Introduction: Accurate forecasting of emergency department (ED) patient visits can inform better resource matching. Calendar variables such as day of week and time of day are routinely used as predictors of ED volume. Further improvement in forecasting will likely come from dynamic variables. The effect of snowfall on ED volumes in colder climates remains poorly understood. We sought to determine whether accounting for snowfall improves ED patient volume forecasting. Our secondary objective was to characterize the magnitude of effect of snowfall on ED volume. Methods: This was a retrospective observational study using historical patient volume data and local snowfall records from April 1st, 2011 to March 31st, 2018 (2,542 days) at a single urban ED. We fit a series of four generalized linear models: a baseline model which included calendar variables and three different snowfall models which contained the variables in the baseline model plus an indicator variable for modelling snowfall. Each snowfall model had a different daily threshold for its indicator variable: any snowfall ( >0cm), moderate snowfall ( > = 1 cm), or high snowfall ( > = 5 cm). We modeled daily ED volume as the dependent variable using a Poisson distribution. To evaluate model fit, we examined the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) in each of the four models. In both cases, a lower number indicates better model fit. Incident rate ratios were calculated to determine the effect of snowfall. We used the delta method to calculate confidence intervals. Results: A total of 2542 days were used to develop the model. All three snowfall models demonstrated improved model fit compared to the baseline model with lower AIC and BIC values. The best fitting model included a binary variable for moderate snowfall ( > = 1cm/day). This model showed a statistically significant decrease in ED volume of 2.65% (95% CI: 1.23% -4.00%) on snowfall days, representing 5.4 (95% CI: 2.5 -8.2) patients per day at our hospital with an average daily volume of 205 patients. Conclusion: The addition of a snowfall variable results in improved forecasting model performance in ED volume forecasting with optimal threshold set at 1 cm of snow in our setting. Snowfall is associated with a modest, but statistically significant reduction in ED volume.

Introduction: Emergency department (ED) congestion is an ongoing threat to quality care. Traditional measures of ED efficiency use census and wait times over extended time intervals (e.g. per year, per day), failing to capture the hourly variations in ED flow. Borrowing from the traffic theory framework used to describe cars on a freeway, ED flow can instead be characterized by three fundamental parameters: flux (patients traversing a care segment per unit time), density (patients in a care segment per unit time), and duration (length of stay in a care segment). This method allows for the calculation of near-instantaneous ED flux and density. To illustrate, we examined the association between stretcher occupancy and time to physician initial assessment (PIA), seeking to identify thresholds where flux and PIA deteriorate. Methods: We used administrative data as reported to government agencies for 115,559 ED visits from April 1, 2014 to March 31, 2016 at a tertiary academic hospital. Time stamps collected at triage, PIA, and departure were verified by nosologists and used to define two care segments: awaiting assessment or receiving care. Using open-source software developed in-house, we calculated flow measures for each segment at 90-minute intervals. Graphical analysis was supplemented by regression analysis, examining PIA times of high (CTAS 1-3) or low (CTAS 4-5) acuity patients against ED occupancy (=density/staffed stretchers) adjusting for the day of the week, season and fiscal year. Results: At occupancy levels below 50%, PIA times remain stable and flux increases with density, reflecting free flow. Beyond 50% occupancy, PIA times increase linearly and flux plateaus, indicating congestion. While PIA times further deteriorate above 100% occupancy, flow is maintained, reflecting care delivery in non-traditional spaces (e.g. hallways). An inflection point where flux decreased with increased crowding was not identified, despite lengthening queues. Conclusion: The operational performance of a modern ED can be captured and visualized using techniques borrowed from the analysis of vehicular traffic. Unlike cars on a jammed roadway, patients behave more like a compressible fluid and ED care continues despite high degrees of crowding. Nevertheless, congestion begins well below 100% occupancy, presumably reflecting the need for stretcher turnover and saturation in subsegmental work processes. This methodology shows promise to analyze and mitigate the many factors contributing to ED crowding.

Introduction: Planning for the future emergency physician (EP) workforce will be a significant challenge for decision makers given the rise in emergency department (ED) visits and no concurrent increase in resident positions. EP workforce planning must incorporate physician supply, as well as current and forecasted patient demand. Nova Scotia has undertaken the process of developing a planning model to support policy decision making. We hypothesize that Nova Scotia will require increased resident positions and recruitment from other provinces to meet future patient demand. Methods: We have developed an age structured population model that tracks the number of clinical full-time equivalent (FTE) EPs by their age and shows the “variance” (i.e., supply – demand = variance) over a 30 year planning horizon. This model represents all Level 1, 2, 3, and 4 EDs in Nova Scotia. Current physician supply was calculated based on FTE staffing levels. The current patient demand was based on historical volume and acuity of patients and converted to an FTE demand estimate. Forecasted demand was predicted to increase at an average rate of 0.5% per year. We varied the number of residents trained and the number of EPs recruited from outside the province to examine the effect on the EP workforce. Our initial model will reflect the current training environment and will be referred to as the “current state”. In our 3 scenarios, we increased the number of residents and recruited physicians by 50%, individually and then together. Our outcome measure will be the variance in FTE. Results: The current state showed that the province will have a deficit of 51 FTE EPs over the next 30 years. In scenario 1, a 50% increase in both resident training streams eliminated all variance, while in scenario 2, the increase in recruitment reduced the FTE variance to 34 FTE positions unfilled. In scenario 3, the variance was 0. Conclusion: We feel that this CTAS weighted volumes perspective is important for clinical services planning but the siting, sizing, and synergizing of EDs in a region will involve other inputs. Its important to recognize that we have made the assumption that all physicians starting to work in Nova Scotia will be a 1 FTE. Future iterations will examine the effect of more realistic FTE definitions that account for administrative, teaching and research activities.

