ED Presentations Research Articles

233TF Making Antibiotics Stick

Learning in medical school still involves the “fire hose model” where vast amounts of information are being unleashed upon the students who are then responsible for recalling as much as possible in order to pass the test. The challenge occurs when they make the transition from fourth year medical student to first year resident, when they need to recall the information learned, more precisely, when they need to know what information to recall. In order to facilitate this transition, I created an online module to help the students recall useful antibiotics once they start their internship and are faced with the responsibility of being a doctor. In creating this module, I am combining two well-known teaching principles, that of microteaching alongside spaced repetition. In order to take these concepts a step forward, I am combining the spaced repetition model with information retrieval as I am using the same quiz at different time intervals in order to facilitate long-term knowledge retention. This course is not intended to teach new information, but rather to allow the learner to review information once known and facilitate long-term recall. Study objective: At the end of this elective, the student/intern will be able to recall specific antibiotics to treat different infections, from common ones to deadly ones. Along with the name of the antibiotic, the student will recall the dose, the route and the frequency. During their emergency medicine core rotation, the learner will be enrolled in an online module composed of three sessions: two knowledge review sessions and a quiz. They will start by completing the quiz first, followed by completion of the two educational sessions during the assigned two weeks. Within one day of completing lesson two, the learner will be required to take the same quiz again. Once the student finishes the EM rotation, they will be enrolled in spaced repetition schedule where they will be receiving emails with links to take the same quiz at progressively longer time intervals. In this way, we can use spaced information retrieval to solidify the knowledge. The two knowledge review sessions were created using Articulate 360 platform and the spaced quizzes were created using “Google forms,” a free online resource. The spaced quizzes would serve both as teaching tools as well as assessment tools. The learner will also be asked to complete a 4-question survey of the course in order to fine tune the material as well as identify other potential topics suited for this format. Teaching antibiotics is just one of the thousands of topics that can be adapted to this teaching format. I can easily see this format being used to teach rare EKG finding such as Brugada, DeWInter’s, etc. This format could also be used to teach management of common ED presentations such as CHF exacerbation, PNA, sepsis etc. I can see creating an online database of such short, teaching modules that the learner can access at any time, modules focused on knowledge needed during “transition to the wards.”

179 Ethnicity and Symptom Onset in the Emergency Department during the SARS-CoV-2 Pandemic at the “Epicenter of the Epicenter”

New York City Health + Hospitals/Elmhurst located in Queens, New York, has one of the most diverse patient populations in the United States, and likely the world, and was deemed the “epicenter” of the Covid-19 pandemic in 2020. Given its unique population, high number of Covid-19 cases and growing concern that Covid-19 disproportionately affects minority patients, this study seeks to examine the correlation between ethnicity and time from symptoms onset to ED presentation to further understand this disparity.

Estimating the proportion of patients who transition to long-term opioid use following oxycodone initiation in the emergency department.

To report the number of patients discharged from ED with oxycodone immediate release (IR) over 12 months and estimate the proportion who potentially transition to long-term opioid use and subsequent injectable heroin use. Retrospective observational data were collected from a major tertiary-referral metropolitan ED in Melbourne, Australia, describing the number of patients discharged with an oxycodone IR prescription and proportion of discharge scripts filled. These data were projected against published data reporting trends on patients' trajectory to long-term opioid use, to subsequently estimate the proportion of patients from this cohort that may transition to injectable heroin use. Of the 87 551 ED presentations in 2018, there were 4843 prescriptions written for oxycodone IR for 4102 different patients. An estimated 279 patients may become long-term opioid users following initial ED presentation. Of these 279 patients, 1.4 patients may potentially transition to injectable heroin use. Modelling opioid use behaviour in an ED population demonstrated the potential development of unintentional long-term opioid use, and associated harms. Prospective study is required to fully understand trajectories of patients dispensed outpatient therapy from Australian EDs.

