Impact of initial high flow nasal cannula flow rates on clinical outcomes in children with bronchiolitis.

PURPOSE: Our aim was to evaluate the impact of the initial high flow nasal cannula (HFNC) flow rate on clinical outcomes in children with bronchiolitis. METHODS: This secondary analysis of retrospective data included children <2 years who required HFNC for bronchiolitis between 10/01/2018-04/20/2019, and following implementation of a revised institutional bronchiolitis pathway between 10/01/2021-04/30/2022. The new pathway recommended weight-based initiation of HFNC at 1.5-2 L/kg/min. We evaluated the effect of low (<1.0 L/kg/min), medium (1-1.5 L/kg/min) and high (>1.5 L/kg/min) HFNC flow rates on need for positive pressure ventilation (PPV), intensive care unit (ICU) transfer, HFNC treatment time, and hospital length of stay (LOS). RESULTS: The majority of the 885 included children had low initial flow rates (low [n = 450, 50.8%], medium [n = 332, 37.5%] and high [n = 103, 11.7%]). There were no significant differences in PPV (high: 7.8% vs. medium: 9.3% vs. low: 8.2%, p = 0.8) or ICU transfers (high: 4.9% vs. medium: 6.0% vs. low: 3.8%, p = 0.3). The low flow group had a significantly longer median HFNC treatment time (High: 29 [18, 45] vs. medium: 29 [16, 50] vs. low: 39 [25, 63], p < .001) and hospital LOS (High: 41 [27, 59] vs. medium: 42 [29, 66] vs. low: 50 (39, 75), p < .001). Logistic and linear regression models did not demonstrate any associations between HFNC flow rates and PPV or hospital LOS. CONCLUSIONS: Initial HFNC flow rates were not associated with significant changes in clinical outcomes in children in children with bronchiolitis.

Outcomes of Ultrasound Guided Peripheral Intravenous Catheters Placed in the Emergency Department and Factors Associated with Survival.

Background: Patients with difficult peripheral intravenous (IV) access are common in emergency departments (EDs). Ultrasound-guided peripheral intravenous catheters (USIVs) are frequently used in this population; however, information regarding the effect of patient and IV characteristics on the dwell time (DT) and survival probability (SP) of USIVs is limited. Objective: Our study aimed to evaluate for associations between patient or IV characteristics and the DT and SP of USIVs. Methods: Retrospective analysis was performed on a database from an ED nurse (RN) USIV training program at an urban, academic hospital. Patients over 18 years with an USIV placed during the study period were included. Subject demographics, history, IV characteristics, insertion, and removal times were collected. Data were analyzed using descriptive statistics and univariable and multivariable Cox regression. USIV survival times for variates of interest were estimated using Kaplan-Meier curves for three censoring points. Results: The final analysis cohort was 388 patients. Mean age was 56.6 years, 66.5% were female, mean BMI was 29.9 kg/m2, and 42.5% were obese (BMI ≥30). Median DT was 40.3 hours in admitted patients (N=340). SP for USIVs at 96 hours was 87.8%. A total of 21 of 340 (6.2%) USIVs failed. USIV location conferred a difference on DT in obese patients when dichotomized into upper arm versus antecubital fossa and forearm together (38.6 hours vs 44.6 hours, p=0.03). No factors were associated with a difference in USIV SP. Conclusion: Median USIV DT of 40.3 hours for admitted patients was higher than in previous studies. Only 7% of USIVs in our study failed. Overall, catheters survived longer than expected.

Comparison of pulmonary congestion severity using artificial intelligence-assisted scoring versus clinical experts: A secondary analysis of BLUSHED-AHF.