Introduction: The Ontario emergency department (ED) Return Visit Quality Program (RVQP) launched in 2016 and aims to promote continuous quality improvement (QI) in the province's largest EDs. The program mandates routine audits of cases involving patients who had ED return visits within 72hrs that led to admission to hospital, in order to identify quality issues that can be tackled through QI initiatives. Our objective was to formally evaluate how well the RVQP achieved its aim of promoting continuous QI at participating sites using the constructivist grounded theory. Methods: Using a semi-structured interview guide, we employed a maximum variation sampling approach to ensure diverse representation across several geographical and institutional experiences (e.g., urban vs. rural, academic vs. community). Selected RVQP program leads were invited to participate in a phone interview to yield maximal insight, additionally using a snowball sampling approach to reach non-lead physicians to capture the penetration of the program. Interviews were conducted until thematic saturation was reached and no new insights were gleaned. Interviews were initially cross-performed by two members of the research team, recorded, transcribed, and de-identified. Data analysis was conducted using a constant comparative approach through the development of a coding framework and triangulation with the respondents’ ED setting. We then grouped, compared and refined our analytic categories through an inductive, iterative approach. Results: Between June and August 2018, we interviewed 32 participants, including 21 RVQP program leads and 11 non-lead physicians, from a total of 23 diverse sites (out of 84). Our analysis suggests that the RVQP provides a structured method for EDs to frame the continuous collection of data in order to channel activities towards quality improvement projects based on identified needs. Success factors included: greater involvement with QI processes prior to the RVQP leading to more openness to improvement, a more collaborative approach to RVQP implementation which led to greater front-line workers’ understanding and engagement, and more resources dedicated to implementing the RVQP as well as tackling the quality issues it identified. Conclusion: This study evaluated the impact of an innovative and large-scale program aimed at improving the culture of quality in Ontario EDs. While the program is still relatively new, early results show that there are key elements of EDs that support building a culture of QI.

Introduction: Access block (AB) is the most important indicator of Emergency Department (ED) crowding, but the impact of AB on consultation time has not been described. Our objectives were to determine if ED AB affects inpatient service consultation time, and operational and patient outcomes. Methods: We conducted a health records review of all ED patients referred and admitted at a university-affiliated tertiary care hospital over 60-days. A computational algorithm determined hourly ED AB at the time of consultation request, and observational cohorts were determined based on ED AB high (>35% ED bed capacity occupied by admitted patients) or low (<35%). The outcomes included total consultation time (TCT), ED physician initial assessment (PIA) time, ED length of stay (LOS), transfer time to inpatient bed (TTB), hospital LOS, return to ED (RTED) within 30 days, and 30-day mortality. Results: We included 2,871 patients (48% male; M = 63 years, IQR 45–78), and the low AB cohort were higher acuity (N = 1,692; 50.4% CTAS 1–2) than the high AB cohort (N = 1,179; 47.1% CTAS 1–2). Median TCT was not significantly different (low = 209min, high = 212min; p = 0.09), and there was no difference in consults completed within the 3-hour institutional time target (low = 41.1%, high = 40.9%; p = 0.89). Median ED PIA time was not significantly different (low = 66min, high = 68min; p = 0.08), however, patients seen within the funding-associated provincial ED PIA time target was significantly less during high AB (high = 82.2%, low = 89.2%; p < 0.001). Median ED LOS was significantly longer during high AB (high = 12.1hr, low = 11.1hr; p = 0.009), but median hospital LOS was not different (high = 109.5hr, low = 112.4hr; p = 0.44). Median TTB was significantly longer during high AB (high = 8.0hr, low = 5.9hr; p = 0.0004). There was no difference in RTED visits (high = 12.4%, low = 10.6%; p = 0.15) or 30-day mortality (high = 8.4%, low = 9.2%; p = 0.51). Conclusion: In conclusion, consultation time is not affected by AB. However, boarding admitted patients in the ED impairs our ability to meet funding-associated performance metrics. Reducing boarding time should be an ED and hospital-wide priority, as it negatively impacts funding and delays patient care.

