Introduction: Acute pharyngitis is a common reason for primary care or emergency department visits, often resulting in antibiotic prescription. Rapid antigen detection tests (RADT) are routinely used to diagnose Group A Streptococcus (GAS) pharyngitis. However, due to its low sensitivity, patient pressures and conflicting guidelines, the RADT often complicates management decisions. Our aim was to assess the impact of RADT in patients presenting with acute GAS pharyngitis on the antibiotic prescription rate and appropriateness of antibiotic management. Methods: We systematically searched Medline, Embase, and Cochrane databases from 1980 to June 2016. Studies were selected according to a predefined PRISMA protocol and data extracted by two independent reviewers. Prospective and retrospective studies that evaluated the impact of RADT on antibiotic prescription for pharyngitis were included. Study quality was assessed using Cochrane Risk of Bias Tool and the Newcastle-Ottawa Scale. Our main outcome was a dichotomous measure of antibiotic prescription, with or without RADT availability. Studies were combined if there was low clinical and statistical heterogeneity (I2<30%). Bivariate Mantel-Haenszel random effects model was used to perform meta analyses using SPSS 22 and Revman 5. Results: We identified 4003 studies: 139 were selected for full text review; 10 met our inclusion criteria (N=10859 participants, median age 31 years, 56.7% female). Mean antibiotic prescription rate in the RADT and control arm was 38.2% (SD 15.6) and 55.9% (SD 16.3), respectively. The use of RADT was associated with lower antibiotic prescription rate in both adults (OR=0.60 [95% CI 0.45-0.80], I2=8%, N=1407) and pediatrics (OR=0.49 [95% CI 0.44-0.55], I2=5%, N=976). There was no overall difference (p<0.3) in antibiotic prescription rate among disease severity (Centor scores 1-4). The use of RADT did not significantly impact the appropriateness of antibiotic management (OR=1.15 95% CI 0.94-1.5). Conclusion: The use of RADT is associated with a reduction in antibiotic prescription for patients with GAS pharyngitis without an increase in appropriate antibiotic use. Despite low prevalence of the disease in the population, antibiotic prescription rates are still high. These findings suggest great potential for antibiotic stewardship and reevaluation of current guidelines for managing GAS pharyngitis.

Introduction: Sepsis protocols call for the acquisition of blood cultures in septic emergency department (ED) patients.However, the criteria for blood cultures are vague, they are costly, only positive 8-12% of the time, with up to half of these being false positives. The objective of this study was to establish if positive blood cultures could be excluded in low-risk sepsis patients with levels of CRP below 20 ml/L. Methods: This was a multicenter prospective cohort study of 765 ED patients at St Paul’s and Mount St Joseph’s hospitals in Vancouver with sepsis (2 or more SIRS criteria and infection) and none of: immuncompromised, injection drug use, indwelling vascular device or septic shock (SBP<90 mmhg). Consecutive patients with sepsis had CRP and blood cultures obtained at the same time.OUTCOMES. True positive blood cultures, false positive blood cultures, positive blood cultures that changed patient management. True and false positive blood cultures were based on Infectious Disease Society of America Guidelines, and change in management was defined as change in type or length of antibiotic therapy and was blindly adjudicated by a medical microbiologist. Results: 765 ED patients with sepsis met inclusion criteria. Mean age was 48.3 years and 57% were male. Blood cultures were positive in 99/765 (12.9%) subjects, of which 19 were false positive (19.2%). CRP was >20 mg/L in 595/765 (77.8%) of patients. Of 170 subjects with a CRP<20 mg/L, 3 had a positive blood culture (1.8%; 95% CI 0.1%- 5%). Management was not changed in any patient with a positive blood culture and CRP level<20 mg/L. Of 19 subjects with a false positive blood culture, CRP was <20 mg/L for 6 (31.6%). Conclusion: In this cohort of low-risk sepsis patients, based on a CRP of <20 mg/L, acquisition of blood cultures could be safely avoided in 22.2% of patients, at significant savings to the health care system.

Introduction: Nightshifts may represent a more challenging work environment due to staff fatigue. Our objective was to determine if an association exists between health outcomes for patients seen in Calgary Zone Emergency Department (ED) during nightshifts as compared to other time periods. Methods: Administrative data from a city-wide electronic health record was collected from four urban EDs on all discharged patients during a 2-year period: January 2015-December 2016. A total of 454,125 patient visits were included and patients with a scheduled return to the ED were excluded. Three primary outcomes were selected to assess the effects of night shifts on the quality of care received by patients in the ED at night; (i) unscheduled returns to the ED within 7 days resulting in admission, (ii) mortality within 48 hrs and, (iii) mortality within 7 days of being seen by a physician. Non-night shifts were defined as patients seen on day and evening or 700-2300. The data was analyzed using descriptive statistics and precision reported via 95% confidence intervals. Results: For the outcome of returns resulting in admission, a 2.6% rate was noted for patients seen at night compared to 2.3% during non-night; OR 1.15 (95% CI 1.09-1.21). Furthermore, patients seen at night had a 0.033% rate of death, while non-night patients had a 0.022% chance of death within 48 hrs of discharge; OR 1.53 (95% CI 0.98-2.38). For mortality within 7 days, the rate of death observed was 0.10% and 0.078% respectively; OR 1.24 (95% CI 0.97-1.60). Conclusion: Our study identified presenting to the ED at night as a potential risk factor for adverse patient outcomes using 3 primary quality of care indicators. An adjusted analysis is needed to account for potential confounding variables and effect modifiers and is underway.

