The effect of increasing emergency department observation volumes on downstream admission rates.

Rising length of stay and inpatient boarding in emergency departments have directly affected patient satisfaction and nearly all provider-to-patient care metrics. Prior studies suggest that ED observation has significant clinical and financial benefits including decreasing hospitalization and length of stay. ED observation is one method long employed to shorten ED length of stay and to free up inpatient beds, yet many patients continue to be admitted to the hospital with an average hospital length of stay of only one day. The objectives of this study were to evaluate whether vigorous tracking and provider reviews of one day hospital admits affected the utilization of ED observation and whether this correlated with significant change in rates of admission from observation status. Between September 2020 and May 2021, in a tertiary care hospital with an annual ED volume of 55,0000, chart reviews of 24-h inpatient discharges were initiated by two senior EM faculty to determine perceived suitability for ED observation. Non-punitive email reviews were then initiated with ED attending providers in order to encourage evaluation of whether these patients would have benefitted from being placed into observation. We then analyzed ED observation patient volumes and subsequent admission rates to the hospital from ED observation and compared these numbers to baseline ED observation volume and admission rates between September 2018 and May 2019. A total of 1448 reviews were conducted on 24-h discharges which correlated with an increase in utilization of ED observation from 11.77% (95% CI [11.62, 12.31]) of total ED volume in our control period to 14.21% (95% CI [13.84, 14.58]) during the study period. We found that the overall admission rate from ED observation increased from 20.12% (95% CI [18.97, 21.26]) baseline to 23.80% (95% CI [22.60, 25.00]) during the same time periods. Our data suggest that increasing the total number of patients placed into observation by 21% correlated with a relative increase in admission rates from ED observation by 18%. This would suggest that our efforts to potentially include more patients into our observation program led to a significant increase in subsequent admission rates. There is likely a balance that must be struck between under- and over-utilization of ED observation, and expanding ED observation may be an effective solution to hospital boarding and ED overcrowding.

Can Asking Emergency Physicians Whether or Not They Would Have Done Something Differently (WYHDSD) be a Useful Screening Tool to Identify Emergency Department Error?

BACKGROUND: Error in emergency medicine remains common and difficult to identify. OBJECTIVE: To evaluate if questioning emergency physician reviewers as to whether or not they would have done something differently (Would you have done something differently? [WYHDSD]) can be a useful marker to identify error. METHODS: Prospective data were collected on all patients presenting to an academic emergency department (ED) between 2017 and 2021. All cases who met the following criteria were identified: 1) returned to ED within 72 h and admitted; 2) transferred to intensive care unit from floor within 24 h of admission; 3) expired within 24 h of arrival; or 4) patient or provider complaint. Cases were randomly assigned to emergency physicians and reviewed using an electronic tool to assess for error and adverse events. Reviewers were then mandated to answer WYHDSD in the management of the case. RESULTS: During the study period, 6672 cases were reviewed. Of the 5857 cases where reviewers would not have done something differently, 5847 cases were found to have no error. The question WYHDSD had a sensitivity of 97.4% in predicting error and a negative predictive value of 99.8%. CONCLUSION: There was a significantly higher rate of near misses, adverse events, and errors attributable to an adverse event in cases where the reviewer would have done something differently (WHDSD) compared with cases where they would not. Therefore, asking reviewers if they WHDSD could potentially be used as a marker to identify error and improve patient care in the ED.

Assessing resident and attending error and adverse events in the emergency department.