AIM: Acute decompensated heart failure (ADHF) is the leading cause of cardiovascular hospitalizations in the United States. Detecting B-lines through lung ultrasound (LUS) can enhance clinicians' prognostic and diagnostic capabilities. Artificial intelligence/machine learning (AI/ML)-based automated guidance systems may allow novice users to apply LUS to clinical care. We investigated whether an AI/ML automated LUS congestion score correlates with expert's interpretations of B-line quantification from an external patient dataset. METHODS AND RESULTS: This was a secondary analysis from the BLUSHED-AHF study which investigated the effect of LUS-guided therapy on patients with ADHF. In BLUSHED-AHF, LUS was performed and B-lines were quantified by ultrasound operators. Two experts then separately quantified the number of B-lines per ultrasound video clip recorded. Here, an AI/ML-based lung congestion score (LCS) was calculated for all LUS clips from BLUSHED-AHF. Spearman correlation was computed between LCS and counts from each of the original three raters. A total of 3858 LUS clips were analysed on 130 patients. The LCS demonstrated good agreement with the two experts' B-line quantification score (r = 0.894, 0.882). Both experts' B-line quantification scores had significantly better agreement with the LCS than they did with the ultrasound operator's score (p < 0.005, p < 0.001). CONCLUSION: Artificial intelligence/machine learning-based LCS correlated with expert-level B-line quantification. Future studies are needed to determine whether automated tools may assist novice users in LUS interpretation.

Comparison of Trauma and Burn Evaluations in a Pediatric Emergency Department During Pre, Early and Late COVID-19 Pandemic.

BACKGROUND: Pediatric trauma epidemiology altered during early COVID-19 pandemic period but the impact of the ongoing pandemic is unknown. OBJECTIVES: To compare pediatric trauma epidemiology between the pre, early and late pandemic periods and to evaluate the association of race and ethnicity on injury severity during the pandemic. METHODS: We performed a retrospective study of trauma consults for an injury/burn in children ≤16 years between January 1, 2019 and December 31, 2021. Study period was categorized into pre (January 1, 2019-February 28, 2020), early (March 1, 2020-December 31, 2020), and late (January 1, 2021-December 31, 2021) pandemic. Demographics, etiology, injury/burn severity, interventions and outcomes were noted. RESULTS: A total of 4940 patients underwent trauma evaluation. Compared to pre-pandemic, trauma evaluations for injuries and burns increased during both the early (RR: 2.13, 95% CI: 1.6-2.82 and RR: 2.24, 95% CI: 1.39-3.63, respectively) and late pandemic periods (RR: 1.42, 95% CI: 1.09-1.86 and RR: 2.44, 95% CI: 1.55-3.83, respectively). Severe injuries, hospital admissions, operations and death were higher in the early pandemic but reverted to pre-pandemic levels during late pandemic. Non-Hispanic Blacks had an approximately 40% increase in mean ISS during both pandemic periods though they had lower odds of severe injury during both pandemic periods. CONCLUSIONS: Trauma evaluations for injuries and burns increased during the pandemic periods. There was a significant association of race and ethnicity with injury severity which varied with pandemic periods. LEVEL OF EVIDENCE: Retrospective comparative study, Level III.

Primer on Logistic Regression for Emergency Care Researchers.

BACKGROUND: Logistic regression plays a fundamental role in the production of decision rules, risk assessment, and in establishing cause and effect relationships. This primer is aimed at novice researchers with minimal statistical expertise. OBJECTIVE: Introduce the logit equation and provide a hands-on example to facilitate understanding of its benefits and limitations. DISCUSSION: This primer reviews the mathematical basis of a logit equation by comparing and contrasting it with the simple straight-line (linear) equation. After gaining an understanding of the meaning of beta coefficients, readers are encouraged to download a free statistical program and database to produce a logistic regression analysis. Using this example, the narrative then discusses commonly used methods to describe model fitness, including the C-statistic, chi square, Akaike and Bayesian Information Criteria, McFadden's pseudo R2, and the Hosmer-Lemeshow test. The authors provide a how-to discussion for variable selection and estimate of sample size. However, logistic regression alone can seldom establish causal inference without further steps to explore the often complex relationship amongst variables and outcomes, such as with the use of a directed acyclic graphs. We present key elements that generally should be considered when appraising an article that uses logistic regression. This primer provides a basic understanding of the theory, hands-on construction, model analysis, and limitations of logistic regression in emergency care research. CONCLUSIONS: Logistic regression can provide information about the association of independent variables with important clinical outcomes, which can be the first step to show predictiveness or causation of variables on the outcomes of interest. © 2022 Elsevier Inc.