Introduction: Highly frequent users (HFU) of the emergency department (ED) remain a poorly defined and complex population. This study describes patient and visit characteristics for HFU of the ED, and analyzes subgroups of patients with mental illness, substance abuse, and/or ≥30 yearly ED visits. Methods: We performed a health records review of 250 randomly selected adults with >99th percentile of ED visit frequency (≥7 visits) at a tertiary care academic hospital with two EDs in 2014. Two reviewers collected demographic variables (age, sex, and comorbidities) and visit data (ED diagnosis, ED length of stay (LOS), ED presentation time (daytime 0800-1559 h, evening 1600-2359 h, overnight 2400-0759 h), consultation services, and final disposition). Data were analyzed using descriptive and univariate analyses, student t and Mann Whitney U tests. Results: Of 897 eligible patients who experienced 9,376 ED visits we included 250 patients (2,670 visits) in our main analyses, and an additional 11 patients (494 visits) outside of the random selection with ≥30 ED visits. Mean age was 53.4±1.3 (SEM), and 55.6% were female. Most patients had a fixed address (88.9%), and a family physician (87.2%). Top comorbidities included gastrointestinal (61.6%), cardiovascular (52%), and chronic pain issues (47.2%). Top ED diagnoses included musculoskeletal pain (9.6%), abdominal pain (8.4%) and alcohol-related presentations (8.5%). Hospital admission was required for 15.6% of visits. From all possible visits (3164 visits), consultations for social workers, geriatric emergency medicine nurses, or Community Care Access Centres were made for 5.9% of visits, with 47.3% of these patients presenting during daytime hours. Among visits requiring these consultations, median ED LOS was greatest in the evening (12.7 hours, range 1.4-45.2 hours), compared to daytime (5.4, 1.2-33.6; p=0.0002) or overnight (7.9, 1.0-38.3, p=0.02). Inter-rater review of 4.5% of abstracted health records revealed a kappa score of 0.8. Conclusion: This study highlights that a remarkably low proportion of HFUs received allied health consultations at the study sites, likely corresponding to a lack of available consultants outside of daytime work hours. Our findings suggest the need to address significant gaps in order to balance the clinical needs of patients who frequent the ED with currently available resources.

Introduction: Headaches are a common emergency department (ED) presentation. The objective of this study was to characterize headache presentations in Alberta over a five-year period and explore the proportion of patients with potentially severe pathology. Methods: Administrative health data for Alberta (years 2011-2015) were obtained from the National Ambulatory Care Reporting System (NACRS) for all adult (>17 years) headache presentations (ICD-10-CA: G43, G44, R51). Patients with a primary or secondary diagnosis code of headache were eligible for inclusion in the study. Exclusions were made using the following criteria: 1) sites without computed tomography (CT) scanners; 2) presentations with a Canadian Triage and Acuity Scale (CTAS) score of 1; 3) patients with trauma or external mechanism of injury (e.g., ICD-10-CA codes S,T,V,W,X,Y); and 4) presentations receiving an enhanced/contrast CT (head). NACRS data were linked with a provincial diagnostic imaging data. Data are reported as means and standard deviation (SD), medians and interquartile range (IQR) or proportions, as appropriate. Results: From 2011-2015, 98,333 presentations were made by 66,970 patients (~0.3 presentations per patient per year; equivalent to one presentation every 3.4 years). Headache presentations increased from 15,643 in 2011 to 21,636 in 2015. The median age was 38 years (IQR: 29, 51 years); more patients were female (69.3%), had a CTAS score of 3 (55%) and arrived at the ED without ambulance (90.3%). The majority of patients had a primary ED diagnosis of headache (88%) and the most common co-diagnosis was benign hypertension (2.8%). Additional diagnoses indicating severe or pathological headaches, included: stroke (0.63%), subarachnoid hemorrhage (0.43%), infection (i.e., meningitis) (0.11%), and other brain hemorrhages (0.08%). Overall, the ED management of approximately 25% of presentations involved a head CT. Most patients were discharged from the ED (89.4%) after a median length of stay of 3.5 hours (IQR: 2.1, 5.2 hours). Conclusion: Headache-related ED presentations are increasing in Alberta, yet few severe/pathological diagnoses are being identified. Efforts to ensure appropriateness of head CT ordering could reduce exposure to ionizing radiation, improve patient flow and reduce health care costs; this imaging represents a target for future interventions.