Unplanned return presentations of older patients to the emergency department: a root cause analysis

BackgroundIn line with demographic changes, there is an increase in ED presentations and unplanned return presentations by older patients (≥70 years). It is important to know why these patients return to the ED shortly after their initial presentation. Therefore, the aim of this study was to provide insight into the root causes and potential preventability of unplanned return presentations (URP) to the ED within 30 days for older patients.MethodsA prospective observational study was conducted from February 2018 to November 2018 in an academic hospital in Amsterdam. We included 83 patients, aged 70 years and older, with an URP to the ED within 30 days of the initial ED presentation. Patients, GPs and doctors at the ED were interviewed by trained interviewers and basic administrative data were collected in order to conduct a root cause analysis using the PRISMA-method.ResultsOne hundred fifty-one root causes were identified and almost half (49%) of them were disease-related. Fifty-two percent of the patients returned to the ED within 7 days after the initial presentation. In 77% of the patients the URP was related to the initial presentation. Patients judged 17% of the URPs as potentially preventable, while doctors at the ED judged 25% and GPs 23% of the URPs as potentially preventable. In none of the cases, there was an overall agreement from all three perspectives on the judgement that an URP was potentially preventable.ConclusionDisease-related factors were most often identified for an URP and half of the patients returned to the ED within 7 days. The majority of the URPs was judged as not preventable. However, an URP should trigger healthcare workers to focus on the patient’s process of care and their needs and to anticipate on potential progression of disease. Future research should assess whether this may prevent that patients have to return to the ED.

Identifying ‘at-risk’ critically ill patients who present to the emergency department and require intensive care unit admission: A retrospective observational cohort study

BackgroundEmergency department (ED) triage is the process of prioritising patients by medical urgency. Delays in intensive care unit (ICU) admission can adversely affect patients. ObjectivesThis study aimed to identify characteristics associated with ICU admission for patients triaged as Australasian Triage Scale (ATS) 3 but subsequently admitted to the ICU within 24 h of triage. MethodsThis retrospective, observational cohort study was conducted in a public teaching hospital in Queensland, Australia. Patients older than 18 y triaged with an ATS 3 and admitted to the ICU within 24 h of triage or admitted to the ward between January 1, 2012, and December 31, 2012, were included. The demographic and clinical profiles of ICU admissions vs. all other ward admissions for patients triaged an ATS of 3 were compared. Multivariable regression analysis compared characteristics of patients triaged with an ATS of 3 who did and did not require ICU transfer. Descriptive data are reported as n (%) and median and interquartile range (IQR). Regression analysis is reported as adjusted odds ratios (aORs) with 95% confidence intervals (95% CIs). ResultsOf the 27 454 adult ED presentations triaged with an ATS of 3, 22.4% (n = 6138) required hospital admission, comprising 5302 individuals, 2.1% of whom (n = 110) were admitted to the ICU within 24 h of triage. Age- and sex-adjusted predictors of ICU admission for patients triaged with an ATS of 3 included infectious (aOR: 3.7; 95% CI: 2.0–6.9), neurological (aOR: 2.8; 95% CI: 1.6–5.0), and gastrointestinal disorders (aOR: 2.2; 95% CI 1.2–3.5); arriving by ambulance; arriving after hours; or arriving on weekends. Regardless of diagnosis or sex, persons older than 80 y were less likely to be admitted to the ICU (aOR: 0.4; 95% CI: 0.2–0.8). ConclusionsPatients triaged as ATS 3 presenting on weekends or after hours, and those with infectious, gastrointestinal, or neurological conditions warrant careful attention as these factors were associated with higher odds of ICU admission. Ongoing staff education regarding triage and signs of deterioration are important to prevent avoidable outcomes.

Chronic disease management in emergency department patients presenting with dyspnoea.

Guideline recommended treatments for chronic conditions are thought to reduce ED presentations. We used data from 1958 ED patients with dyspnoea to describe medication use in patients with chronic conditions. A total of 1233 (63.5%) patients had one or more of: chronic obstructive pulmonary disease 547 (28%), asthma 454 (23%), atrial fibrillation 368 (19%) or heart failure 401 (21%). Approximately, 70% were prescribed appropriate preventative medication for their chronic condition when they presented to ED with dyspnoea. Prescription of guideline recommended therapies for chronic conditions in patients presenting to the ED in Australasia with acute dyspnoea is similar or higher than reported previously.

