% CI: {95.8%, 99.8%}), and the NPV was 38.5% (95% CI: {15.1%, 67.7%}). On average, ACP opinion of DI indication was made 54.1 minutes (95% CI: {48.0, 60.2}) earlier that of the MD. Conclusion: The overall agreement between MDs and ACPs was almost 95%. ACPs are more likely to under-order x-rays than to over-order them, lowering the risk of increasing radiation exposure and cost. ACP DI ordering may decrease the time of processing of patients with extremity injuries by almost an hour.

Introduction: The Physician Navigator (PN) is a novel position created to manage patient flow in real-time at a very-high volume emergency department (ED). When paired with an emergency physician, PNs actively track patient wait times, and direct the physician to see and re-assess patients in a particular order to improve measures of emergency department efficiency, and maximize patient flow. Anecdotal evidence has shown that PNs decrease length-of-stay times for non-resuscitative patients in the setting of increased patient volumes, and without additional nursing or physician hours. The objective was to study the operational impact of PN on emergency department patient flow. Methods: A 48-month pre-/post-intervention retrospective chart review at an urban community emergency department from September 2011 to September 2015. The PN program started on March 1, 2013. The main outcome is emergency department length-of-stay (LOS). Secondary outcomes include time to physician-initial-assessment (PIA), left-without-being-seen rates (LWBS), left-against-medical-advice (LAMA), and physician satisfaction rates. Autoregressive integrated moving average models were generated for Canadian Triage and Acuity Scale (CTAS) 2 to 5 patients to quantify the immediate impact of the intervention on the outcome levels, and whether the impact was sustained over time. Results: Interim results are provided. 399,958 patients attended the ED during the study period. Daily patient volumes increased 11.2% during the post-intervention period. There were no significant increases in the number of physicians shifts/day, and physician hours/day during the post-intervention period. Post-intervention, for CTAS 2-5 patients, there was a reduction in average LOS by 0.04 hours/PN (p<0.05), and 90th-percentile LOS by 0.14 hours/PN (p<0.05). For secondary outcomes, there was a decrease in overall average PIA by 6.37 minutes/PN (p<0.05), and 90th-percentile PIA by 8.29 minutes/PN (p<0.05). LWBS rates decreased by 40.8% (p<0.05). There were no significant changes in LAMA rates. Conclusion: The implementation of Physician Navigators is associated with significant reductions in LOS, PIA, and LWBS rates for non-resuscitative patients at a very-high volume emergency department.

Introduction: TIAs precede about 30% of strokes, with 4-10% having a stroke within 90 days of their TIA. In patients with a TIA due to symptomatic carotid disease, diagnosis and treatment within 2 weeks has been shown to have much better outcomes, while delay beyond 12 weeks no longer reduces subsequent stroke risk. The objective of this study was to determine the clinical findings associated with symptomatic critical disease following an ED visit for TIA to indicate patients requiring prompt carotid imaging. Methods: We performed a prospective Canadian multicenter cohort study, at 13 academic sites, of ED patients with TIA or non-disabling stroke from 2006-2014. Treating ED physicians indicate clinical features on standardized data collection forms. Symptomatic carotid disease was carotid stenosis 50-99%, or carotid dissection, adjudicated by stroke neurology to be the etiology of the index event. Patients were followed by medical review and telephone up to 90 days. Univariate analysis was conducted for clinical features associated with patients who were eventually found to have symptomatic carotid disease as a cause for their TIA. Results: The cohort included 305 patients with and 5,277 without symptomatic carotid disease. Positive predictors of symptomatic carotid disease included older age (74.0 yrs vs 68.0 yrs p<0.0001), male sex (62.9% vs 47.9%; p<0.0001), history of weakness (63.3% vs 41.4%; p<0.0001), language disturbance (52.1% vs 40.0%; p<0.0001), weakness on physical exam (25.5% vs 17.1%; p=0.0002), history of hypertension (74.8% vs 59.5%; p<0.0001), and known history of carotid stenosis (18.9% vs 3.1%; p<0.0001). Negative predictors of symptomatic carotid disease included first ever TIA (56.8% vs 68.8%; p<0.0001), history of altered sensation (39.4% vs 45.8%; p=0.0322), lightheadedness (13.0% vs 22.4%; p=0.0002), and vertigo (3.6% vs 12.7%; p<0.0001). Conclusion: TIA patients with older age, male sex, weakness, language disturbance or history of carotid stenosis need to be promptly imaged to assess for symptomatic carotid disease.

Introduction: National guidelines (NICE, AHA) for management of Acute Cerebrovascular Syndrome (ACVS) in the Emergency Department (ED) recommend the use of ABCD2 score to risk stratify patients despite its poor specificity and low diagnostic accuracy. The SpecTRA project previously developed a clinical classifier for ACVS vs. Mimic derived from historical clinical data collected during a 5-year period at an outpatient stroke clinic (Victoria, BC). Here we present a prospective evaluation of the performance of our clinical classifier on prospectively collected ED patient data compared to the industry-standard ABCD2. Methods: The prospective cohort consisted of ED patients (N=555, Male=54%, Mean (SD) Age=68.7(15.5), ACVS=70%) enrolled between Jan 2014 and May 2015 at Victoria General Hospital (BC) and Foothills Medical Centre (Calgary, AB). ABCD2 and clinical classifier scores were calculated from clinical data from the ED. We compared the performance of the two classifiers using DeLong’s test of Dependent Receiver Operating Curves (ROC). In keeping with national guidelines, we used a score of 4 or more to assess sensitivity, specificity and accuracy (sens/spec and acc) of the ABCD2; for our clinical classifier, we used the cut point previously determined to maximize agreement between predictions and true class labels in the historical data. Results: Our new clinical classifier significantly outperformed the ABCD2 (z=2.44, p=0.015) with an AUC of 0.72, (95% CI: 0.68, 0.77) vs. 0.66 (0.61, 0.71). In terms of sens/spec and acc, our classifier achieved 0.78/0.55 with acc 71% compared to 0.75/0.46 with acc 66% for the ABCD2 (using the previously specified cut points). Conclusion: Our ACVS clinical classifier showed better performance than the ABCD2 score on a prospective sample of ED patients. The improved specificity of the clinical classifier relative to existing prognostic tools would reduce the number of non-ACVS patients referred for early treatment as well as conserve medical resources. Our ongoing multi-site study will evaluate the utility of the ACVS classifier embedded in a logic-enabled e-fillable form. This form will also provide risk-based thresholds guiding timely ordering of CTA as well as links to clinical treatment guidelines. Longer-term, the e-form and classifiers will be further enhanced to include plasma-based protein biomarker data.