Characteristics of Adult Rapid Response Events in a Freestanding Children's Hospital.

OBJECTIVES: To describe nonhospitalized adult rapid response events (adult RREs) in a freestanding children's hospital and examine the relationship between various demographic and clinical factors with the final patient disposition. METHODS: We retrospectively reviewed records for nonhospitalized patients ≥18 years of age from events that occurred in a freestanding pediatric hospital between January 2011 through December 2020. We examined the relationship between adult RREs and patient demographic information, medical history, interventions, and patient disposition following an adult RRE. RESULTS: Four hundred twenty-nine events met inclusion criteria for analysis. Most events (69%) occurred in females, 49% of events occurred in family members of patients, and 47% occurred on inpatient floor and ICU areas. The most common presenting complaint was syncope or dizziness (36%). Delivery of bad news or grief response was associated with 14% of adult RREs. Overall, 46% (n = 196) of patients were transferred to the pediatric emergency department (ED). Patients requiring acute intervention or with cardiac or neurologic past medical histories were more likely to be transferred to the pediatric ED. Acute advanced cardiac life support interventions were infrequent but, of the patients taken to the pediatric ED, 1 died, and 3 were admitted to the ICU. CONCLUSIONS: Adult RREs are common in freestanding children's hospitals and, although rare, some patients required critical care. Expertise in adult critical care management should be available to the rapid response team and additional training for the pediatric rapid response team in caring for adult nonpatients may be warranted.

Increasing illness severity is associated with global myocardial dysfunction in the first 24 hours of sepsis admission.

BACKGROUND: Septic cardiomyopathy was recognized more than 30 years ago, but the early phase remains uncharacterized as no existing studies captured patients at the time of Emergency Department (ED) presentation, prior to resuscitation. Therapeutic interventions alter cardiac function, thereby distorting the relationship with disease severity and outcomes. The goal of this study was to assess the impact of illness severity on cardiac function during the first 24 h of sepsis admission. METHODS: This was a pre-planned secondary analysis of a prospective observational study of adults presenting to the ED with suspected sepsis (treatment for infection plus either lactate > 2 mmol/liter or systolic blood pressure < 90 mm/Hg) who received < 1L IV fluid before enrollment. Patients had 3 echocardiograms performed (presentation, 3, and 24 h). The primary outcome was the effect of increasing sepsis illness severity, defined by ED Sequential Organ Failure Assessment (SOFA) score, on parameters of cardiac function, assessed using linear mixed-effects models. The secondary goal was to determine whether cardiac function differed between survivors and non-survivors, also using mixed-effects models. RESULTS: We enrolled 73 patients with a mean age of 60 (SD 16.1) years and in-hospital mortality of 23%. For the primary analysis, we found that increasing ED SOFA score was associated with worse cardiac function over the first 24 h across all assessed parameters of left-ventricular systolic and diastolic function as well as right-ventricular systolic function. While baseline strain and E/e' were better in survivors, in the mixed models analysis, the trajectory of Global Longitudinal Strain and septal E/e' over the first 24 h of illness differed between survivors and non-survivors, with improved function at 24 h in non-survivors. CONCLUSIONS: In the first study to capture patients prior to the initiation of resuscitation, we found a direct relationship between sepsis severity and global myocardial dysfunction. Future studies are needed to confirm these results, to identify myocardial depressants, and to investigate the link with adverse outcomes so that therapeutic interventions can be developed.

Early echocardiographic assessment of cardiac function may be prognostically informative in unresuscitated patients with sepsis: A prospective observational study.