Recruitment of adolescents with suicidal ideation in the emergency department: lessons from a randomized controlled pilot trial of a youth suicide prevention intervention

BackgroundEmergency Departments (EDs) are a first point-of-contact for many youth with mental health and suicidality concerns and can serve as an effective recruitment source for randomized controlled trials (RCTs) of mental health interventions. However, recruitment in acute care settings is impeded by several challenges. This pilot RCT of a youth suicide prevention intervention recruited adolescents aged 12 to 17 years presenting to a pediatric hospital ED with suicide related behaviors.MethodsRecruitment barriers were identified during the initial study recruitment period and included: the time of day of ED presentations, challenges inherent to study presentation, engagement and participation during an acute presentation, challenges approaching and enrolling acutely suicidal patients and families, ED environmental factors, and youth and parental concerns regarding the study. We calculated the average recruitment productivity for published trials of adolescent suicide prevention strategies which included the ED as a recruitment site in order to compare our recruitment productivity.ResultsIn response to identified barriers, an enhanced ED-centered recruitment strategy was developed to address low recruitment rate, specifically (i) engaging a wider network of ED and outpatient psychiatry staff (ii) dissemination of study pamphlets across multiple areas of the ED and relevant outpatient clinics. Following implementation of the enhanced recruitment strategy, the pre-post recruitment productivity, a ratio of patients screened to patients randomized, was computed. A total of 120 patients were approached for participation, 89 (74.2%) were screened and 45 (37.5%) were consented for the study from March 2018 to April 2019. The screening to randomization ratio for the study period prior to the introduction of the enhanced recruitment strategies was 3:1, which decreased to 1.8:1 following the implementation of enhanced recruitment strategies. The ratio for the total recruitment period was 2.1:1. This was lower than the average ratio of 3.2:1 for published trials.ConclusionsEDs are feasible sites for participant recruitment in RCTs examining new interventions for acute mental health problems, including suicidality. Engaging multi-disciplinary ED staff to support recruitment for such studies, proactively addressing anticipated concerns, and creating a robust recruitment pathway that includes approach at outpatient appointments can optimize recruitment.Trial registrationClinicalTrials.gov: NCT03488602, retrospectively registered April 4, 2018.

Presentations of stroke and acute myocardial infarction in the first 28 days following the introduction of State of Emergency restrictions for COVID-19.

SOE measures to achieve physical distancing and restrict community contact were not associated with significant changes in time to ED presentation after acute stroke and AMI. Time to initial investigation and life‐saving interventions were also unchanged in the first 28 days. Public health messaging encouraging patients with acute illness or injury to seek medical assessment should be continued.

Nurse practitioner administered point-of-care ultrasound compared with X-ray for children with clinically non-angulated distal forearm fractures in the ED: a diagnostic study

BackgroundPaediatric distal forearm fractures are a common ED presentation. They can be diagnosed with point-of-care ultrasound (POCUS) as an alternative to X-rays. Given that ED nurse practitioners (NPs) are relied...

Impact of COVID-19 State of Emergency restrictions on presentations to two Victorian emergency departments.

To determine if COVID-19 State of Emergency (SOE) restrictions were associated with a reduction in presentations to two urban EDs in Melbourne, Victoria. This retrospective observational study included adult patients presenting to The Alfred and Sandringham Hospital EDs during the first month of stage 2 and 3 SOE restrictions (26 March-25 April 2020). Patients transferred from other hospitals or diagnosed with COVID-19 were excluded. The primary outcome was the average number of presentations per day. Secondary outcomes included the average daily number of presentations for pre-specified subgroups defined by triage category and diagnosis. The independent impact of SOE restrictions, adjusted for underlying trends in attendance, was determined using negative binomial regression and reported as an incident rate ratio (IRR) with a 95% confidence interval (CI). Average daily attendance during the exposure period was 174.7. In the absence of SOE restrictions, 278.8 presentations per day were predicted, a reduction of 37.3% (IRR 0.63, 95% CI 0.59-0.67). Attendance was lower than anticipated for all triage categories (especially category 5 [IRR 0.51, 95% CI 0.44-0.59]) and diagnostic groups (including circulatory problems [IRR 0.62, 95% CI 0.50-0.76] and injury [IRR 0.58, 95% CI 0.53-0.63]). There were fewer than predicted presentations for several sentinel diagnoses, including gastroenteritis (IRR 0.27, 95% CI 0.17-0.42) and renal colic (IRR 0.55, 95% CI 0.33-0.92). SOE restrictions were associated with a significant reduction in ED presentations across a range of triage categories and diagnoses. Public health messaging should emphasise the importance of timely ED attendance for acute illness and injury.

Lunacy in a tertiary emergency department: A 3-year cohort study of the association between moon cycles and occupational violence and aggression.

