Evidence-Based Guidelines for Prehospital Airway Management: Methods and Resources Document.

INTRODUCTION: Emergency airway management is a common and critical task EMS clinicians perform in the prehospital setting. A new set of evidence-based guidelines (EBG) was developed to assist in prehospital airway management decision-making. We aim to describe the methods used to develop these EBGs. METHODS: The EBG development process leveraged the four key questions from a prior systematic review conducted by the Agency for Healthcare Research and Quality (AHRQ) to develop 22 different population, intervention, comparison, and outcome (PICO) questions. Evidence was evaluated using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework and tabulated into the summary of findings tables. The technical expert panel then used a rigorous systematic method to generate evidence to decision tables, including leveraging the PanelVoice function of GRADEpro. This process involved a review of the summary of findings tables, asynchronous member judging, and online facilitated panel discussions to generate final consensus-based recommendations. RESULTS: The panel completed the described work product from September 2022 to April 2023. A total of 17 summary of findings tables and 16 evidence to decision tables were generated through this process. For these recommendations, the overall certainty in evidence was "very low" or "low," data for decisions on cost-effectiveness and equity were lacking, and feasibility was rated well across all categories. Based on the evidence, 16 "conditional recommendations" were made, with six PICO questions lacking sufficient evidence to generate recommendations. CONCLUSION: The EBGs for prehospital airway management were developed by leveraging validated techniques, including the GRADE methodology and a rigorous systematic approach to consensus building to identify treatment recommendations. This process allowed the mitigation of many virtual and electronic communication confounders while managing several PICO questions to be evaluated consistently. Recognizing the increased need for rigorous evidence evaluation and recommendation development, this approach allows for transparency in the development processes and may inform future guideline development.

Evidence-Based Guideline for Prehospital Airway Management.

Airway management is a cornerstone of emergency medical care. This project aimed to create evidence-based guidelines based on the systematic review recently conducted by the Agency for Healthcare Research and Quality (AHRQ). A technical expert panel was assembled to review the evidence using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology. The panel made specific recommendations on the different PICO (population, intervention, comparison, outcome) questions reviewed in the AHRQ review and created good practice statements that summarize and operationalize these recommendations. The recommendations address the use of ventilation with bag-valve mask ventilation alone vs. supraglottic airways vs. endotracheal intubation for adults and children with cardiac arrest, medical emergencies, and trauma. Additional recommendations address the use of video laryngoscopy and drug-assisted airway management. These recommendations, and the associated good practice statements, offer EMS agencies and clinicians an opportunity to review the available evidence and incorporate it into their airway management strategies.

Emergency department utilization by persons with rheumatoid arthritis: a population-based cohort study.

Some emergency department (ED) visits by persons with rheumatoid arthritis (RA) may be avoidable. This study aims to describe ED use by persons with RA in Alberta, Canada over a 10-year period. Using linked population-based administrative datasets, the annual frequency of ED visits, timing of visits, acuity at presentation assessed (Canadian Triage Acuity Scale (CTAS)), return visits within 72 h, and final disposition were assessed. Most responsible diagnoses assessed by the ED provider were categorized. Between 2008 and 2017, a total of 48,633 persons with RA had 416,964 unique ED visits. There was a 41% relative increase in visits over the study period and within a fiscal year 37% of persons with RA on average attended an ED. Half of the visits were assessed as CTAS 4 'Less Urgent' (31%) and CTAS 5 'Non-Urgent' (19%). No specific diagnosis could be assigned in 36% of visits and RA was listed as the most responsible diagnosis in 2.5% of all visits. Hospital admissions, occurring on average for 14% of ED visits, increased by 15% over the 10 years, and were rare for CTAS 4 (6.4%) and CTAS 5 (1.4%) presentations. Male patients (difference to female 1.2%, 95%CI 0.6, 1.7) and urban patients (difference to rural 8.4%, 95%CI 7.7, 9.2) were more frequently admitted to hospital. Persons with RA have increased ED utilization over time, with a significant volume of less urgent and non-urgent visits. Opportunities for appropriate ambulatory care provision to reduce acute care use should be identified.

In ischemic stroke, EVT improved 90-d function more than usual care across admission SBP levels.