PURPOSE: The goal of this study was to explore the association cardiac function at Emergency Department (ED) presentation prior to the initiation of resuscitation, and its change at 3-hours, with adverse outcomes in patients with sepsis. METHODS: This was a prospective observational study of patients presenting to an urban ED with suspected sepsis. Patients had a point-of-care echocardiogram performed prior to initiation of resuscitation and again 3 hours later. Left-ventricular (LV) parameters recorded included e', and E/e', and ejection fraction (EF); right-ventricular (RV) function was evaluated using tricuspid annular plane systolic excursion (TAPSE). Logistic and generalized linear regression were used to assess the association of echocardiographic parameters and ≥ 2-point increase in SOFA score at 24 hours (primary outcome) and 24-hours SOFA score and in-hospital mortality (secondary outcomes). RESULTS: For ΔSOFA ≥ 2 and 24-hour SOFA score, declining LVEF was associated with better outcomes in patients with greater baseline SOFA scores, but worse outcomes in patients with lower baseline scores. A similar relationship was found for ΔTAPSE at 3 hours. Reduced LVEF at presentation was associated with increased mortality after adjusting for ED SOFA score (odds-ratio (OR) 0.76 (CI 0.60-0.96). No relationship between diastolic parameters and outcomes was found. IVF administration was similar across ΔLVEF/TAPSE sub-groups. CONCLUSIONS: Our results suggest that early change in LV and RV systolic function are independently prognostic of sepsis illness severity at 24-hours. Further study is needed to determine if this information can be used to guide treatment and improve outcomes.

Success of implementation of a systemwide point-of-care ultrasound privileging program for emergency medicine faculty.

Objectives: Point-of-care ultrasound (POCUS) is widely used in the emergency department (ED). Not all practicing emergency physicians received POCUS training during residency, leaving a training gap that is reflected in POCUS privileging. The purpose of this study was to evaluate the success of meeting privileging criteria as well as associated factors, following implementation of a basic POCUS training and privileging program within a large emergency medicine department. Methods: We implemented a POCUS training and privileging program, based on national guidelines, for faculty physicians who worked at one of the following EDs staffed by the same emergency medicine department: a pediatric tertiary site, two tertiary academic sites, and seven community sites. POCUS examinations included aorta, cardiac, first-trimester obstetrics (OB), and extended focused assessment with sonography in trauma. Pediatric emergency medicine faculty were taught soft tissue and thoracic US instead of aorta and OB. Completion of the program required 16 h of didactics, ≥25 quality-assured US examinations by examination type, and passing a series of knowledge-based examinations. Descriptive statistics were calculated. Associations between physician characteristics and successfully becoming privileged in POCUS were modeled using Firth's logistic regression. Results: A total of 176 faculty physicians were eligible. A total of 145 (82.4%) achieved basic POCUS privileging during the study period. Different pathways were used including 86 (48.9%) practice-based, nine (5.1%) fellowship-based, and 82 (46.9%) residency-based. POCUS privileging was lower for those working in a community versus academic setting (odds ratio 0.3, 95% confidence interval 0.1-0.9). A greater number of scans completed prior to the privileging program was associated with greater success. Conclusions: Implementation of a POCUS training and privileging program can be successful in a large emergency medicine department that staffs hospitals in a large-scale health care system composed of both academic and community sites. Faculty physicians with at least some prior exposure to POCUS were more successful.

The Parasternal Long Axis View in Isolation: Is it Good Enough?

