Introduction: As part of our audit and feedback process, Emergency Physicians (EP) are provided feedback on flow metrics and resource utilization. We analysed the relationship between two specific metrics (adjusted workload measurement (AWM), with the number of patients seen per hour adjusted according to CTAS, and percentage of revisits within 72 hours and diagnostic imaging use. Unfortunately, we are unable to evaluate quality of care, nor appropriateness of DI indication at this stage. Methods: We used data from 86 physicians at an academic ED, from June 1, 2015 to May31, 2017. The Data Envelope Analysis (DEA) model incorporated performance quality measures as outputs and efficiency measures as inputs. DEA is a method widely used in physician performance analysis. The method provides a score (optimal performance efficiency-OPE) for each EP based on maximization of the performance (AWM) in proportion to the combination of efficient use of resources, diagnostic imaging (DI). The score was used to regress against demographic characteristics and training. Results: The median AWM was 6.8 (quartiles Q1-Q3 = 6.4-7.4) with the median diagnostic imaging use of percentages of CT (median = 10.1, 8.6-11.9), US (median = 4.7, 3.6-5.6) and x-ray (80, 74-84). The EPs who had highest AWM combined with least use of DI (OPE = 100%), provided median AWM of 9.1 (range 8.9-9.7) with percentage CT, US and x-ray medians at 5.8% (range 5.8-6.2), 2.7% (range 2.4-3.6) and 59% (range 59-72). These provided benchmarks for optimal performance indicators. We found statistically significant differences of OPE scores based on gender (men 4.1 times higher, p < 0.001) and degree (RCPS < CCFPEM, Other < CCFPEM, p < 0.001). Overall AWM diminishes at the rate of 14% (95%CI: 9-20%) for a combination of 100 DI tests ordered. In order to reach the optimal level of performance, to reach an OPE of 100%, the median CT use percentage needs to be reduced by 6% (quartile range 3.9- 7.7%), US by 2.2% (quartile range 1.5-3.4%) and x-rays by 37.2% (quartile range: 26.8-44.3%). Return visit rates were not associated with DI use, possibly due to homogeneity in the percentage of return visits. Conclusion: We found significant performance variations in terms of average workload measurement in proportion to the weighted average of diagnostic imaging use, with increased use of DI being associated with decreasing AWM. Percentage of return visits does not appear to be useful as a performance indicator.

Introduction: Clinicians treating children in the emergency department (ED) are especially concerned with the efficacy and safety of imaging. Interventions to limit imaging have been proposed to maximize benefits and avoid risks; however, the types and effectiveness of interventions employed in pediatric EDs have not been examined in detail. Methods: Electronic databases and grey literature were systematically searched by a medical librarian. Comparative studies of ED-based interventions reporting computed tomography (CT), radiography (XR), or ultrasound (US) outcomes were included. Interventions introducing new imaging equipment or personnel to the ED, ED diversion strategies, and pre-admission protocols were excluded. At least two independent reviewers assessed each study for inclusion based on pre-defined criteria and extracted data. Disagreements were resolved through consensus. Descriptive results are reported. Results: Overall, 38 pediatric studies were included. Most (66%) interventions implemented two or more components; the most common intervention components were clinical guidelines or pathways (87%) and education or information (66%). Studies were categorized by presentation type: traumatic (n = 27); non-traumatic (n = 19), or combined ‘all-comers’ (n = 2). Included studies reported 62 imaging outcomes (CT = 29; XR = 20; US = 13). Among traumatic studies, 26 imaging outcomes were reported; CT was the most commonly reported outcome (CT = 15; XR = 9; US = 1). Of the CT outcomes, 33% reported significant decreases and five decreased but were either not significant or did not report significance. XR significantly decreased in 44% (4/9). In the non-traumatic studies, the most common imaging outcome remained CT (12 outcomes); 58% of which reported significant decreases. XR was the second most frequent outcome, with 63% reporting significant reductions. Combined success of the interventions to reduce CT and XR was 60%. Reported changes in ordering were less consistent in US. Conclusion: Multifaceted passive interventions have been implemented to reduce imaging in pediatric EDs. Most reported some success changing ordering practices, specifically among patients with non-trauma presentations. Future research exploring relationships between intervention content, effectiveness, and fidelity may provide insight into how to develop more effective interventions to change image ordering in the ED and guide which presentations to target.