SOURCE CITATION: Samuels N, van de Graaf RA, Mulder MJHL, et al; HERMES Collaborators. Admission systolic blood pressure and effect of endovascular treatment in patients with ischaemic stroke: an individual patient data meta-analysis. Lancet Neurol. 2023;22:312-319. 36931806.

In large acute ischemic stroke, adding endovascular thrombectomy to medical therapy improved function at 90 d.

SOURCE CITATION: Sarraj A, Hassan AE, Abraham MG, et al; SELECT2 Investigators. Trial of endovascular thrombectomy for large ischemic strokes. N Engl J Med. 2023;388:1259-1271. 36762865.

The impact of the COVID-19 pandemic on transfers between long-term care and emergency departments across Alberta.

BACKGROUND: Long-term care (LTC) was overwhelmingly impacted by COVID-19 and unnecessary transfer to emergency departments (ED) can have negative health outcomes. This study aimed to explore how the COVID-19 pandemic impacted LTC to ED transfers and hospitalizations, utilization of community paramedics and facilitated conversations between LTC and ED physicians during the first four waves of the pandemic in Alberta, Canada. METHODS: In this retrospective population-based study, administrative databases were linked to identify episodes of care for LTC residents who resided in facilities in Alberta, Canada. This study included data from January 1, 2018 to December 31, 2021 to capture outcomes prior to the onset of the pandemic and across the first four waves. Individuals were included if they visited an emergency department, received care from a community paramedic or whose care involved a facilitated conversation between LTC and ED physicians during this time period. RESULTS: Transfers to ED and hospitalizations from LTC have been gradually declining since 2018 with a sharp decline seen during wave 1 of the pandemic that was greatest in the lowest-priority triage classification (CTAS 5). Community paramedic visits were highest during the first two waves of the pandemic before declining in subsequent waves; facilitated calls between LTC and ED physicians increased during the waves. CONCLUSIONS: There was a reduction in number of transfers from LTC to EDs and in hospitalizations during the first four waves of the pandemic. This was supported by increased conversations between LTC and ED physicians, but was not associated with increased community paramedic visits. Additional work is needed to explore how programs such as community paramedics and facilitated conversations between LTC and ED providers can help to reduce unnecessary transfers to hospital.

A mapping review of interventions to address patients who frequently seek care in the emergency department.

BACKGROUND: The high utilization of acute care services, particularly emergency departments (ED), continues to be a significant concern for healthcare providers. Numerous approaches have been studied to meet the care needs of patients who frequently seek care in the ED; however, there is no comprehensive review of the current literature base. As such, a current understanding of the interventions initiated within the ED to address the needs of frequent users is required. This mapping review had three objectives: identify the characteristics associated with the need to frequently seek care in the ED; identify interventions implemented to address the needs of this population; and identify gaps in the current evidence base. METHODS: A knowledge map was created by scoping the literature to identify characteristics associated with frequent ED use and interventions implemented to address frequent use. Then, a literature search was conducted to determine what has been implemented by EDs to reduce frequent ED use. The literature was searched from 2013 to January 2023. MeSH terms and keywords were used to identify relevant studies. Studies implementing an intervention for those with characteristics associated with frequent ED use and reporting on ED use were included. RESULTS: Twenty-three (23) controlled trials and 35 observational studies were included. The most common populations were older adults, those with chronic conditions, and generic "frequent users". No studies assessed Indigenous Peoples or racial minorities, and few assessed patients with a disability or patients experiencing homelessness. The most common interventions were referrals, care plans, case management, care coordination, and follow-up phone calls. Most studies reported ED revisits, hospitalization, costs, length-of-stay, or outpatient utilization. Few assessed patient or staff perspectives. About one-third of studies (n = 24) reported significant reductions in ED revisits. CONCLUSIONS: Similar interventions, mainly focused on care coordination and planning, have been implemented to address frequent use of the ED. There are still significant gaps in the populations that have been studied. Efforts now must be undertaken to study more diverse populations whose care needs are not being met elsewhere and thus frequent the ED often.

Evidence-Based Guidelines for Prehospital Pain Management: Literature and Methods.