Introduction: Medication discrepancies are unintended differences between a patient’s outpatient and inpatient medication regimens, and occur in up to 60% of hospital admissions. Canadian emergency departments (EDs) have implemented medication reconciliation forms that are pre-populated with outpatient medication dispensing data in order to reduce medication discrepancies and resultant adverse drug events. However, these forms may introduce errors of commission by prompting prescribers to reorder discontinued or potentially harmful medications. Our objective was to evaluate the incidence of medication discrepancies and errors of commission after the implementation of pre-populated medication reconciliation forms. Methods: This chart review included admitted patients who were enrolled in a parent study in which a research pharmacist prospectively collected best-possible medication histories (BPMHs) in the ED using all available information sources. Following discharge, research assistants uninvolved with the parent study compared medication orders documented within 48 h of admission with the BPMH to identify medication discrepancies and errors of commission. Errors of commission were defined as inappropriate continuations of medications and reordering discontinued medications. An independent panel adjudicated the clinical significance of the errors. We used regression methods to identify factors associated with errors. The sample size was limited by enrolment into the parent study. Results: Of 151 patients, 71 (47%; 95%CI 39.2-54.9) were exposed to 112 medication errors. Of these errors, 75.9% (85/112; 95%CI 67.1-82.9) were discrepancies, of which 18.8% (16/85; 95%CI 12.0-28.4) were clinically significant. Errors of commission made up 24.1% (27/112; 95%CI 17.3-32.8) of all errors, of which 37.0% (10/27; 95%CI 18.8-55.2) were clinically significant. Taking 8 or more medications was associated with a 5-fold greater odds of experiencing a medication error after controlling for confounders (OR 5.00; 95%CI 2.45-10.17; p<0.001). Conclusion: Clinically significant medication discrepancies and errors of commission remain common despite the implementation of electronically pre-populated medication reconciliation forms. Prospective studies are needed to evaluate whether using pre-populated medication reconciliation forms increases the risk of introducing errors of commission.

Introduction: Almost every domain of quality is reduced in crowded emergency departments (ED), with significant challenges around the definition, measurement and interventions for ED crowding. We wished to determine if a combination of 3 easily measurable variables could perform as well as standard tools (NEDOCS score and a NEDOCS-derived LOCAL tool) in predicting ED crowding at a tertiary hospital with 57,000 visits per year. Methods: Over a 2-week period, we recorded ED crowding predictor variables and calculated NEDOCS and LOCAL scores. These were compared every 2 hours to a reference standard Physician Visual Analog Scale (range 0 to 10) impression of crowding to determine if any combination of variables outperformed NEDOCS and LOCAL (crowded=5 or greater). Five numeric variables performed well under univariate analysis: i) Total ED Patients; ii) Patients in ED beds + Waiting Room; iii) Boarded Patients; iv) Waiting Room Patients; v) Patients in beds To Be Seen. These underwent multivariate, log regression with stratification and bootstrapping to account for incomplete data and seasonal and daily effect. Results: 143 out of a possible 168 observations were completed. Two different combinations of 3 variables outperformed NEDOCS and LOCAL. The most powerful combination was: Boarded Patients; plus Waiting Room Patients; plus Patients in beds To Be Seen, with Sensitivity 81% and Specificity 76% (r=0.844, β=0.712, p<0.0001, strong positive correlation). This compared favourably with NEDOCS and LOCAL, each with Sensitivity 71% and Specificity 64%[PA1] (r=0.545 and r=0.640 respectively). We will also present a sensitivity and specificity analysis of all combinations of predictor variables, using various reference standard cut-offs for crowding. Conclusion: A combination of 3 easily measurable ED variables (Boarded Patients; plus Waiting Room Patients; plus Patients in beds To Be Seen) performed better than the validated NEDOCS tool and a NEDOCS-derived LOCAL score at predicting ED crowding. Work is on going to design a simple tool that can predict crowding in real time and facilitate early interventions. Correlation with ED system and clinical outcomes should be studied in different ED environments.

Introduction: Patients with advanced malignant and non-malignant disease (advanced disease—AD) who do not want or benefit from aggressive resuscitation may unfortunately receive such treatments if unable to communicate in an emergency. Timely access to patients’ resuscitation wishes is imperative for treating physicians and for medical information systems. Our aim was to determine what proportion of emergency department (ED) patients with AD have accurate, readily accessible resuscitation status documentation. Methods: This cross-sectional, prospective study was conducted at a tertiary care ED during purposefully sampled random accrual times in summer 2016. We enrolled all patients with: 1) palliative care consultation, 2) metastatic malignancy, 3) COPD or CHF on home oxygen, 4) hemodialysis, or 5) advanced neurodegenerative disease/dementia. The primary outcome was the retrieval of any existing resuscitation status documents. Documentation was obtained from a standardized review of forms accompanying the patient (“arrival documents”) and electronic medical record (“EMR”). We measured the time to retrieve this documentation, and interviewed consenting patients to corroborate documentation with their current wishes. Results: Of 85 enrolled patients, only 33 (39%) had any documentation of resuscitation status: 28 (33%) had goals of care retrieved from the hospital EMR, and 11 (15%) from arrival documents (some had both). Patients from long-term care facilities were more likely to have documentation available (odds ratio 13 [95% CI 2.5-65] vs community-living). Of 32 patients who were able to be interviewed, 20 (63%) expressed “do not resuscitate” wishes. Ten of these 20 lacked any documents to support their expressed resuscitation wishes. Previously expressed resuscitation wishes took more than 5 minutes to be retrieved in 3 cases when not filed “one click deep” in our EMR. Conclusion: The majority of patients with AD, including half of those who would not wish resuscitation from cardiorespiratory arrest, did not have goals of care documents readily available upon arrival to the ED. Patients living in the community with AD appear to be at high risk for unwanted resuscitative treatments should they present to hospital in extremis. Having documentation of their goals of care that is easily retrievable from the EMR shows promise, though issues of retrieval, accuracy, and validity remain important considerations.