Introduction: In addition to its clinical utility, the Canadian Triage and Acuity Scale (CTAS) has become an administrative metric used by governments to estimate patient care requirements, emergency department (ED) funding and workload models. The electronic Canadian Triage and Acuity Scale (eCTAS) initiative aims to improve patient safety and quality of care by establishing an electronic triage decision support tool that standardizes that application of national triage guidelines across Ontario. The objective of this study was to evaluate triage times and score agreement in ED settings where eCTAS has been implemented. Methods: This was a prospective, observational study conducted in 7 hospital EDs, selected to represent a mix of triage processes (electronic vs. manual), documentation practices (electronic vs. paper), hospital types (rural, community and teaching) and patient volumes (annual ED census ranged from 38,000 to 136,000). An expert CTAS auditor observed on-duty triage nurses in the ED and assigned independent CTAS in real time. Research assistants not involved in the triage process independently recorded triage time. Interrater agreement was estimated using unweighted and quadratic-weighted kappa statistics with 95% confidence intervals (CIs). Results: 1491 (752 pre-eCTAS, 739 post-implementation) individual patient CTAS assessments were audited over 42 (21 pre-eCTAS, 21 post-implementation) seven-hour triage shifts. Exact modal agreement was achieved for 567 (75.4%) patients pre-eCTAS, compared to 685 (92.7%) patients triaged with eCTAS. Using the auditor's CTAS score as the reference standard, eCTAS significantly reduced the number of patients over-triaged (12.0% vs. 5.1%; Δ 6.9, 95% CI: 4.0, 9.7) and under-triaged (12.6% vs. 2.2%; Δ 10.4, 95% CI: 7.9, 13.2). Interrater agreement was higher with eCTAS (unweighted kappa 0.89 vs 0.63; quadratic-weighted kappa 0.91 vs. 0.71). Research assistants captured triage time for 3808 patients pre-eCTAS and 3489 post implementation of eCTAS. Median triage time was 312 seconds pre-eCTAS and 347 seconds with eCTAS (Δ 35 seconds, 95% CI: 29, 40 seconds). Conclusion: A standardized, electronic approach to performing CTAS assessments improves both clinical decision making and administrative data accuracy without substantially increasing triage time.

Introduction: CTAS triage acuity and CEDIS complaint categories are used to prioritize patients for rapid treatment and ED resource allocation. Our objective was to evaluate CTAS and CEDIS validity for risk stratification of ED patients with chest pain using data from two Canadian cities. Methods: This administrative database study included patients seen over a five-year period with a triage complaint of chest pain. Our composite primary outcome included 7-day mortality, cardiac arrest, acute coronary syndrome (ACS) diagnosis (STEMI, NSTEMI, unstable angina{UA}), admission to a critical care unit, or hospitalization with CHF, pulmonary embolism, dysrhythmia, aortic pathology, neurologic or respiratory diagnosis. We dichotomized triage assignments to cardiac vs. noncardiac chest pain and high (CTAS 1,2) vs. low (3,4,5) triage acuity. For our secondary outcome we reported the components of the primary composite outcome. Results: We studied 111,824 patients. The most common overall diagnoses were chest pain NYD (53.8%), ACS (8.9%), musculoskeletal (7.4%), and acute respiratory (5.5%) or GI (5.1%) conditions. Of all patients studied, 85,888 (76.8%) were placed in the “cardiac features” group, and 93,257 (83.4%) fell into high acuity CTAS 1-2. Patients triaged into the “cardiac features” group were more likely to have a composite outcome event (16.6% v. 6.7%; p < 0.001), to be admitted (21.8% v. 9.0%), to require critical care (6.0% v. 0.7%), to receive an ACS diagnosis (11.3% v. 0.9%), and to die within 7 days (0.5% v. 0.2%). Patients in high acuity triage levels were also more likely to have a composite outcome event (15.8% v. 3.3%; p < 0.001), to be admitted (25.4% v. 14.3%), to require critical care (8.2% v. 1.2%), to receive an ACS diagnosis (10.5% v. 0.9%), and to die within 7 days (0.5% v. 0.2%). Conclusion: This study shows that triage assignment is strongly correlated with important patient outcomes and that both the chief complaint and acuity level are powerful risk predictors. These findings may differ at other sites and hospitals should assess and evaluate their data.

Innovation Concept: High-quality cardiopulmonary resuscitation saves lives; however, current certification standards can leave providers poorly prepared to perform effective chest compressions (CCs). We designed a training program based on the emerging model of skill maintenance through frequent short practice sessions. The ideal frequency of training is currently unknown. Our goal was to provide medical students with access to efficient and effective CC training and to determine an optimal training interval. Methods: Thirty-six second-year medical students were randomized to three groups that trained at different frequencies: once every two months (q2m) (n = 12), once every four months (q4m) (n = 13), and control (n = 11). Study duration was eight months with the intervention groups, q2m and q4m, participating in five and three sessions respectively. The control group was assessed at study start and end, receiving no training in between. At each session, participants completed a one-minute pre-test of CC performance, viewed a one-minute training video, practiced CCs for two minutes with real-time feedback, and completed a one-minute post-test. Performance parameters measured were CC depth, rate, release, and hand positioning. A final “compression score” assessed integrated performance across these parameters and served as our primary outcome. Participants also reported pre- and post-training comfort with performing CCs which served as our secondary outcome. Curriculum, Tool or Material: Our “Quick Refresher Sessions” (QRS) were completed by participants independently without requiring an assessor or facilitator. A manikin with the ability to record and provide real-time quantitative feedback on CC quality was connected to a laptop running a customized interface. Participants typed in an individualized code and were guided through their six-minute sessions automatically. Conclusion: Immediately following the first training session, subjects had significant improvement in compression score (p < 0.001) and skill comfort (p < 0.001). At eight months, both intervention groups, q2m and q4m, achieved higher compression scores than control (p = 0.001 and p = 0.011) and showed greater increase in comfort level (p = 0.002 and p = 0.010). Performance between intervention groups at eight months was not statistically different. Overall, we conclude that independent QRS training every two or four months led to improved CC quality and provider comfort. Future directions include increasing sample size and tailoring training intervals to individual performance.