BACKGROUND: Focused cardiac ultrasound (FOCUS) is a vital tool to evaluate patients at the bedside, but its use can be limited by patient habitus, sonographer skill, and time to perform the examination. OBJECTIVE: Our primary goal was to determine the diagnostic accuracy of the parasternal long axis (PSLA) view in isolation for identifying pericardial effusion, left ventricular (LV) dysfunction, and right ventricular (RV) dilatation compared with a four-view FOCUS examination. METHODS: This was a retrospective study looking at FOCUS images. Examinations were blinded and randomized for review by point-of-care ultrasound faculty. The primary objective was measured by comparing ultrasound findings on PSLA view in isolation with findings on a full four-view FOCUS examination, which served as the criterion standard. Sensitivity and specificity were calculated. RESULTS: Of 100 FOCUS examinations; 36% were normal, 16% had a pericardial effusion, 41% had an LV ejection fraction < 50%, and 7% had RV dilatation. Sensitivity and specificity for identifying pericardial effusion, LV dysfunction, and RV dilatation were 81% (confidence interval [CI] 0.54-0.95) and 98% (95% CI 0.91-0.99), 100% (95% CI 0.88-1) and 91% (95% CI 0.80-0.97), and 71% (95% CI 0.30-0.94) and 99% (95% CI 0.93-1), respectively. All moderate to large effusions were identified correctly. Overall, there were only four clinically significant disagreements between PSLA alone and the four-view interpretations. CONCLUSIONS: In isolation, the PSLA view was highly sensitive and specific for identifying LV ejection fraction and moderate to large pericardial effusions. It was highly specific for identifying RV dilatation, but had only moderate sensitivity.

Do septic patients with reduced left ventricular ejection fraction require a low-volume resuscitative strategy?

BACKGROUND: Many clinicians are wary of administering 30 cc/kg of intravenous fluid (IVF) to septic patients with reduced left-ventricular ejection fraction (rLVEF), fearing volume overload. Prior studies have used history of heart failure, rather than LVEF measured at presentation, thereby potentially distorting the relationship between rLVEF, IVF, and adverse outcomes. Our goal was to assess the relationship between IVF volume and outcomes in patients with, versus without, rLVEF. METHODS: This was a prospective observational study performed at an urban Emergency Department (ED). Included patients were adults with suspected sepsis, defined as being treated for infection plus either systolic blood pressure <90 mm/Hg or lactate >2 mmol/L. All patients had LVEF assessed by ED echocardiogram, prior to receipt of >1 l IVF. MEASUREMENTS AND MAIN RESULTS: We enrolled 73 patients, of whom 33 had rLVEF, defined as <40%. Patients with rLVEF were older, had greater initial lactate, more ICU admission, and more vasopressor use. IVF volume was similar between LVEF groups at 3-h (2.2 (IQR 0.8) vs 2.0 (IQR 2.4) liters) while patients with rLVEF were more likely to achieve 30 cc/kg (61% (CI 44-75) vs 45% (CI 31-60). In the reduced versus not-reduced LVEF groups, hospital days, ICU days, and ventilator days were similar: 8 (IQR 7) vs 6.5 (8.5) days, 7 (IQR 7) vs 5 (4) days, and 4 (IQR 8) vs. 5 (10) days, respectively. CONCLUSIONS: Septic patients with rLVEF at presentation received similar volume of IVF as those without rLVEF, without an increase in adverse outcomes attributable to volume overload. While validation is needed, our results suggest that limiting IVF administration in the setting of rLVEF is not necessary.

Factors Associated With Voluntary Refusal of Emergency Medical System Transport for Emergency Care in Detroit During the Early Phase of the COVID-19 Pandemic.