Introduction: Previous studies have shown a link between Emergency Department (ED) overcrowding and worse clinical outcomes, increased risk of in-hospital mortality, higher costs, and longer times to treatment. Prolonged ED Length of Stay (LoS) of admitted patients awaiting a bed on in-patient units has been identified as a major driver of ED overcrowding. The purpose of this study is to provide a descriptive analysis of ED LoS among admitted patients, and determine the impact of prolonged ED LoS on total hospital in-patient length of stay (IP LoS). Methods: We conducted a single-site retrospective study for the period between January 1-December 31, 2015 at a very high volume community hospital. All patients aged ≥18 years admitted from the ED to acute in-patient Medicine units were identified. We carried out overall descriptive analysis (including analysis of day-of-the-week variability) on ED LoS. The mean total IP LoS for those patients with ED LoS<12 hours, 12-24 hours, and ≥24 hours were calculated and analyzed using ANOVA and Tukey HSD tests. Results: A total of 6,961 individuals were admitted to the medical units over the 12-month period. The median and mean ED LoS for admitted patients were 22.9 hrs (IQR: 13.9 hrs- 33.1 hrs) and 25.6 hrs respectively. Using ANOVA, there was a statistically significant difference in means of ED LoS as a function of the day of the week (p<0.0001), with Mondays having the highest mean ED LoS (27.6 hrs), and Fridays having the lowest (23.1 hrs). The mean IP LoS for those with ED LoS<12 hours, 12-24 hours, and ≥24 hours, were 6.8 days, 6.9 days, and 8.5 days respectively, with a statistically significant difference between group means (p<0.0001). Multiple pairwise comparisons of group means showed a statistically significant (p<0.05) difference between mean IP LOS of those with an EDLOS≥24 hours and those with an EDLOS<24 hours. Conclusion: Preliminary results indicate that ED LoS≥24 hours among admitted patients was associated with an increase in total IP LoS.*In the next 1-2 months, we intend to explore the role of other independent variables (age, sex, comorbidity, isolation status, and telemetry) on total ED LoS, and its association with IP LoS.

Introduction: Improving emergency department productivity has been a priority across Canada. In the United States, medical scribes have been utilized to increase the number of patients seen per hour (PPH) per physician; however, it is not well known if these outcomes can be translated to Canada. The purpose of this pilot evaluation was to (a) establish proof-of-concept of medical scribes in Canada and (b) gain experience in scribe implementation so as to inform future directions for the use of scribes in Canada. It was hypothesized that use of medical scribes would result in a greater PPH per physician. Methods: We conducted a four-month pilot evaluation of medical scribes in the emergency department of the Queensway-Carleton Hospital in Ottawa, Ontario. Eleven scribes were utilized in the study ranging in age from 18 to 23 years old. Following scribe training and an initial two-month acclimation period for both scribes and physicians, data collection began January 2015. Twenty-two full or part time emergency physicians were followed in this study, who received shifts with and without a scribe over the next four months. Physician work hours as well as the number of patients seen by each physician on each shift was documented. From these metrics, PPH per physician was calculated for each shift. Across the four months, the average PPH was determined for each physician during shifts with a scribe and shifts without a scribe. Two-tailed paired-samples t-tests (α=0.05) were used to compare mean (SD) PPH within physicians based on presence or absence of a scribe. Results: A total of 463 physician hours were documented without use of a scribe and 693.75 physician hours were documented with use of a scribe. Across all 22 physicians in the study, 18 (81.8%) demonstrated a greater PPH with use of a scribe. Overall, PPH per physician was significantly greater (12.9%) during shifts with a scribe (mean 2.81, SD 0.78) compared to shifts without a scribe (mean 2.49, SD 0.60) (p=0.006). Sensitivity analyses revealed that PPH per physician during shifts without a scribe during the study period were similar to the year prior, before scribes were introduced to the hospital (p=0.315). Conclusion: Use of medical scribes resulted in an increased PPH per physician in our hospital. While these results were from an evaluation at a single centre, they support broader implementation and evaluation of scribes in more centres across Canada.