Innovation Concept: The outcome of emergency medicine training is to produce physicians who can competently run an emergency department (ED) shift. While many workplace-based ED assessments focus on discrete tasks of the discipline, others emphasize assessment of performance across the entire shift. However, the quality of assessments is generally poor and these tools often lack validity evidence. The use of entrustment scale anchors may help to address these psychometric issues. The aim of this study was to develop and gather validity evidence for a novel tool to assess a resident's ability to independently run an ED shift. Methods: Through a nominal group technique, local and national stakeholders identified dimensions of performance reflective of a competent ED physician. These dimensions were included in a new tool that was piloted in the Department of Emergency Medicine at the University of Ottawa during a 4-month period. Psychometric characteristics of the items were calculated, and a generalizability analysis used to determine the reliability of scores. An ANOVA was conducted to determine whether scores increased as a function of training level (junior = PGY1-2, intermediate = PGY3, senior = PGY4-5), and varied by ED treatment area. Safety for independent practice was analyzed with a dichotomous score. Curriculum, Tool or Material: The developed Ottawa Emergency Department Shift Observation Tool (O-EDShOT) includes 12-items rated on a 5-point entrustment scale with a global assessment item and 2 short-answer questions. Eight hundred and thirty-three assessment were completed by 78 physicians for 45 residents. Mean scores differed significantly by training level (p < .001) with junior residents receiving lower ratings (3.48 ± 0.69) than intermediate residents who received lower ratings (3.98 ± 0.48) than senior residents (4.54 ± 0.42). Scores did not vary by ED treatment area (p > .05). Residents judged to be safe to independently run the shift had significantly higher mean scores than those judged not to be safe (4.74 ± 0.31 vs 3.75 ± 0.66; p < .001). Fourteen observations per resident, the typical number recorded during a 1-month rotation, were required to achieve a reliability of 0.80. Conclusion: The O-EDShOT successfully discriminated between junior, intermediate and senior-level residents regardless of ED treatment area. Multiple sources of evidence support the O-EDShOT producing valid scores for assessing a resident's ability to independently run an ED shift.

Innovation Concept: Assessment of residents' Point of Care Ultrasound (PoCUS) competency currently relies on heterogenous and unvalidated methods, such as the completion of a number of proctored studies. Although number of performed studies may be associated with ability, it is not necessarily a surrogate for competence. Our goal was to create a single Ultrasound Competency Assessment Tool (UCAT) using domain-anchored entrustment scoring. Methods: The UCAT was developed as an anchored global assessment score, building on a previously validated simulation-based assessment tool. It was designed to measure performance across the domains of Preparation, Image Acquisition, Image Optimization, and Clinical Integration, in addition to providing a final entrustment score (i.e., OSCORE). A modified Delphi method was used to establish national expert consensus on anchors for each domain. Three surveys were distributed to the CAEP Ultrasound Committee between July-November 2018. The first survey asked members to appraise and modify a list of anchor options created by the authors. Next, collated responses from the first survey were redistributed for a re-appraisal. Finally, anchors obtaining >65% approval from the second survey were condensed and redistributed for final consensus. Curriculum, Tool or Material: Twenty-two, 26, and 22 members responded to the surveys, respectively. Each anchor achieved >90% final agreement. The final anchors for the domains were: Preparation – positioning, initial settings, ensures clean transducer, probe selection, appropriate clinical indication; Image Acquisition – appropriate measurements, hand position, identifies landmarks, visualization of target, efficiency of probe motion, troubleshoots technical limitations; Image Optimization – centers area of interest, overall image quality, troubleshoots patient obstacles, optimizes settings; Clinical Integration – appropriate interpretation, understands limitations, utilizes information appropriately, performs multiple scans if needed, communicates findings, considers false positive and negative causes of findings. Conclusion: The UCAT is a novel assessment tool that has the potential to play a central role in the training and evaluation of residents. Our use of a modified Delphi method, involving key stakeholders in PoCUS education, ensures that the UCAT has a high degree of process and content validity. An important next step in determining its construct validity is to evaluate the use of the UCAT in a multi-centered examination setting.