Importance: Emergency department (ED) and emergency medical services (EMS) volumes decreased during the COVID-19 pandemic, but the amount attributable to voluntary refusal vs effects of the pandemic and public health restrictions is unknown. Objective: To examine the factors associated with EMS refusal in relation to COVID-19 cases, public health interventions, EMS responses, and prehospital deaths. Design, Setting, and Participants: A retrospective cohort study was conducted in Detroit, Michigan, from March 1 to June 30, 2020. Emergency medical services responses geocoded to Census tracts were analyzed by individuals' age, sex, date, and community resilience using the Centers for Disease Control and Prevention Social Vulnerability Index. Response counts were adjusted with Poisson regression, and odds of refusals and deaths were adjusted by logistic regression. Exposures: A COVID-19 outbreak characterized by a peak in local COVID-19 incidence and the strictest stay-at-home orders to date, followed by a nadir in incidence and broadly lifted restrictions. Main Outcomes and Measures: Multivariable-adjusted difference in 2020 vs 2019 responses by incidence rate and refusals or deaths by odds. The Social Vulnerability Index was used to capture community social determinants of health as a risk factor for death or refusal. The index contains 4 domain subscores; possible overall score is 0 to 15, with higher scores indicating greater vulnerability. Results: A total of 80 487 EMS responses with intended ED transport, 2059 prehospital deaths, and 16 064 refusals (62 636 completed EMS to ED transports) from 334 Census tracts were noted during the study period. Of the cohort analyzed, 38 621 were women (48%); mean (SD) age was 49.0 (21.4) years, and mean (SD) Social Vulnerability Index score was 9.6 (1.6). Tracts with the highest per-population EMS transport refusal rates were characterized by higher unemployment, minority race/ethnicity, single-parent households, poverty, disability, lack of vehicle access, and overall Social Vulnerability Index score (9.6 vs 9.0, P = .002). At peak COVID-19 incidence and maximal stay-at-home orders, there were higher total responses (adjusted incident rate ratio [aIRR], 1.07; 1.03-1.12), odds of deaths (adjusted odds ratio [aOR], 1.60; 95% CI, 1.20-2.12), and refusals (aOR, 2.33; 95% CI, 2.09-2.60) but fewer completed ED transports (aIRR, 0.82; 95% CI, 0.78-0.86). With public health restrictions lifted and the nadir of COVID-19 cases, responses (aIRR, 1.01; 0.97-1.05) and deaths (aOR, 1.07; 95% CI, 0.81-1.41) returned to 2019 baselines, but differences in refusals (aOR, 1.27; 95% CI, 1.14-1.41) and completed transports (aIRR, 0.95; 95% CI, 0.90-0.99) remained. Multivariable-adjusted 2020 refusal was associated with female sex (aOR, 2.71; 95% CI, 2.43-3.03 in 2020 at the peak; aOR 1.47; 95% CI, 1.32-1.64 at the nadir). Conclusions and Relevance: In this cohort study, EMS transport refusals increased with the COVID-19 outbreak's peak and remained elevated despite receding public health restrictions, COVID-19 incidence, total EMS responses, and prehospital deaths. Voluntary refusal was associated with decreased EMS transports to EDs, disproportionately so among women and vulnerable communities.

LIMEs and LEMONs: Critically Examining the Effect of a Blog Post on Junior Faculty Learners.

INTRODUCTION: The usage of asynchronous resources such as blogs and podcasts is pervasive in academic medicine, despite little understanding of their actual effect on learner knowledge acquisition. This study sought to examine the objective effect of a blog post on knowledge acquisition and application among junior faculty in emergency medicine (EM) via randomized controlled study. METHODS: All accredited EM residency programs in the United States and Canada were contacted to identify assistant and associate program directors and medical education fellows for recruitment into this study. Upon enrollment, participants were randomized as to whether they received access to a supplemental blog post prior to listening to a podcast episode. After listening to the podcast episode, all participants completed an assessment that included a test of knowledge application and knowledge acquisition; demographic information was also obtained. RESULTS: Ultimately, 103 participants completed the study; the study closed for enrollment in July 2019. Data were nonnormally distributed and groups were compared using the Wilcoxon rank-sum test. There were no significant differences between the demographics of the two groups nor was there a significant difference in knowledge between the two groups. CONCLUSION: The addition of a supplementary blog post did not increase junior faculty knowledge of a podcast episode.

B-line quantification: comparing learners novice to lung ultrasound assisted by machine artificial intelligence technology to expert review.