Occupational violence and aggression (OVA) in the ED is an issue of global concern and increasing incidence. The empirical evidence for the relationship between the lunar cycle and 'lunatics' remains equivocal. The present study aims to examine the association between OVA in ED and the full moon (FM). Data on all presentations were extracted from The Alfred Hospital ED records for consecutive patients over a 3-year period (January 2013-December 2015). The primary outcome of the present study is OVA among patients in the ED. Univariable and multivariable logistic regression were used to determine the association between aspects of the lunar cycle and OVA. There were 184 059 ED presentations during the 3 years, 6234 (3.4%) of which occurred on a FM. There were 1853 episodes of OVA, 57 (3.1%) of which occurred on a FM. OVA among patients presenting to ED was not associated with the FM (adjusted odds ratio [OR] 0.92 [95% confidence interval 0.70-1.20]; P = 0.53). However, the first quarter (FQ) (adjusted OR 1.38 [1.11-1.72]; P < 0.01) and third quarter (TQ) (adjusted OR 1.29 [95% confidence interval 1.03-1.62]; P = 0.03) moons of the lunar cycle were independently associated with OVA. Contrary to traditional beliefs, the FQ and TQ of the lunar cycle but not the FM were associated with OVA. This highlights a relatively unexplored relationship that has previously been overshadowed by the FM in the literature. Prediction models of violence in the ED could consider incorporating the FQ and TQ of the lunar cycle in their models.

A 10-year retrospective study of paediatric emergency department visits for dental conditions in Montreal, Canada.

Trends of paediatric emergency visits (ED) for dental conditions have been broadly discussed; however, little has been published in the Canadian context. To describe the utilization of ED for dental conditions among children and to investigate demographic characteristics influencing its use. A comprehensive review of records of all children aged 1-17years who attended the ED service of a paediatric hospital in Montreal, Canada, for dental conditions over a 10-year period (2004-2013) was completed. Information on the child's principal dental diagnosis, sociodemographic data, and source of referral was obtained. A total of 10905 paediatric ED visits were seen during the study period. Among the children, 54.7% were male and the majority was younger than 6years old. Dental caries constituted the most common reason for ED presentation comprising close to 43% of total visits for a dental complaint. Females, teenagers, and self-referred children were more likely to experience ED visits due to non-traumatic dental conditions. The utilization of ED for dental conditions has increased among pre-school children in the last decade and was mostly due to caries-related dental problems. Effective preventive strategies are needed to improve the oral health condition of children.

An evaluation of the presentation and severity of Australian football injury in children

Introduction Child participation in sport is important for physical, cognitive and psychosocial wellbeing. Australian rules football has high participation, but also carries a high risk of injury due to the contact nature of the sport. This study aimed to evaluate changes in the presentation and hospital admission of paediatric Australian rules football-related injuries, and to compare the severity of these injuries with those from other team ball sports. Materials and methods At an Australian paediatric major trauma service, ED and hospital trauma registry data relating to Australian rules football injury between 2009 and 2015 were obtained. Data from other common team ball sports with a shared field of play were also identified. Results During the study period, there were 10,003 ED presentations, and 1110 admissions resulting from team ball sports. With 4751 ED presentations and 616 admissions, Australian rules football accounted for almost one-third of all sports-related presentations and admissions, and around half of the team ball sports cohort. Compared to other team ball sports patients, Australian rules football-related patients were 40% more likely to be admitted, and nearly twice as likely to be classified as severe injury. Australian rules football players presented with different injury patterns were compared to other team ball sports players; admitted players were significantly more likely to have sustained head or neck injuries, and were more than twice as likely to sustain truncal injury. Conclusions Australian rules football is a common cause of ED presentations and results in substantial morbidity, both overall and when compared with other team ball sports. Australian rules football should remain a focus for ongoing and active research into strategies which reduce injury risk.

Impact of a new specialist alcohol and drug brief intervention service model integrated into the emergency department: An interrupted time series analysis.

To describe and examine the impact of a new specialist drug and alcohol brief intervention team (DABIT) model integrated into the ED on the identification of individuals at risk of future alcohol and other drug (AOD)-related harm. A cost-outcome analysis was conducted to assess the impact on costs per referral. An interrupted time series analysis examined the changes in number of referrals following the implementation of the DABIT model over 2 years (January 2015-December 2016) within a large 436-bed public hospital. The primary outcome of interest was the number of AOD-related referrals per month identified following ED presentations. The independent variables were: time (measured in months), implementation periods (pre-implementation; a transition period of adjustment during which the new DABIT model of care was developed; post-implementation period with a fully operational DABIT model); and the number of full-time equivalent staff per month to account for the increase in labour productivity. In a second time series analysis, the outcome was cost per referral per month. After controlling for changes in labour productivity, the number of referrals was significantly higher following the implementation of the DABIT model when compared to those during the pre-implementation and transition periods. Costs were significantly lower following DABIT implementation resulting in $1096 net cost savings per referral. Integration of a specialist brief intervention AOD model to support ED care may increase uptake of specialist AOD treatment and could be beneficial from an economic efficiency viewpoint.