Background: Greater than 80% of patient visits to emergency departments (EDs) are for a pain-related concerns. Approximately 38,000 patients per year have such complaints in our academic hospital ED. 3,300 (8.6%) of those visits are for musculoskeletal (MSK) pain (i.e. back or extremity injury/pain), which are typically triaged as low-acuity presentations, leading to longer times to clinician assessment. Delays to adequate analgesia result in unnecessary suffering, worse patient care and satisfaction, and increased patient complaints. Aim Statement: We aimed to reduce the time-to-analgesia (TTA; time from patient triage to receipt of analgesia) for patients with MSK pain in our ED by 55% (to under 60 minutes) in 9 months’ time (May 2018). Measures & Design: Our outcome measures were TTA (in minutes) and ED length of stay (LOS; in minutes). Process measures included nurses’ use of medical directive and rate of analgesia administration. Balancing measures included patient adverse events and time spent triaging for nurses. We utilized weekly data capture for the Statistical Process Control (SPC) chart, and we used Mann-Whitney U test for our before-and-after evaluation. Utilizing the Model for Improvement, we performed wide stakeholder engagement and root cause analyses, and we created a Pareto chart. This led to our Plan-Do-Study-Act (PDSA) cycles: 1) nurse-initiated analgesia (NIA) at triage; 2) new triage documentation aid for medication administration; 3) quick reference medical directive badge tag for nurses; 4) weekly targeted feedback of the project's progress at clinical team huddle. Evaluation/Results: TTA decrease from 129 minutes (n = 153) to 100 minutes (22.5%; n = 87, p < 0.05). ED LOS decreased from 580 minutes (n = 361) to 519 minutes (10.5%; n = 187; p = 0.77). Special cause variation was identified on the ED LOS SPC chart with eight consecutive points below the midline, after PDSA 1. The number of patients who received any analgesia increased from 42% (n = 361) to 47% (n = 187; p = 0.13). The number of patients who received medications via medical directives increased from 22% (n = 150) to 44% (n = 87; p < 0.001). Balancing measures were unchanged. Discussion/Impact: The significant reduction in the TTA and increase in the use of medical directives in the before-and-after analyses were likely due to our front-line focused improvements and deliberate nursing engagement. With continued success and sustainable processes, we are planning to spread our project to other EDs and broaden our initiative to all pain-related concerns.

Background: Choosing Wisely (CW) recommends patients under age 50 with uncomplicated, recurrent renal colic do not require CT scans. Despite this, CT use has risen dramatically in the past two decades, resulting in unnecessary radiation, cost and prolonged length of stay (LOS). Additionally, a common alternative – formal ultrasound (US) – is not always available. Returning for US can add 10 hours to LOS. We introduced a clinical management pathway (CMP) for low-risk patients with renal colic utilizing point-of-care ultrasound (POCUS) and evaluated its impact on emergency department (ED) CT rates and LOS. Aim Statement: By April 2019, we aim to reduce CT utilization by 50% and time from physician initial assessment (PIA) to discharge by 1 hour for patients under age 50 presenting to Sunnybrook ED with uncomplicated, recurrent renal colic. Measures & Design: The primary intervention was a CMP developed collaboratively with local urologists. The CMP uses POCUS to assess for hydronephrosis (HN) as a marker of nephrolithiasis. Patients with HN receive follow-up in urology clinic without confirmatory imaging. Patients without HN proceed to usual care. An Ishikawa diagram helped identify barriers to success. Subsequent PDSA cycles included the introduction of reference cards, POCUS workshops and online modules. Outcome measures were ED CT utilization and PIA to discharge times. Process measures were referrals to urology clinic and proportion of patients receiving XR, US and no imaging. Balancing measures were urology CT utilization, alternate diagnoses and return ED visits. Data was plotted on a run chart. Evaluation/Results: Data collection is ongoing and will conclude by April 2019. Interim data shows patients enrolled in the CMP have a reduction in mean PIA-to-discharge time of 173 minutes. Fidelity – specifically, the willingness of ED physicians to use POCUS compared to the ease of ordering CTs – is the biggest challenge to success. Discussion/Impact: This study addresses the feasibility of CW recommendations and utilizes POCUS as a tool for recurrent renal colic. Collaboration with Urology will provide insight into the CMP's sustainability and downstream impact. Reduction of unnecessary CTs will lead to improved patient safety and reduced costs. Decreased PIA-to-discharge times will reduce overcrowding, shorten wait times and improve access to imaging for other patients. Finally, this project may encourage use of POCUS for low-risk patients with renal colic.