Introduction: Variation in image ordering exists across Alberta emergency departments (EDs). Evidence-based, pocket-sized knowledge dissemination tools were developed for two conditions (acute asthma [AA] and benign headache [BHA]) for which imaging (chest x-ray [CXR] and computed tomography [CT], respectively) has limited utility. This study explored tool acceptability among ED patients and emergency physicians (EPs). Methods: Tool feedback was provided by EPs, via online survey, and adult patients with AA and BHA via in-person survey. EPs qualitative interviews further explored communication tools. Preliminary descriptive analyses of survey responses and content analysis of interview data were conducted. Results: Overall, 55 EPs (55/192; 29%) and 38 consecutive patients participated in the AA study; 73 EPs (73/192; 38%) and 160 patients participated in the BHA study. In both studies, approximately 50% of EPs felt comfortable using the tool; however, they suggested including radiation risk details and imaging indications and removing references to imaging variation and health system cost. In the BHA study, EPs opposed the four Choosing Wisely® campaign questions fearing they would increase imaging expectations. In both conditions, most patients ( >90%) understood the content and 68% felt the information applied to them. Less than half (AA:45%; BHA: 38%) agreed that they now knew more about when a patient should have imaging workup done. Following tool review, 71% of AA and 50% of BHA patients stated they would discuss their imaging needs with their ED care provider today or during a future presentation. Both patient groups suggested including: additional imaging details (i.e., indications, risk, clinical utility), removing imaging overuse references, and including instructions that encourage patients to ask their EP questions. EP interviews (n = 12) identified preferences for personalized and interactive tools. Tensions were perceived around ED time pressure as well as remuneration schemes that fail to prioritize patient conversation. Tool centralization, easy access, and connection with outpatient support were also key themes. Conclusion: Both patients and EPs provided valuable information on how to improve ED knowledge dissemination tools, using two chronic conditions to demonstrate how these changes would improve tool utility. Implementing these recommendations, and considering preferences of EPs and patients, may improve future tool uptake and impact.

Introduction: Clinical context is critical for accurate radiologic interpretation of neuroimaging investigations. The aim of this study was to determine the impact of a change in the Emergency Department (ED) computerized provider order entry (CPOE) interface on the quality of clinical information conveyed in ED neuroimaging requisitions for suspected stroke patients. Methods: Four local EDs utilizing a common CPOE ED Stroke order set were studied before and after the introduction of a mandatory blank free text field requiring clinical information for the radiologist before a computed tomography angiography (CTA) request could be submitted. Prior to this modification, the indication (acute stroke) was pre-filled in the CTA request for convenience with the option of providing additional information at the discretion of the ordering physician. ED physicians were informed of the change as well as the rationale for its implementation. A retrospective pre (90 days) post (30 days) analysis was conducted across four local EDs to evaluate the impact of the CPOE user interface change on the quality of clinical information provided on neuroimaging orders. Patients aged 18 with CTA head and/or neck orders submitted from the order set were included. Patients were excluded if the CTA order was submitted outside of the ED Stroke order set, if order entry was by non-physician personnel, or if the order was modified by the diagnostic imaging department after ED submission. Clinical information from CTA orders were scored as complete, partial, or absent/uninformative based on a standardized rubric of critical elements, including: description of neurological deficit(s), lateralization, and timing of symptom onset or duration. Results were analyzed using chi square analysis. Results: Pre-implementation data from Oct 1, 2015 Jan 1, 2016 (N=652) was compared to post-implementation data from Nov 1 30, 2016 (N=227). The proportion of complete, partial, and absent/uninformative clinical histories were: 45.3%, 31.4%, and 23.3% in the pre-implementation period and 62.6%, 37.4%, and 0% in the post-implementation period respectively. There was a 38.2% relative increase in complete clinical histories, a 19.1% relative increase in partial clinical histories, and a 100% reduction in absent/uninformative clinical histories (p<0.001). Conclusion: The introduction of a mandatory free text field significantly increased the overall quality of clinical information provided on ED neuroimaging orders. This CPOE strategy has the potential to improve diagnostic accuracy and reduce unnecessary delays to imaging interpretation caused by lack of clinical information.