Introduction: Burnout rates for emergency physicians (EP) continue to be amongst the highest in medicine. One of the commonly cited sources of stress contributing to disillusionment is bureaucratic tasks that distract EPs from direct patient care in the emergency department (ED). The novel position of Physician Navigator was created to help EPs decrease their non-clinical workload during shifts, and improve productivity. Physician Navigators are non-licensed healthcare team members that assist in activities which are often clerical in nature, but directly impact patient care. This program was implemented at no net-cost to the hospital or healthcare system. Methods: In this retrospective study, 6845 clinical shifts worked by 20 EPs over 39 months from January 1, 2012 to March 31, 2015 were evaluated. The program was implemented on April 1, 2013. The primary objective was to quantify the effect of Physician Navigators on measures of EP productivity: patient seen per hour (Pt/hr), and turn-around-time (TAT) to discharge. Secondary objectives included examining the impact of Physician Navigators on measures of ED throughput for non-resuscitative patients: emergency department length of stay (LOS), physician-initial-assessment times (PIA), and left-without-being-seen rates (LWBS). A mixed linear model was used to evaluate changes in productivity measures between shifts with and without Physician Navigators in a clustered design, by EP. Autoregressive modelling was performed to compare ED throughput metrics before and after the implementation of Physician Navigators for non-resuscitative patients. Results: Across 20 EPs, 2469 shifts before, and 4376 shifts after April 1, 2013 were analyzed. Daily patient volumes increased 8.7% during the period with Physician Navigators. For the EPs who used Physician Navigators, Pt/hr increased by 1.07 patients per hour (0.98 to 1.16, p<0.001), and TAT to discharge decreased by 10.6 minutes (-13.2 to -8.0, p<0.001). After the implementation of the Physician Navigators, overall LOS for non-resuscitative patients decreased by 2.6 minutes (1.0%, p=0.007), and average PIA decreased by 7.4 minutes (12.0%, p<0.001). LBWS rates decreased by 43.9% (0.50% of daily patient volume, p<0.001). Conclusion: The use of a Physician Navigator was associated with increased EP productivity as measured by Pt/hr, and TAT to discharge, and reductions in ED throughput metrics for non-resuscitative patients.

Introduction: Cognitive dysfunction is getting more common in geriatric emergency department (ED) patients, as the number of seniors visiting our EDs is increasing. ED guidelines recommend a systematic mental status screening for seniors presenting to the ED. As the existing tools are not suitable for the busy ED environment, we need quicker and easier ways to assess altered mental status, such as the O3DY. The purpose of this study is to assess the effectiveness of the French version of the O3DY to screen for cognitive dysfunction in seniors presenting to the ED. Methods: This is a planned sub-study of the INDEED project, which was conducted between February and May 2016 in 4 hospitals across the province of Québec. Inclusion criteria were: patients aged ≥65, with an 8-hour ED stay, admitted on a care unit, independent or semi-independent in their activities of daily living. Exclusion criteria were: patient living in a long-term nursing facility, with an unstable medical condition, pre-existing psychiatric condition or severe dementia, a delirium within the 8-hour exposure to the ED. A trained research assistant collected the following data upon initial interview: socio-demographic information, cognitive assessment (TICS-m), functional assessment (OARS) and delirium screening (CAM). The O3DY was also administered at initial interview and during patient follow-ups, as well as the CAM. Results: This study population was composed of 305 participants, of which 47.7% were men. Mean age was 76 years old (SD: 10.8). Nine of these participants had a previous history of dementia. 151 of these participants (47,04%) had a negative O3DY and 154 (47,98%) a positive O3DY at the initial encounter. When compared to the CAM, the O3DY presents a sensitivity of 85.0% (95% CI [62.1, 96.8]) and a specificity of 57.7% (95% CI [51.8, 63.6]) for prevalent delirium. When compared to the TICS, the O3DY presents a sensitivity of 76.7% (95% CI [66.4, 85.2]) and a specificity of 68.1% (95% CI [61.3, 74.3]) for cognitive impairment. The combined measure presents a sensitivity of 76.7% (95% CI [66.6, 84.9]) and a specificity of 68.4% (95% CI [61.7, 74.5]). Conclusion: A negative result to the O3DY indicates the absence of prevalent delirium or undetected cognitive impairment. The O3DY could be a useful tool for the triage nurses in the ED.