Background: The Choosing Wisely campaign aims to reduce unnecessary testing. Over testing for urinary tract infections and concomitant overtreatment of asymptomatic bacteriuria is a target of this campaign, aiming to decrease healthcare costs and the risks of side effects such as Clostridium difficile infection, adverse reactions, and antimicrobial resistance. During the study baseline (2017), 95 urine cultures (UC) were sent for every 1000 ED visits (9.5%). Of these, fewer than 20% were positive. Aim Statement: The aim of this improvement initiative was to reduce UC testing in the ED, by 50%, from a baseline average of nearly 100 cultures per 1000 ED patients visits, to 50 cultures per 1000 visits, by May 31st, 2018. Measures & Design: This was an interrupted time series study, analyzed using Statistical Process Control (SPC) methodology. Root cause analysis was performed using an Ishikawa diagram. A Pareto chart was completed via multi-voting. A Driver Diagram was developed using the highest ranked items from the Pareto chart to identify locally relevant and feasible interventions. Interventions 1) Medical directives were modified; Routine paired sending of UC with urinalysis by nurses was removed. 2) Physician Education and implementation of a clinical decision aid (CDA); A CDA was created using PDSA methodology, using an iterative approach from development through implementation. Outcome measure: rate of Urine Cultures sent per 1000 ED patient visits Process measure: percent of positive cultures Balancing measures: rate of 14-day ED return visits and hospital admission for patients diagnosed with UTI/Urosepsis/Pyelonephritis. Evalution/Results: At the study's conclusion, there was a decrease in UC rate, from 95 per 1000 ED visits, to 59 per 1000 ED visits (RR 38%, AR 3.6%) There was evidence of special cause variation on the SPC chart. Positive cultures increased from 19% to 34%. There was no increase in the rate of ED 14-day return visits or hospital admission for patients with a diagnosis of UTI, urosepsis or pyelonephritis. Discussion/Impact: The study interventions of uncoupling routine sending of UA and UC, and physician education and use of a clinical decision aid, effectively decreased the rate of UC testing during the study period. A reduction in inappropriate UC testing is important to limit avoidable patient morbidity and reduce unnecessary health care spending. Further studies are indicated to target interventions on patient subgroups and to reduce unnecessary antibiotic prescriptions.

Background: Canadian Stroke Guidelines recommend that Transient Ischemic Attack (TIA) patients at highest risk of stroke recurrence should undergo immediate vascular imaging. Computed tomography angiography (CTA) of the head and neck is recommended over carotid doppler because it allows for enhanced visualization of the intracranial and posterior circulation vasculature. Imaging while patients are in the emergency department (ED) is optimal for high-risk patients because the risk of stroke recurrence is highest in the first 48 hours. Aim Statement: At our hospital, a designated stroke centre, less than 5% of TIA patients meet national recommendations by undergoing CTA in the ED. We sought to increase the rate of CTA in high risk ED TIA patients from less than 5% to at least 80% in 10 months. Measures & Design: We used a multi-faceted approach to improve our adherence to guidelines including: 1) education for staff ED physicians; 2) agreements between ED and radiology to facilitate rapid access to CTA; 3) agreements between ED and neurology for consultations regarding patients with abnormal CTA; and 4) the creation of an electronic decision support tool to guide ED physicians as to which patients require CTA. We measured the rate of CTA in high risk patients biweekly using retrospective chart review of patients referred to the TIA clinic from the ED on a biweekly basis. As a balancing measure, we also measured the rate of CTA in non-high risk patients. Evaluation/Results: Data collection is ongoing. An interim run chart at 19 weeks shows a complete shift above the median after implementation, with CTA rates between 70 and 100%. At the time of submission, we had no downward trends below 80%, showing sustained improvement. The CTA rate in non-high risk patients did also increase. Disucssion/Impact: After 19 weeks of our intervention, 112 (78.9%) of high risk TIA patients had a CTA, compared to 10 (9.8%) in the 19 weeks prior to our intervention. On average, 10-15% of high risk patients will have an identifiable lesion on CTA, leading to immediate change in management (at minimum, an inpatient consultation with neurology). Our multi-faceted approach could be replicated in any ED with the engagement and availability of the same multi-disciplinary team (ED, radiology, and neurology), access to CTA, and electronic orders.

Background: Massive transfusion protocols (MTP) are widely used to rapidly deliver blood products to bleeding trauma patients. Every minute delay in blood product administration in bleeding trauma patients is associated with a 5% increased odds of death. In-situ simulation (ISS) is simulation that takes place in the actual clinical work environment. We used ISS as a novel, prospective and iterative quality improvement (QI) approach to identify and improve MTP steps that impact time to blood delivery (TTBD) during actual trauma resuscitations. Aim Statement: To reduce the TTBD for bleeding trauma patients by 20% over a 12-month ISS-based QI initiative. Measures & Design: We conducted twelve high-fidelity, interprofessional ISS sessions at a Level-1 trauma center in Toronto, Canada. We used clinician video review as well as extensive stakeholder involvement, including with nurses, porters, blood bank and human factors experts, to develop Plan-Do-Study-Act (PDSA) cycles for MTP improvement. Our three major PDSA cycles revolved around: 1) decreasing MTP activation time; 2) reducing the unpredictable and inefficient transport times for the blood itself; and 3) improving the notification of blood product arrival in the trauma bay. Each PDSA cycle was iteratively tested with ISS prior to implementation into clinical care. Outcome measure was the mean TTBD for trauma patients requiring MTP (in minutes, standard deviation [SD]). Process measures included time to MTP activation and porter transport times. Balancing measures included stakeholder satisfaction. Evaluation/Results: Our baseline TTBD for MTP patients was 11.58min (n = 41, SD 6.8). There were 54 trauma patients that had MTP during the ISS-based QI initiative, and their mean TTBD was 10.44min (SD 6.1). The TTBD after the QI initiative was 9.12min, sustained over 1 year (n = 50, SD 5.3; 21.2% relative reduction, p < 0.05). A run chart did not show special cause variation chronologically related to our interventions. Patients in each group were similar in demographic data, trauma characteristics and injury severity score. Discussion/Impact: We achieved a 21.2% reduction in TTBD for trauma patients requiring MTP with an ISS-based QI initiative. ISS represents a novel approach to the identification and iterative testing of process improvements within trauma care. This methodology can and should be included in QI projects in order to safely test and improve processes of care before they impact real patients.