A Service Evaluation after 4 year’s use of the Virtual Fracture Clinic model by a District General Hospital in the South West of England.

IntroductionA Virtual fracture clinic (VFC) was set up in 2015. An initial patient satisfaction survey demonstrated satisfaction with the service. The purpose of this service evaluation was to re-evaluate the VFC by reporting on patient satisfaction. A small audit was undertaken alongside to examine the time taken from ED presentation to VFC review against the British Orthopaedic Association Society for Trauma (BOAST) guidelines of 72 h. Patients and methodsAll patients discharged from VFC in August 2019 were eligible to take part in the patient satisfaction survey. The Electronic Patients Records System (EPRS) was used to generate data regarding time of review and patient return for follow up appointments. ResultsThe results demonstrated that 88% of patients would recommend the service to friends. More than 80% of patients were satisfied with various elements of the service and 80% of patients are seen within the target time of 72 h. ConclusionsPatients continue to be satisfied with the VFC. There seems to be a greater acceptance of not being reviewed by a doctor. In general, patients were seen within 72 h of ED presentation but more work is needed to streamline the process of tertiary referrals.

Evaluating an emergency department‐based mental health liaison nurse service: A multi‐site translational research project

We aimed to translate and evaluate a model of mental health liaison nursing (MHLN) care that was embedded within EDs. The 12 month mixed-methods translational research study incorporated descriptive data on ED presentations, waiting times for MHLN intervention, time spent in ED and discharge arrangements across three EDs in New South Wales. The study involved an inner-city ED (where the model was first established) and two rural sites. Surveys were conducted on a subset of ED patients (n = 58), and emergency and psychiatry staff (n = 52). Triage category 3 presentations accounted for 49% of the MHLN team workload. Response times and ED length of stay varied between city and rural sites, with rural sites demonstrating prompt response times and reduced ED length of stay. The model was strongly endorsed by patients and staff, with 95% of staff and 85% of patients across the three sites recommending the model be implemented in other emergency settings. The need for adequate resources to maintain designated levels of staffing and sustain this model of care was highlighted. Findings from the present study indicate that a model of ED-based MHLN care developed in a metropolitan setting was successfully translated to two rural sites. However, the model needs to adhere to certain key principles, and be adequately resourced in order to be sustainable and improve outcomes for ED patients and access to community care.

Association between potential primary care emergency service and general practitioner care utilisation in New South Wales.

To examine patterns of potential primary care (PPC) ED presentations and any association between PPC ED presentations and frequency of general practitioner (GP) care utilisation in New South Wales, Australia. Retrospective cross-sectional study of 6 221 762 New South Wales patients who had at least one service in public hospitals, EDs or Medical Benefit Schedule claimable for GP service between 2013/2014 and 2014/2015 is conducted to examine association between PPC ED presentations and GP care utilisation using logistic regression adjusting for comorbidity index and a number of other covariates. Data over 2010/2011 to 2014/2015 are included for analyses of trends and patterns in PPC ED and GP care utilisation. Forty-one percent of the ED presentations were PPC ED presentations over the 5 years 2010/2011 to 2014/2015. Population rates of PPC ED presentations and GP care both increased over the period, with higher PPC ED presentation rates in regional areas, and higher GP care rates in major cities. GP care utilisation was associated with reduced odds for PPC ED presentations, with the adjusted odds ratios ranging from 0.28 for patients with one GP care service to 0.48 for patients with five or more GP care services compared with patients with none. Increased comorbidity index was also associated with increased risk of PPC ED presentations. GP care utilisation was associated with reduced risk for any PPC ED presentations after adjusting for comorbidity index and the other factors.

Patients with mental health conditions in the emergency department: Why so long a wait?