Introduction: Hip fractures affect over 35,000 Canadians each year. Delirium, or acute confusion, occurs in up to 62% of patients following a hip fracture. Delirium substantially increases hospital length of stay and doubles the risk of nursing home admissions and death. The primary objective of this study was to identify risk factors independently associated with acute in-hospital delirium within 72 hours of emergency department (ED) arrival for patients diagnosed with a hip fracture. Methods: This was a retrospective chart review of patients aged 65 years and older presenting to one of two academic EDs with a discharge diagnosis of hip fracture from January 1st 2014 to December 31st 2015. Multivariable logistic regression analysis was used to determine variables independently associated with the development of acute in-hospital delirium within 72 hours of ED arrival. Results: Of the 668 included patients, mean (SD) age was 84.1 (8.0) years and 501 (75%) were female. 521 (78.0%) patients received an opioid analgesic and/or femoral nerve block in the ED. The most common analgesics used in the ED were intravenous (IV) morphine (35.8%), IV hydromorphone (35.2%), or dual therapy with both IV hydromorphone and IV morphine (2.2%). Femoral nerve blocks were initiated for 36 (5.4%) patients and successfully completed in 35 (5.2%) patients in the ED. 181 (27.1%) patients developed delirium within 72 hours of ED arrival. History of neurodegenerative disease or dementia (OR: 5.7, 95% CI: 3.9, 8.4), age >75 (OR: 2.8, 95% CI: 1.4, 5.6) and absence of analgesia in the ED (OR: 2.1, 95% CI: 1.3, 3.2) were independently associated with acute in-hospital delirium. Conclusion: The development of in-hospital delirium is common in patients diagnosed with a hip fracture. We have identified modifiable and non-modifiable risk factors independently associated with acute in-hospital delirium, which can be identified in the ED. Clinicians should be aware of these risk factors in order to implement strategies directed at reducing the development of acute delirium. Additionally, further research is needed in order to understand the relationship between analgesia delivered in the ED and the development of delirium for patients diagnosed with a hip fracture.

Introduction: We examined persons transported to hospital after police use of force to determine whether Emergency Department (ED) assessment and/or mode of transport could be predicted. Methods: A multi-site prospective consecutive cohort study of police use of force with data on ED assessment for individuals ≥18 yrs was conducted over 36 months (Jan 2010-Dec 2012) in 4 cities in Canada. Police, EMS and hospital data were linked by study ID. Stepwise logistic regression examined the relationship between the police call for service and subject characteristics on subsequent ED assessment and mode of transport. Results: In 3310 use of force events, 86.7% of subjects were male, median age 29 yrs. ED transport occurred in 26% (n=726). Odds of ED assessment increased by 1.2 (CI 1.1, 1.3) for each force modality >1. Other predictors of ED use: if the nature of police call was for Mental Health Act (MHA) (Odds 14.3, CI 10.6, 19.2), features of excited delirium (ExD) (Odds 2.7, CI 1.9, 3.7), police-assessed emotional distress (EDP) not an MHA (Odds 2.1, CI 1.5, 3.0) and combined drugs, alcohol and EDP (Odds 1.7, CI 1.9, 3.7). Those with alcohol impairment alone were less likely to go to ED from the scene: OR 0.6 (CI 0.5, 0.7). EMS transported 55% of all patients (n=401), although police transported ~100 people who EMS attended at the scene but did not subsequently transport. For patients brought to the ED, 70% had a retrievable chart (512/726) with a discernible primary diagnosis: 25% for physical injury, 32% for psychiatric and 43% for drug and/or alcohol intoxication. For use of force events that began as MHA calls, patient transport was more often by police car than ambulance OR 1.8 (CI 1.2, 2.5), while those with drug intoxication or ≥3 ExD features were more often brought by ambulance: odds of police transport 0.5 (CI 0.3, 0.9) and 0.4 (CI 0.3, 0.7). Violence or aggression did not predict mode of transport in our study. Conclusion: About one quarter of police use of force events lead to ED assessment; 1 in 4 patients transported had a physical injury of some description. Calls including the Mental Health Act or individuals with drug intoxication or excited delirium features are most predictive of ED use following police use of force. In MHA calls with use of force, persons are nearly twice as likely to go to ED by police car than by ambulance.

Introduction: Screening for depression in the emergency department (ED) has been recommended for the last two decades. It is estimated that 1 in 5 adults presenting to the ED meet the criteria for depression, making this setting an ideal point of care for proper and early referral to general practitioners and/or specialist mental health services. One of the barriers to assessment of depression in the ED is a lack of validated tools to screen for depression in this context of use. The purpose of this study is to test the extent to which the commonly used Patient Health Questionnaire (PHQ-9) is valid and reliable to screen for depression in adults presenting to the ED. Methods: Adults, aged 19 years and over, presenting to an inner-city, academic ED with an acute mental health complaint (AMHC) completed a questionnaire package that included demographic questions, the PHQ-9, and 5 other questionnaires for validation purposes. Traditional and Rasch Measurement (RM) methods were applied to the data to examine how well the items: captured the 95% range (±2 logits) of the concept of interest, were reliable and valid, and met the criteria for unidimensional and invariant measurement. Results: Preliminary prospective data from 108/200 adults (mean age 39.7±13.6 years; 65% male) completed the questionnaire package. A total of 58.9% of the sample met the criteria for moderate-severe depression (PHQ-9≥15), with 37% reporting thoughts of suicide and/or self-harm nearly every day for the past two weeks. Analysis of these items showed good overall fit to the Rasch model (χ2=28.3, df=18, p=.06), good reliability (rp=0.84), an ordered 4-point response scale structure, excellent individual item fit, and no item bias for gender, age, level of education, or employment status. Items covered between -1.45 to 1.52 logits, spanning 74% of the targeted theoretical continuum, with gaps at each end of the range. Item #3 (trouble falling or staying asleep) was the easiest item (indicating lower depression) and Items #8 and #9 (moving slowly and thoughts of harm/suicide) were the more difficult items (indicating more severe depression). Conclusion: This study supports the PHQ-9 as a reliable and valid screen for depression in the ED. Incorporating standardized and uniform assessment in Canadian EDs will begin the process of advancing the role of the ED to initiate evidence-based care to optimize the outcomes of Canadians with an AMHC.