Background: Urinary tract infection (UTI) is a common diagnosis in children presenting to the Emergency Department (ED) and often leads to empiric antibiotic treatment prior to culture results. A recent study at our centre found that 47% of children diagnosed with a UTI and discharged on antibiotics had a negative urine culture. None of these patients were notified of the negative result or to discontinue antimicrobial treatment. Aim Statement: The aim of this study was to improve UTI diagnostic accuracy by 50% while promoting antimicrobial stewardship through timely antibiotic discontinuation and standardized antimicrobial treatment for uncomplicated UTIs over the next 12 months. Measures & Design: Three interventions were developed using plan-do-study-act (PDSA) cycles. In collaboration with the hospital's Choosing Wisely campaign and antimicrobial stewardship program, an evidence-based empiric UTI diagnostic algorithm was created to aid with diagnostic decision-making and reduce practice variation. A daily call-back system was also implemented for urine cultures where patients who had a negative urine culture were contacted to stop antibiotics. Lastly, a practice alert was integrated in the EMR as a reminder of appropriate antimicrobial prescription duration. The main outcome measures were the percentage of inappropriately diagnosed UTIs and percentage with timely antimicrobial discontinuation. Process measures included antibiotic days saved, treatment duration, and physician adherence to the algorithm. As a balancing measure, positive urine cultures were reviewed to assess accuracy of the algorithm to detect UTIs and potential harm from delayed UTI diagnoses. Evaluation/Results: Early results from the 530 children included in the analysis demonstrated a 14% reduction in inappropriate UTI diagnoses. With the initiation of the call-back system, the antibiotic days saved increased from 0 to 495 days. Call-backs for negative cultures increased from 0% to 68% of the time. Of those positive cultures with a missed UTI diagnosis, only 5 patients in 5 months had a return visit within 72 hours and none required admission. Discussion/Impact: Appropriate diagnosis and treatment of UTIs in our ED has improved with the implementation of a diagnostic algorithm. A larger impact is anticipated once the algorithm is embedded in the EMR as a form of decision support, but these changes take time to implement. Although labour intensive, the call-back system has greatly impacted the antimicrobial days saved and reduced risk for harm in this population.

Background: Emergency Department (ED) communication between patients and clinicians is fraught with challenges. A local survey of 65 ED patients revealed low patient satisfaction with ED communication and resultant patient anxiety. Aim Statement: To increase patient satisfaction with ED communication and to decrease patient anxiety related to lack of ED visit information (primary aims), and to decrease clinician-perceived patient interruptions (secondary aim), each by one point on a 5-point Likert scale over a six-month period. Measures & Design: We performed wide stakeholder engagement, surveyed patients and clinicians, and conducted a patient focus group. An inductive analysis followed by a yield-feasibility-effort grid led to three interventions, introduced through sequential and additive Plan-Do-Study-Act (PDSA) cycles. PDSA 1: clinician communication tool (Acknowledge-Empathize-Inform [AEI] tool), based on survey themes and a literature review, and introduced through a multi-modal education approach. PDSA 2: patient information pamphlets developed with stakeholder input. PDSA 3: new waiting room TV screen with various informational ED-specific videos. Measures were conducted through anonymous surveys: Primary aims towards the end of the patient ED stay, and the secondary aim at the end of the clinician shift. We used Statistical Process Control (SPC) charts with usual special cause variation rules. Two-tailed Mann-Whitney tests were used to assess for statistical significance between means (significance: p < 0.05). Evaluation/Results: Over five months, 232 patient and 104 clinician surveys were collected. Wait times, ED processes, timing of typical steps, and directions were reported as the most important communication gaps, they and were included in the interventions. Patient satisfaction improved from 3.28 (5 being best, all means; n = 65) to 4.15 (n = 59, p < 0.0001). Patient anxiety improved from 2.96 (1 being best; n = 65) to 2.31 (n = 59, p < 0.01). Clinician-perceived interruptions went from 4.33 (1 being best; n = 30) to 4.18 (n = 11, p = 0.98). SPC charts using Likert scales did not show special cause variation. Discussion/Impact: A sequential, additive approach undertaken with pragmatic and low-cost interventions based on both clinician and patient input led to increased patient satisfaction with communication and decreased patient anxiety due to lack of ED visit information after PDSA cycles. These approaches could easily be replicated in other EDs to improve the patient experience.