INTRODUCTION: Emergency Medical Services (EMS) clinicians commonly encounter patients with acute pain. A new set of evidence-based guidelines (EBG) was developed to assist in the prehospital management of pain. Our objective was to describe the methods used to develop these evidence-based guidelines for prehospital pain management. METHODS: The EBG development process was supported by a previous systematic review conducted by the Agency for Healthcare Research and Quality (AHRQ) covering nine different population, intervention, comparison, and outcome (PICO) questions. A technical expert panel (TEP) was formed and added an additional pediatric-specific PICO question. Identified evidence was evaluated using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework and tabulated into Summary of Findings tables. The TEP then utilized a rigorous systematic method, including the PanelVoice function, for recommendation development which was applied to generate Evidence to Decision Tables (EtD). This process involved review of the Summary of Findings tables, asynchronous member judging, and facilitated panel discussion to generate final consensus-based recommendations. RESULTS: The work product described above was completed by the TEP panel from September 2020 to April 2021. For these recommendations, the overall certainty of evidence was very low or low, data for decisions on cost effectiveness and equity were lacking, and feasibility was rated well across all categories. Based on the evidence, one strong and seven conditional recommendations were made, with two PICO questions lacking sufficient evidence to generate a recommendation. CONCLUSION: We describe a protocol that leveraged established EBG development techniques, the GRADE framework in conjunction with a previous AHRQ systematic review to develop treatment recommendations for prehospital pain management. This process allowed for mitigation of many confounders due to the use of virtual and electronic communication. Our approach may inform future guideline development and increase transparency in the prehospital recommendations development processes.

Evidence-Based Guidelines for Prehospital Pain Management: Recommendations.

This project sought to develop evidence-based guidelines for the administration of analgesics for moderate to severe pain by Emergency Medical Services (EMS) clinicians based on a separate, previously published, systematic review of the comparative effectiveness of analgesics in the prehospital setting prepared by the University of Connecticut Evidence-Based Practice Center for the Agency for Healthcare Research and Quality (AHRQ). A technical expert panel (TEP) was assembled consisting of subject matter experts in prehospital and emergency care, and the development of evidence-based guidelines and patient care guidelines. A series of nine "patient/population-intervention-comparison-outcome" (PICO) questions were developed based on the Key Questions identified in the AHRQ systematic review, and an additional PICO question was developed to specifically address analgesia in pediatric patients. The panel made a strong recommendation for the use of intranasal fentanyl over intravenous (IV) opioids for pediatric patients without intravenous access given the supporting evidence, its effectiveness, ease of administration, and acceptance by patients and providers. The panel made a conditional recommendation for the use of IV non-steroidal anti-inflammatory drugs (NSAIDs) over IV acetaminophen (APAP). The panel made conditional recommendations for the use of either IV ketamine or IV opioids; for either IV NSAIDs or IV opioids; for either IV fentanyl or IV morphine; and for either IV ketamine or IV NSAIDs. A conditional recommendation was made for IV APAP over IV opioids. The panel made a conditional recommendation against the use of weight-based IV ketamine in combination with weight-based IV opioids versus weight-based IV opioids alone. The panel considered the use of oral analgesics and a conditional recommendation was made for either oral APAP or oral NSAIDs when the oral route of administration was preferred. Given the lack of a supporting evidence base, the panel was unable to make recommendations for the use of nitrous oxide versus IV opioids, or for IV ketamine in combination with IV opioids versus IV ketamine alone. Taken together, the recommendations emphasize that EMS medical directors and EMS clinicians have a variety of effective options for the management of moderate to severe pain in addition to opioids when designing patient care guidelines and caring for patients suffering from acute pain.

In suspected PE, YEARS rule + D-dimer was noninferior to D-dimer alone to rule out PE.

SOURCE CITATION: Freund Y, Chauvin A, Jimenez S, et al. Effect of a diagnostic strategy using an elevated and age-adjusted D-dimer threshold on thromboembolic events in emergency department patients with suspected pulmonary embolism: a randomized clinical trial. JAMA. 2021;326:2141-9. 34874418.

In ischemic stroke upon wakening, IV thrombolysis or EVT improves functional outcome at 90 d.

SOURCE CITATION: Roaldsen MB, Lindekleiv H, Mathiesen EB. Intravenous thrombolytic treatment and endovascular thrombectomy for ischaemic wake-up stroke. Cochrane Database Syst Rev. 2021;12:CD010995. 34850380.

WHO provides 2 conditional recommendations for casirivimab-imdevimab combination therapy in COVID-19.

SOURCE CITATION: Rochwerg B, Agarwal A, Siemieniuk RA, et al. Update to living WHO guideline on drugs for covid-19. BMJ. 2021;374:n2219. 34556469.