Introduction: The risk of recurrent stroke following a transient ischemic attack (TIA) has been estimated to be as much as 5 percent in the first 48 hours and ten percent in the first week following initial TIA symptoms, but can be modified as a result of intensive risk factor management. Care pathways for these patients vary between different regions within Alberta with Edmonton admitting more TIA patients and Calgary using computed tomography angiography (CTA) based triage. To examine regional differences in the quality of care, the rate of admission for stroke within 90 days of an index ED visit for TIA/minor stroke was investigated. Methods: Data analysts from the Data Integration, Measurement and Reporting (DIMR) branch of Alberta Health Services (AHS) used the National Ambulatory Care Reporting System (NACRS) to identify patients in Alberta who were admitted for stroke within 90-days of an index emergency department (ED) visit for TIA/minor stroke from April 2010 to March 2016. Information extracted included patient demographics, region of residence (Edmonton, Calgary or non-major urban [NMU]), return diagnosis and timing of return ED visit. Analysis included descriptive summaries and proportions were compared using a χ2 test. Results: During the study period, there were 26,232 index visits to Alberta EDs for TIA/minor stroke. 5426 (26.1%) of patients were admitted on their index visit. Calgary (22.5%) had lower rates of admission on index visit followed by Edmonton (31.4%) and the NMU (46%). 20,806 (79.3%) were discharged home following their index visit. Of the patients discharged on their index visit 729 (3.5%) had an admission for stroke within 90-days of their index ED visit with rates in Edmonton (3.8%) and the NMU regions (3.8%) being significantly higher than Calgary (2.8%, p<0.01). Conclusion: Our study demonstrates significantly lower rates of admission for stroke within 90-days of ED visit for minor stroke/TIA in Calgary compared to Edmonton and the NMU. Further work should focus on validating this result and consideration of standardized care pathways that promote effective resource utilization and quality of care.

Introduction: Age and systolic blood pressure (SBP) are important predictors of Acute Cerebrovascular Syndrome (ACVS). Yet, the effect of SBP is confounded by age, making its independent contribution to ACVS risk difficult to quantify. Here we use logistic regression to explore the role of SBP in younger and older ED patients. Methods: Data comprised 1019 ED patients (ACVS 70%, 30% non-ACVS) enrolled during a 28-month period of an ongoing prospective, observational, multi-site stroke biomarker study (SpecTRA). We used logistic regression to examine the effects of age, sex, and the age:SBP interaction as predictive markers of the diagnosis of ACVS. Results: Participants (53% male) ranged in age from 18 to 97 years (Q1=58, median=70, Q3=80). SBP ranged from 84 to 248 mmHg (Q1=137, median=154, Q3=174). In our initial regression model, age, sex, SBP and the age:SBP interaction were all significant (p<0.01). Using cubic regression splines for age, sbp and their interaction yields the same conclusion (p<0.01). To better understand the role of SBP in younger vs. older patients, we stratified the sample at the median age (70 years of age). In the younger group (n=510), participants were 55% male, 60% ACVS, and had SBP ranging from 91 to 236 mmHg (Q1=133, median=148, Q3=165). In the older group (n=509), participants were 51% male, 82% ACVS and had SBP ranging from 84 to 248 mmHg (Q1=143, median=159, Q3=179), a shift of approximately 10 mmHg between the groups. The logistic regression model was then fit separately to each group without the age:SBP interaction term. In the younger group, we found SBP to be highly significant (p<0.001), with an odds-ratio (OR) of 1.18 per 10 mmHg (95% CI: 1.10-1.29). In the older group, we found that SBP was not significant (p=0.91), with an OR of 1.00 per 10 mmHg (95% CI: 0.91-1.08). Age and sex were also significant risk factors in the younger group (each p<0.01), though not in the older group (both p>0.07). Conclusion: Our findings suggest that for ED patients suspected of ACVS, SBP is a clinically relevant predictor for younger patients, with higher SBP associated with an increased risk of ACVS, regardless of patient age and sex. SBP does not appear to be a strong predictor for patients over 70. ED physicians can leverage this finding by attributing greater importance to elevated SBP in younger patients than older patients when working toward a clinical suspicion of ACVS.