Background: Sepsis is a life-threatening syndrome, and delays to appropriate antibiotic therapy increases mortality. Order sets have shown decrease in time to antibiotics in pneumonia, and in sepsis, the implementation of order sets resulted in more intravenous fluids, appropriate initial antibiotics and lower mortality. Aim Statement: The goal was to create an order set for an approach to septic patients, to improve sepsis management. We sought to improve time from triage to first antibiotics, by 15 minutes, for Emergency Department (ED) patients with sepsis in three months after implementation compared to three months before. Measures & Design: We used a literature review, as well as comparison to existing order sets at other EDs to design our initial order set. We underwent multiple revisions based on stakeholder feedback. We educated physician and nursing teams about the order sets, although use was ultimately at physician discretion. We implemented the order set on April 9, 2017. After three months, an electronic retrospective chart review identified patients with a final sepsis diagnosis admitted to the critical care unit. For each patient, we captured triage time using the electronic record, and time to antibiotics from when the antibiotic was taken out of the medication cart. Finally, utilization of order sets was checked via manual chart audit. Evaluation/Results: A run chart did not demonstrate any shifts or trends suggesting a change after implementation. Median time to antibiotics in minutes, 3 months prior (n = 45) and post (n = 55) intervention, increased from 245 to 340 minutes, although the range was very large. Chart audits demonstrated clinicians were not using the order sets. There was 10% usage for 2 of the months and 0% usage the other month, post-intervention. Disucssion/Impact: There was insufficient uptake of the Sepsis Order Set by the Sunnybrook ED to result in any impact on time to antibiotics. Order sets require more than just implementation to be effective. Difficulties in implementation were due to the document not being readily available to physicians. To mediate, we have organized nursing staff to attach the order set onto charts based on triage assessment and will re-assess with another PDSA cycle after this intervention.

Introduction: Substance use is prevalent in Canada yet treatment for alcohol use disorder (AUD) and opioid use disorder (OUD) is often inaccessible. Consequently, alcohol and opioid-related diagnoses such as intoxication, withdrawal, and overdose are a major reason for frequent emergency department (ED) visits. The Rapid Access to Addiction Medicine (RAAM) Clinic opened at the University Health Network (UHN) in January 2018 as part of a larger network of clinics in Toronto, and provides rapid, low barrier access to medical treatment for substance use disorder (SUD). Patients attended via self-referral, peer-referral, or referral by the ED, primary care, internal medicine or withdrawal management services. This study describes the demographic profile and short-term outcomes for patients attending a new RAAM clinic in its first 26 weeks of operation, including substance use and treatment retention for AUD and OUD. Methods: We reviewed the electronic medical record at the clinic over its first 26 weeks of operation. We assessed SUD diagnoses, referral source, prescribed medications, self-reported outcomes and retention rates. We calculated descriptive statistics using proportions for categorical variables and means with standard error for continuous variables. A student's t-test was used for all statistical analyses using Microsoft Excel. We reviewed the electronic medical record at the clinic over its first 26 weeks of operation. We assessed SUD diagnoses, referral source, prescribed medications, self-reported outcomes and retention rates. We calculated descriptive statistics using proportions for categorical variables and means with standard error for continuous variables. A student's t-test was used for all statistical analyses using Microsoft Excel. Results: The clinic saw 64 unique patients: 66% had an AUD, 39% had an OUD and 20% had a stimulant use disorder. 55% of patients were referred from outpatient care providers, 30% from the emergency department and 11% from withdrawal management services. 42% remained ongoing patients, 23% were discharged to other care and 34% were lost to follow-up. Gabapentin (38%), naltrexone (33%), and acamprosate (20%) were most frequently prescribed for AUD. Patients with AUD reported a significant decrease (p < 0.05) in alcohol consumption at their most recent visit compared to their initial visit. Most patients (78%) with OUD were prescribed buprenorphine, and most (89%) patients with OUD on buprenorphine had a negative urine screen at their most recent visit. Conclusion: A new RAAM outpatient clinic demonstrates the early success of a low-barrier addictions model in addressing unmet needs in substance use treatment. We see a reduction in both alcohol consumption and opioid use, and increased access to evidence-based pharmacotherapy for SUDs.

Introduction: Emergency hospital admissions are a growing concern for patients and health systems, globally. The objective of this study was to systematically review the evidence for diagnostic, medical, and surgical interventions that reduce emergency hospital admissions. Methods: We conducted a systematic review of systematic reviews by searching MEDLINE, PubMED, the Cochrane Database of Systematic Reviews, Google Scholar, and grey literature. Systematic reviews of any diagnostic, surgical, or medical interventions examining the effect on emergency hospital admissions among adults were included. The quality of reviews was assessed using AMSTAR and the quality of evidence was assessed using GRADE. The subsequent analysis was restricted to interventions with moderate or high-quality evidence only. Results: 13 051 titles and abstracts and 1 791 full-text articles were screened from which 42 systematic reviews were included. The reviews included an underlying evidence base of 215 randomized controlled trials with 135 282 patients. Of 20 unique diagnostic, medical, and surgical interventions identified, four had moderate (n = 4) or high (n = 0) quality evidence for significant reductions in hospital admissions in five patient populations. These were: cardiac resynchronization therapy for heart failure and atrial fibrillation, percutaneous aspiration for pneumothorax, early/routine coronary angiography for acute coronary syndrome (alone or comorbid with chronic kidney disease), and natriuretic peptide guided therapy for heart failure. Conclusion: We identified four interventions across five populations that when optimized, may lead to reductions in emergency hospital admissions. These finding can therefore help guide the development of quality indicators, standards, or practice guidelines.