In acute ischemic stroke, early IV tenecteplase was noninferior to alteplase for excellent functional outcome.

SOURCE CITATION: Menon BK, Buck BH, Singh N, et al. Intravenous tenecteplase compared with alteplase for acute ischaemic stroke in Canada (AcT): a pragmatic, multicentre, open-label, registry-linked, randomised, controlled, non-inferiority trial. Lancet. 2022;400:161-9. 35779553.

Emergency physician gender is associated with early pregnancy loss management: a multisite retrospective cohort study.

BACKGROUND: Patients experiencing early pregnancy loss often first present to the emergency department (ED) where they can be managed non-operatively through expectant or medical management, or surgically by the obstetrical team. Studies have reported that physician gender can influence clinical decision making, but there is limited research on this phenomenon in the ED. The objective of this study was to determine whether emergency physician gender is associated with early pregnancy loss management. METHODS: Data were retrospectively collected from patients who presented to Calgary EDs with a non-viable pregnancy from 2014 to 2019. Pregnancies >12 weeks gestational age were excluded. The emergency physicians included saw at least 15 cases of pregnancy loss over the study period. The primary outcome was obstetrical consult rates by male versus female emergency physicians. Secondary outcomes included rates of initial surgical evacuation via dilation and curettage (D&C) procedures, ED returns, returns to care for D&Cs and total D&C rates. Data were analysed using χ2, Fisher's exact and Mann-Whitney U tests, as appropriate. Multivariable logistic regression models accounted for physician age, years of practice, training programme and type of pregnancy loss. RESULTS: 98 emergency physicians and 2630 patients from 4 ED sites were included. 76.5% of the physicians were male accounting for 80.4% of pregnancy loss patients. Patients seen by female physicians were more likely to receive an obstetrical consultation (adjusted OR (aOR) 1.50, 95% CI 1.22 to 1.83) and initial surgical management (aOR 1.35, 95% CI 1.08 to 1.69). ED return rates and total D&C rates were not associated with physician gender. CONCLUSION: Patients seen by female emergency physicians had higher rates of obstetrical consultation and initial operative management compared with those seen by male emergency physicians, but outcomes were similar. Additional research is required to determine why these gender differences exist and how these discrepancies may impact the care of early pregnancy loss patients.

Acute outreach service to nursing homes: A systematic review with GRADE and triple aim approach.

BACKGROUND: People living in nursing homes face the risk of visiting the emergency department (ED). Outreach services are developing to prevent unnecessary transfers to ED. AIMS: We aim to assess the performance of acute care services provided to people living in nursing homes or long-term homecare, focusing on ED transfer prevention, safety, cost-effectiveness and experiences. MATERIALS & METHODS: This review was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Studies were eligible for inclusion if they were peer-reviewed and examined acute outreach services dedicated to delivering care to people in nursing homes or long-term homecare. The service models could also have preventive components. The databases searched were Scopus and CINAHL. In addition, Robins-I and SIGN checklists were used. The primary outcomes of prevented ED transfers or hospitalisations and the composite outcome of adverse events (mortality/Emergency Medical Service or ED visit after outreach service contact related to the same clinical condition) were graded with GRADE. RESULTS: Fifteen relevant original studies were found-all were observational and focused on nursing homes. The certainty of evidence for acute outreach services with preventive components to prevent ED transfers or hospitalisations was low. Stakeholders were satisfied with these services. The certainty of evidence for solely acute outreach services to prevent ED transfers or hospitalisations was very low and inconclusive. Reporting of adverse events was inconsistent, certainty of evidence for adverse events was low. CONCLUSION: Published data might support adopting acute outreach services with preventive components for people living in nursing homes to reduce ED transfers, hospitalisations and possibly costs. If an outreach service is started, it is recommended that a cluster-randomised or quasi-experimental research design be incorporated to assess the effectiveness and safety of the service. More evidence is also needed on cost-effectiveness and stakeholders' satisfaction. Systematic review registration number: PROSPERO CRD42020211048, date of registration: 25.09.2020.

Factors to Consider During Identification and Invitation of Individuals in a Multi-stakeholder Research Partnership.