Patients presenting with a mental health disorder (MHD) wait longer in the ED compared to those presenting for other reasons, potentially placing vulnerable patients at further risk for deterioration. The present study identified factors associated with a long ED stay for patients with a MHD. Linked ambulance, emergency, pathology, imaging and admission data for an 18-month period were analysed for ED presentations diagnosed with an MHD ICD-10 at a large teaching hospital. Admissions and discharges were considered separately; a long ED stay was defined as the 90th percentile length of stay. Multivariable generalised linear models were built, identifying predictors of a long ED stay for presentations diagnosed with a MHD. The sample comprised 1163 admissions and 2242 discharges. For admissions, significant predictors for long ED stay were investigations (pathology or imaging tests), a triage score of 1 or 2, arrival out-of-hours (18.00-05.59 hours) and arrival by ambulance. For discharges, significant predictors of a long ED stay were investigations (pathology or imaging tests), arrival out-of-hours, arrival by ambulance and increasing age. Some factors predictive of a long ED stay for patients presenting to the ED and diagnosed with a MHD varied based on their disposition. For admissions, the most urgent presentations were likely to stay longest. Strategies to reduce ED stay for both admissions and discharges should consider addressing modifiable aspects, including the need for certain investigations, and non-modifiable aspects, including the need for further access to after-hours mental health services in hospital and in the community.

Preventing emergency department visits among patients with cancer: a scoping review.

The objective of this scoping review was to examine the effectiveness of supportive care interventions designed to reduce ED visits among patients receiving active cancer treatment. Literature search involving nine electronic databases and grey literature. Inclusion criteria considered studies assessing the impact of any intervention to reduce ED utilization among patients with active cancer. Dichotomous and continuous outcomes were summarized as risk ratio (RR) or mean difference (MD) with 95% confidence intervals (CIs) using a random-effects model, wherever appropriate. A total of 25 studies were included. Interventions identified in these studies comprised the following: routine and symptom-based patient follow-up, oncology outpatient clinics, early symptom detection, comprehensive inpatient management, hospital at home, and patient navigators. Six out of eight studies assessing oncology outpatient clinics reported a decrease in the proportion of patients presenting to the ED. A meta-analysis of three of these studies did not demonstrate reduction in ED utilization (RR 0.78; 95% CI 0.56 to 1.08; I2= 77%) when comparing oncology outpatient clinics with standard care; however, sensitivity analysis supported a decrease in ED visits (RR 0.86; 95% CI 0.74 to 0.99; I2= 47%). Three studies assessing patient follow-up interventions showed no difference in ED utilization (RR 0.69; 95% CI 0.38 to 1.25; I2= 86%). A variety of supportive care interventions designed to mitigate ED presentations by patients receiving active cancer treatment have been developed and evaluated. Limited evidence suggests that an oncology outpatient clinic may be an effective strategy to reduce ED utilization; however, additional high-quality studies are needed.

Predicting hospital admission for older emergency department patients: Insights from machine learning

BackgroundEmergency departments (ED) are a portal of entry into the hospital and are uniquely positioned to influence the health care trajectories of older adults seeking medical attention. Older adults present to the ED with distinct needs and complex medical histories, which can make disposition planning more challenging. Machine learning (ML) approaches have been previously used to inform decision-making surrounding ED disposition in the general population. However, little is known about the performance and utility of ML methods in predicting hospital admission among older ED patients. We applied a series of ML algorithms to predict ED admission in older adults and discuss their clinical and policy implications. Materials and methodsWe analyzed the Canadian data from the interRAI multinational ED study, the largest prospective cohort study of older ED patients to date. The data included 2274 ED patients 75 years of age and older from eight ED sites across Canada between November 2009 and April 2012. Data were extracted from the interRAI ED Contact Assessment, with predictors including a series of geriatric syndromes, functional assessments, and baseline care needs. We applied a total of five ML algorithms. Models were trained, assessed, and analyzed using 10-fold cross-validation. The performance of predictive models was measured using the area under the receiver operating characteristic curve (AUC). We also report the accuracy, sensitivity, and specificity of each model to supplement performance interpretation. ResultsGradient boosted trees was the most accurate model to predict older ED patients who would require hospitalization (AUC = 0.80). The five most informative features include home intravenous therapy, time of ED presentation, a requirement for formal support services, independence in walking, and the presence of an unstable medical condition. ConclusionTo the best of our knowledge, this is the first study to predict hospital admission in older ED patients using a series of geriatric syndromes and functional assessments. We were able to predict hospital admission in older ED patients with good accuracy using the items available in the interRAI ED Contact Assessment. This information can be used to inform decision-making about ED disposition and may expedite admission processes and proactive discharge planning.