Introduction: Utilization of CT pulmonary angiography (CTPA) to rule out pulmonary embolism (PE) has risen dramatically but diagnostic yield has fallen over the past several decades, suggesting that lower risk patients are being tested. Given little evidence to suggest improved patient outcomes with higher CTPA utilization, and increasing evidence of harm, evidence-based guidelines have been developed to reduce unnecessary CTPA use. The objective of this study was to assess the impact of an electronic clinical decision support (CDS) intervention to reduce unnecessary CTPA utilization for emergency department (ED) patients with suspected PE. Methods: This was a cluster-randomized, controlled trial with physicians as the unit of randomization. All emergency physicians (EPs) at 4 urban adult EDs and 1 urgent care center were randomly assigned to receive either evidence-based imaging CDS for patients with suspected PE (intervention) or no CDS (control) over a 1-year study period. CDS was launched in an external web browser whenever an intervention EP ordered a CTPA from the computerized physician order entry software for ED patients CTAS 2-5; however, physician interaction with CDS was voluntary. The CDS tool enabled calculation of patient-specific information, including the patients Wells score, PERC score, and age-adjusted D-dimer, as well as prediction of each patients pre-test risk of PE along with an imaging/no imaging recommendation. CDS recommendations could be printed for the medical record as could educational patient handouts to support physician decision-making. The primary outcome was CTPA utilization for patients with CEDIS chief complaints of shortness of breath or chest pain on the index visit. Secondary outcomes included index visit length of stay (LOS), and CTPA use or VTE diagnosis within 90-days. This study was REB approved. Results: Demographics were similar among intervention and control EPs; however, during a 2-year pre-intervention period control EPs had a higher baseline CTPA rate (8.5% vs 7.7%, p<0.001). In the first 8-months following CDS implementation, 94 intervention EPs saw 9,609 patients and voluntarily interacted with the CDS tool on 43.2% of eligible encounters while 91 control EPs saw 9,498 patients. CTPA utilization was higher among intervention EPs than control (9.6% vs 8.3%, p<0.001) as was ED LOS (302 vs 287 minutes, p<0.001). There was no difference in 90-day CTPA use or VTE diagnoses. Conclusion: In one of the largest RCTs of CDS to date, exposure to CDS was associated with higher rates of CTPA utilization and longer ED LOS on the index visit, and no difference in 90-day CT use or VTE diagnoses. These results differ from a concurrent study of CDS for patients with mild traumatic brain injury in the same physician population and may relate to the implementation of the CDS intervention and/or complexity of the underlying evidence-based algorithms.

Introduction: Practice guidelines discourage routine imaging for low back pain and recommend selective use when serious underlying conditions are suspected. Evidence about prevalence of serious pathologies and accuracy of red flags for decision-making is limited. We describe rates of serious low back pathology, assess the accuracy of three red flags and model the utility of combining administratively available red flags to reduce imaging. Methods: A seven-year retrospective study of patients presenting with low back pain to four emergency departments in Nova Scotia, Canada. Patient characteristics were available from administrative data. We test sensitivity, specificity, positive and negative predictive values, and likelihood ratios of individual and combinations of red flags. We use decision curve analyses to assess the clinical utility of three red flags to inform imaging. Results: We included data from 38,714 patients presenting with low back pain. Serious low back pathology was diagnosed in 1196(3.09%): 847 (2.19%) were vertebral fractures, 184 (0.48%) unstable fracture, 262 (0.68%) cancer, 57 (0.15%) cauda equina syndrome, and 30 (0.08%) spinal infection. Value of combining three red flags (age>65, female sex, and trauma) was found: positive likelihood ratios of 4.36 and 9.74 for vertebral fracture and unstable fracture, respectively. Imaging for low back pain could be reduced by 28 per 100 patients using a model that incorporates sex, age >65, and trauma. Conclusion: Serious low back pathology is extremely rare in patients presenting with low back pain. Combinations of red flags readily available in emergency departments have the potential to reduce unnecessary imaging tests.

Introduction: Pulmonary embolism (PE) is a diagnostic challenge, since it shares symptoms with other conditions. Missed diagnosis puts patients at a risk of a potentially fatal outcome, while false positive results leave them at risk of side effects (bleeding) from unnecessary treatment. Diagnosis involves a multi-step pathway consisting of clinical prediction rules (CPRs), laboratory testing, and diagnostic imaging, but the best strategy in the Canadian context is unclear. Methods: We carried out a systematic review of the diagnostic accuracy, clinical utility, and safety of diagnostic pathways, CPRs, and diagnostic imaging for the diagnosis of PE. Clinical prediction rules were studied by an overview of systematic reviews, and pathways and diagnostic imaging by a primary systematic review. Where feasible, a diagnostic test meta-analysis was conducted, with statistical adjustment for the use of variable and imperfect reference standards across studies. Results: The Wells CPR rule showed greater specificity than the Geneva, but the relative sensitivities were undetermined. Application of a CPR followed by with D-dimer laboratory testing can safely rule out PE. In diagnostic test accuracy meta-analysis, computed tomography (CT) (sensitivity 0.973, 95% CrI 0.921 to 1.00) and ventilation/perfusion single-photon emission CT (VQ-SPECT) (sensitivity 0.974, 95% CrI 0.898 to 1.00) had the highest sensitivity) and CT the highest specificity (0.987, 95% CrI 0.958 to 1.00). VQ and VQ-SPECT had a higher proportion of indeterminate studies, while VQ and VQ-SPECT involved lower radiation exposure than CT. Conclusion: CPR and D-dimer testing can be used to avoid unnecessary imaging. CT is the most accurate single modality, but radiation risk must be assessed. These findings, in conjunction with a recent health technology assessment, may help to inform clinical practice and guidelines.