BACKGROUND: Health research teams increasingly partner with stakeholders to produce research that is relevant, accessible, and widely used. Previous work has covered stakeholder group identification. OBJECTIVE: We aimed to develop factors for health research teams to consider during identification and invitation of individual representatives in a multi-stakeholder research partnership, with the aim of forming equitable and informed teams. DESIGN: Consensus development. PARTICIPANTS: We involved 16 stakeholders from the international Multi-Stakeholder Engagement (MuSE) Consortium, including patients and the public, providers, payers of health services/purchasers, policy makers, programme managers, peer review editors, and principal investigators. APPROACH: We engaged stakeholders in factor development and as co-authors of this manuscript. Using a modified Delphi approach, we gathered stakeholder views concerning a preliminary list of 18 factors. Over two feedback rounds, using qualitative and quantitative analysis, we concentrated these into ten factors. KEY RESULTS: We present seven highly desirable factors: 'expertise or experience', 'ability and willingness to represent the stakeholder group', 'inclusivity (equity, diversity and intersectionality)', 'communication skills', 'commitment and time capacity', 'financial and non-financial relationships and activities, and conflict of interest', 'training support and funding needs'. Additionally, three factors are desirable: 'influence', 'research relevant values', 'previous stakeholder engagement'. CONCLUSIONS: We present factors for research teams to consider during identification and invitation of individual representatives in a multi-stakeholder research partnership. Policy makers and guideline developers may benefit from considering the factors in stakeholder identification and invitation. Research funders may consider stipulating consideration of the factors in funding applications. We outline how these factors can be implemented and exemplify how their use has the potential to improve the quality and relevancy of health research.

A Systematic Review of Interventions to Reduce Computed Tomography Usage in the Emergency Department.

STUDY OBJECTIVE: Unnecessary computed tomography (CT) scans burden the health care system, leading to increased emergency department (ED) wait times and lengths of stay, costing almost a billion dollars annually. This study aimed to describe ED-based interventions that are most effective at reducing CT imaging while maintaining diagnostic accuracy and patient safety. METHODS: Adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, MEDLINE, Embase, CINAHL, Cochrane Central Register of Controlled Trials, and Google Scholar were searched until December 31, 2020. Randomized and nonrandomized studies that assessed the effect of an ED-based intervention on CT scan usage were included. Abstract screening, data extraction, and quality assessment were conducted in duplicate. The Grading of Recommendation Assessment, Development and Evaluation framework, with the Risk of Bias 2 and Risk of Bias in Nonrandomized Studies - of Interventions tools, was used to determine the certainty of evidence. Significant clinical and statistical heterogeneity precluded meta-analysis; hence, a narrative synthesis was conducted. RESULTS: A total of 149 studies were included of 5,667 screened abstracts, with substantial interrater reliability among reviewers (Cohen's κ>0.60). The CT reduction strategies were categorized into 15 single and 11 multimodal interventions by consensus review. Interventions that consistently reduced CT usage included diagnostic pathways, alternative test availability, specialist involvement, and provider feedback. Family/patient education, clinical decision support tools, or passive guideline dissemination did not consistently reduce usage. Only 44% of studies reported unintended consequences of reduction strategies; however, these showed no increase in missed diagnoses or patient harm. The interventions that engaged multiple specialties during planning/implementation had a greater reduction effect than ED only. The certainty of evidence for the primary outcome was very low. CONCLUSION: Multidisciplinary-led interventions that provided an alternative to CT imaging were the most effective at reducing usage and did so without compromising patient safety.

Functional Decline After Nonhospitalized Injuries in Older Patients: Results From the Canadian Emergency Team Initiative Cohort in Elders.

STUDY OBJECTIVE: To estimate the cumulative incidence of functional decline over 6 months following emergency department (ED) assessments of nonhospitalized injuries and to identify its main determinants. METHODS: We conducted a prospective multicenter cohort of older adults discharged home following assessment for injuries in 8 Canadian EDs. Participants were assessed at 3 time points: baseline in the ED, 3 months, and 6 months. The primary outcome, functional decline, was defined as a 2-points loss from baseline on the Older American Resources Scale (OARS). Other measures included demographics, comorbidities, injury characteristics, frailty, cognition, mobility status, etc. Cumulative incidences were estimated using proportions with 95% confidence intervals. Log-binomial regressions and the "least absolute shrinkage and selection operator" (LASSO) were used to identify significant functional decline determinants. RESULTS: Among 2,919 participants, 403 (13.8%) were lost to follow-up. Mean age was 76.2±7.6 years, 65.3% were women, 9% were frail, and 40.0% prefrail. Main injury mechanisms were falls (65.5%) and motor vehicle accidents (18.6%). The cumulative incidence of functional decline over 6 months was 17.0% (95% confidence interval 12.5% to 23.0%). Occasional use of walking devices, less than 5 outings/week, frailty, and older age were significant baseline determinants of functional decline. CONCLUSION: A significant 17% of older adults with "minor" injuries experience a persistent functional decline over 6 months following their ED visit. Four frailty-related determinants were identified: occasional use of a walking device, less than 5 outings/week, frailty, and older age. Further work is needed to assess if these can help ED clinicians screen seniors at risk and initiate interventions at discharge.