BACKGROUND: The goal of this study was to assess the ability of machine artificial intelligence (AI) to quantitatively assess lung ultrasound (LUS) B-line presence using images obtained by learners novice to LUS in patients with acute heart failure (AHF), compared to expert interpretation. METHODS: This was a prospective, multicenter observational study conducted at two urban academic institutions. Learners novice to LUS completed a 30-min training session on lung image acquisition which included lecture and hands-on patient scanning. Learners independently acquired images on patients with suspected AHF. Automatic B-line quantification was obtained offline after completion of the study. Machine AI counted the maximum number of B-lines visualized during a clip. The criterion standard for B-line counts was semi-quantitative analysis by a blinded point-of-care LUS expert reviewer. Image quality was blindly determined by an expert reviewer. A second expert reviewer blindly determined B-line counts and image quality. Intraclass correlation was used to determine agreement between machine AI and expert, and expert to expert. RESULTS: Fifty-one novice learners completed 87 scans on 29 patients. We analyzed data from 611 lung zones. The overall intraclass correlation for agreement between novice learner images post-processed with AI technology and expert review was 0.56 (confidence interval [CI] 0.51-0.62), and 0.82 (CI 0.73-0.91) between experts. Median image quality was 4 (on a 5-point scale), and correlation between experts for quality assessment was 0.65 (CI 0.48-0.82). CONCLUSION: After a short training session, novice learners were able to obtain high-quality images. When the AI deep learning algorithm was applied to those images, it quantified B-lines with moderate-to-fair correlation as compared to semi-quantitative analysis by expert review. This data shows promise, but further development is needed before widespread clinical use.

The Use of Handheld Ultrasound Devices in Emergency Medicine.

PURPOSE OF REVIEW: Ultraportable handheld ultrasound (HHU) devices are being rapidly adopted by emergency medicine (EM) physicians. Though knowledge of the breadth of their utility and functionality is still limited compared to cart-based systems, these machines are becoming more common due to ease-of-use, extreme affordability, and improving technology. RECENT FINDINGS: Images obtained with HHU are comparable to those obtained with traditional machines but create unique issues regarding billing and data management. HHU devices are increasingly used successfully to augment the education of practitioners-in-training, by emergency physicians in austere environments, and in the burgeoning fields of "tele-ultrasound" and augmented reality scanning. SUMMARY: This review seeks to describe the current state of use of HHU devices in the emergency department (ED) including device overview, institutional concerns, unique areas of use, recent literature since their adoption into clinical EM, and their future potential.

Risk factors associated with hospital admission in COVID-19 patients initially admitted to an observation unit.

BACKGROUND: No set guidelines to guide disposition decisions from the emergency department (ED) in patients with COVID-19 exist. Our goal was to determine characteristics that identify patients at high risk for adverse outcomes who may need admission to the hospital instead of an observation unit. METHODS: We retrospectively enrolled 116 adult patients with COVID-19 admitted to an ED observation unit. We included patients with bilateral infiltrates on chest imaging, COVID-19 testing performed, and/or COVID-19 suspected as the primary diagnosis. The primary outcome was hospital admission. We assessed risk factors associated with this outcome using univariate and multivariable logistic regression. RESULTS: Of 116 patients, 33 or 28% (95% confidence interval [CI] 20-37%) required admission from the observation unit. On multivariable logistic regression analysis, we found that hypoxia defined as room-air oxygen saturation < 95% (OR 3.11, CI 1.23-7.88) and bilateral infiltrates on chest radiography (OR 5.57, CI 1.66-18.96) were independently associated with hospital admission, after adjusting for age. Two three-factor composite predictor models, age > 48 years, bilateral infiltrates, hypoxia, and Hispanic race, bilateral infiltrates, hypoxia yield an OR for admission of 4.99 (CI 1.50-16.65) with an AUC of 0.59 (CI 0.51-0.67) and 6.78 (CI 2.11-21.85) with an AUC of 0.62 (CI 0.54-0.71), respectively. CONCLUSIONS: Over 1/4 of suspected COVID-19 patients admitted to an ED observation unit ultimately required admission to the hospital. Risk factors associated with admission include hypoxia, bilateral infiltrates on chest radiography, or the combination of these two factors plus either age > 48 years or Hispanic race.