Introduction: The BIG score is a new pediatric trauma score composed of the admission base deficit (BD), the international normalized ratio (INR) and the Glasgow Coma Scale (GCS). A score<16 identifies children with a high probability of survival following blunt trauma.The objective of this study was to measure the criterion validity of the BIG score to predict in-hospital mortality among children visiting an emergency department with blunt trauma requiring an admission to the intensive care unit. Methods: This was a retrospective cohort study performed in a single tertiary care pediatric hospital between 2008 and 2016. Participants were all children (<18 years) visiting the emergency department for a blunt trauma requiring intensive care unit admission or who died at the emergency department. All charts were reviewed by a member of the research team using a standardized report form. To insure quality of data abstraction, 10% of the charts were reviewed in duplicate by a second rater blinded to the first evaluation. The primary outcome was in-hospital mortality. Baseline demographics, initial components of the BIG score, Injury Severity Score (ISS) and disposition were extracted. The primary analysis was the association between a BIG score ≥16 and in-hospital mortality. It was calculated that the inclusion of at least 25 deaths would provide confidence intervals of +/- 0.20 for proportions in the worst-case scenario. Results: Twenty-eight children died among the 336 who met the inclusion criteria. The inter-rater agreement for data abstraction was excellent with kappa scores or intraclass correlation coefficients higher than 0.8 for all variables. Two hundred eighty-four children had information on the three components of the BIG score and they were included in the primary analysis. A BIG score ≥16 demonstrated a sensitivity of 0.93 (95%CI: 0.76-0.98) and specificity of 0.83 (95%CI 0.78-0.87) to identify mortality. Using ROC curves, the area under the curve was higher for the BIG score (0.97; 95%IC: 0.95-0.99) in comparison to the ISS (0.78; 95%IC: 0.71-0.85). Conclusion: The BIG score is an excellent predictor of survival for children visiting the emergency department following a blunt trauma.

Introduction: Approximately 20% of Canadians who present to emergency departments (EDs) with acute asthma relapse within 4 weeks of discharge. The reasons are likely multi-factorial; however, the lack of timely primary care provider (PCP) follow-up and inadequate patient self-management are thought to be important variables. Therefore, we tested the effectiveness of ED-directed multifaceted interventions that targeted PCPs and enhanced patient self-management to reduce asthma relapse following ED discharge. Methods: Adults with acute asthma discharged from 6 Alberta EDs were randomly allocated, in a centralized and concealed manner, to receive usual care (UC), opinion leader [OL] guidance to their PCPs, or OL guidance + nurse case-management [OL+CM] for patients (NCT01079000). The main outcome was asthma relapse within 90-days of ED discharge. Secondary outcomes included PCP visits, time to relapse, hospitalizations and asthma-related quality of life (QoL). Outcomes were collected independently and assessors were masked to intervention assignment. Results: From 943 screened patients, 367 patients were allocated to the study arms (UC=146; OL=110; OL+CM=111). Median age was 28 years, 64% were women, median peak flow at discharge was 350 L/min; 77% were discharged home on prednisone and 85% on either inhaled corticosteroids (ICS) or ICS/long-acting β2-agonists. Compared with UC, both interventions significantly increased rates of relapse at 90-days: UC=12%, OL=28%, OL+CM=19%; p=0.006. Based on an absolute increased risk of 0.16 (95% CI: 0.05, 0.25), the number needed to treat for harm was 6 (95% CI: 3.9, 19.0) for the OL arm. Across study differences in PCP follow-up visits, time to relapse, hospitalizations or asthma-related QoL were not identified. Conclusion: Two different theory-informed and evidenced-based interventions intended to decrease asthma relapse robustly and significantly increased rates of relapse compared with UC. While the reasons for these unintended consequences require further study, we caution against the adoption of similar interventions by other EDs.

Introduction: Many patients presenting to Emergency Departments (EDs) with acute asthma have limited or no access to health care providers, medications and preventive resources. This study explored outpatient care gaps among subjects presenting to the ED for acute asthma, before being discharged. Methods: Cross-sectional analysis of data obtained in a comparative effectiveness trial conducted in six EDs in Alberta (NCT01079000). Data were collected through patient interviews and chart reviews at ED presentation. Two clinician-investigators independently reviewed and adjudicated the following preventive actions: use of spacer devices, written asthma action plans (AAPs) and asthma medication; influenza immunization, cigarette smoking, and referral to asthma education. Agreement between adjudicators was calculated based on kappa (k) statistics. Results: The median age of the study population (n=367) was 28 years and 64% were women. Overall, 26% of patients reported not having a regular family physician. Agreement between reviewers was excellent (k=0.96). More than half (59%) reported not using spacer devices despite being indicated and 3% reported having a written AAP. Following the recommendations of the current asthma guidelines, 38% of the patients required the initiation of inhaled corticosteroids (ICS), 11% required the addition of ICS/long-acting β-agonists combination agents and 39% required reinforcement of adherence with preventer medications. Finally, 37% reported receiving influenza vaccination in the past year, 7% had been referred to asthma education in the last 10 years, and 31% were still smoking, suggesting that cessation counselling was indicated. Conclusion: The ED encounter for patients with acute asthma represents a unique opportunity to establish important partnerships across the continuum of asthma care (e.g., link them with a family doctor). This study provided a robust assessment of the outpatient care gaps in this patient population, which identified many areas for targeted interventions. The method of delivery and type of messaging needs further study.