Prevalence, management and outcomes of unrecognized delirium in a National Sample of 1,493 older emergency department patients: how many were sent home and what happened to them?

BACKGROUND: Retrospective studies estimate Emergency Department (ED) delirium recognition at <20%; few prospective studies have assessed delirium recognition and outcomes for patients with unrecognized delirium. OBJECTIVES: To prospectively measure delirium recognition by ED nurses and physicians, document their confidence in diagnosis and disposition, actual dispositions, and patient outcomes. METHODS: Prospective observational study of people ≥65 years. We assessed delirium using the Confusion Assessment Method, then asked ED staff if the patient had delirium, confidence in their assessment, if the patient could be discharged, and contacted patients 1 week postdischarge. We report proportions and 95% confidence intervals (Cls). RESULTS: We enrolled 1,493 participants; mean age was 77.9 years; 49.2% were female, 79 (5.3%, 95% CI 4.2-6.5%) had delirium. ED nurses missed delirium in 43/78 cases (55.1%, 95% CI 43.4-66.4%). Nurses considered 12/43 (27.9%) patients with unrecognized delirium safe to discharge. Median confidence in their delirium diagnosis for patients with unrecognized delirium was 7.0/10. Physicians missed delirium in 10/20 (50.0%, 95% CI 27.2-72.8) cases and considered 2/10 (20.0%) safe to discharge. Median confidence in their delirium diagnosis for patients with unrecognized delirium was 8.0/10. Fifteen patients with unrecognized delirium were sent home: 6.7% died at 1 week follow-up vs. none in those with recognized delirium and 1.1% in the rest of the cohort. CONCLUSION: Delirium recognition by nurses and physicians was sub-optimal at ~50% and may be associated with increased mortality. Research should explore root causes of unrecognized delirium, and novel strategies to systematically improve delirium recognition and patient outcomes.

The impact of emergency department crowding on admission decisions and patient outcomes.

OBJECTIVES: The objective of this study is to evaluate the impact of emergency department (ED) crowding levels on patient admission decisions and outcomes. METHODS: A retrospective study was performed based on 2-year electronic health record data from a tertiary care hospital ED in Alberta, Canada. Using modified Poisson regression models, we studied the association of patient admission decisions and 7-day revisit probability with ED crowding levels measured by: 1) the total number of patients waiting and in treatment (ED census), 2) the number of boarding patients (boarder census), and 3) the average physician workload, calculated by the total number of ED patients divided by the number of physicians on duty (physician workload census). The control variables included age, gender, treatment area, triage level, and chief complaint. A subgroup analysis was performed to evaluate the heterogeneous effects among patients of different acuity levels. RESULTS: Our dataset included 141,035 patient visit records after cleaning from August 2013 to July 2015. The patient admission probability was positively correlated with ED census (relative risk [RR] = 1.006, 95% confidence interval [CI] = 1.005 to 1.007) and physician workload census (RR = 1.029, 95% CI = 1.027 to 1.032), but inversely correlated with boarder census (RR = 0.991, 95% CI = 0.989 to 0.993). We further found that the 7-day revisit probability of discharged patients was positively associated with boarder census (RR = 1.009, 95% CI = 1.004 to 1.014). CONCLUSIONS: Patient admission probability was found to be directly associated with ED census and physician workload census, but inversely associated with the boarder census. The effects of boarder census and physician workload census were stronger for patients of triage levels 3-5. Our results suggested that (i) insufficient physician staffing may lead to unnecessary patient admissions; (ii) too many boarding patients in ED leads to an increase in unsafe discharges, and as a result, an increase in 7-day revisit probability.