BACKGROUND: There is a paucity of data looking at resident error or contrasting errors and adverse events among residents and attendings. This type of data could be vital in developing and enhancing educational curricula OBJECTIVES: Using an integrated, readily accessible electronic error reporting system the objective of this study is to compare the frequency and types of error and adverse events attributed to emergency medicine residents with those attributed to emergency medicine attendings. METHODS: Individual events were classified into errors and/or adverse events, and were attributed to one of three groups-residents only, attendings only, or both (if the event had both resident and attending involvement). Error and adverse events were also classified into five different categories of events-systems, documentation, diagnostic, procedural and treatment. The proportion of error events were compared between the residents only and the attendings only group using a one-sample test of proportions. Categorical variables were compared using Fisher's exact test. RESULTS: Of a total of 115 observed events over the 11-month data collection period, 96 (83.4%) were errors. A majority of these errors, 40 (41.7%), were attributed to both residents and attendings, 20 (20.8%) were attributed to residents only, and 36 (37.5%) were attributed to attendings only. Of the 19 adverse events, 14 (73.7%) were attributed to both residents and attendings, and 5 (26.3%) adverse events were attributed to attendings only. No adverse events were attributed solely to residents (Table 1). Excluding events attributed to both residents and attendings, there was a significant difference between the proportion of errors attributed to attendings only (64.3%, CI: 50.6, 76.0), and residents only (35.7%, CI: 24.0, 49.0), p = 0.03. (Table 2). There was no significant difference between the residents only and the attendings only group in the distribution of errors and adverse events (Fisher's exact, p = 0.162). (Table 2). There was no statistically significant difference between the two groups in errors that did not result in adverse events and the rate of errors proceeding to adverse events (Fisher's exact, p = 0.15). (Table 3). There was no statistically significant difference between the two groups in the distribution of the types of errors and adverse events (Fisher's exact, p = 0.09). Treatment related errors were the most common error types, for both the attending and the resident groups. CONCLUSIONS: Resident error, somewhat expectedly, is most commonly related to treatment interventions, and rarely is due to an individual resident mistake. Resident error instead seems to reflect concomitant error on the part of the attending. Error, in general as well as adverse events, are more likely to be attributed to an attending alone rather than to a resident.

Errors in Decisionmaking in Emergency Medicine: The Case of the Landscaper's Back and Root Cause Analysis.

Root cause analysis is often suggested as a means of conducting quality assurance, but few physicians are familiar with the actual process. We describe a detailed approach to conducting root cause analysis, with an illustrative case to explain the technique. By studying how root cause analysis is applied to the case of a missed epidural abscess, the reader will see how the process reveals systems improvements that reduce the risk that such a miss will happen again. Following this process will be helpful in using root cause analysis to fix not just individuals' issues but also but systemwide quality assurance issues to improve patient care.

A Tale of Two Cities: Applying the Boston Syncope Criteria to Jerusalem.

BACKGROUND: Syncope is a common reason for emergency department (ED) visits; however, the decision to admit or discharge patients after a syncopal episode remains challenging for emergency physicians. Decision rules such as the Boston Syncope Criteria have been developed in an attempt to aid clinicians in identifying high-risk patients as well as those who can be safely discharged, but applying these rules to different populations remains unclear. OBJECTIVES: To determine whether the Boston Syncope Criteria are valid for emergency department patients in Israel. METHODS: This retrospective cohort convenience sample included patients who visited a tertiary care hospital in Jerusalem from August 2018 to July 2019 with a primary diagnosis of syncope. Thirty-day follow-up was performed using a national health system database. The Boston Syncope Criteria were retrospectively applied to each patient to determine whether they were at high risk for an adverse outcome or critical intervention, versus low risk and could be discharged. RESULTS: A total of 198 patients fulfilled the inclusion criteria and completed follow-up. Of these, 21 patients had either an adverse outcome or critical intervention. The rule detected 20/21 with a sensitivity of 95%, a specificity of 66%, and a negative predictive value of 99. CONCLUSIONS: The Boston Syncope Criteria may be useful for physicians in other locations throughout the world to discharge low-risk syncope patients as well as identify those at risk of complications.

Year-to-year trends in emergency medicine morbidity and mortality cases.

BACKGROUND: Morbidity and Mortality (M&M) rounds are peer review conferences during which cases with adverse outcomes and difficult management decisions are presented. Their primary objective is to learn from complications and errors, modify behavior and judgment based on previous experiences, and prevent repetition of errors leading to complications. The objective of this study was to determine if M&M conferences can reduce repetitive error making demonstrated by a shift of the incidence of cases presented at M&M by chief complaint (CC) and experience of attendings. METHODS: All M&M cases from 1/1/2014-12/31/2017 derived from an urban, tertiary referral Emergency Department were reviewed and grouped into 12 different CC categories and by attending years of experience (1-4, 5-9 and 10+). Number and percent of M&M cases by CC and years of attending experience were calculated by year and a chi-squared analysis was performed. RESULTS: 350 M&M cases were presented over the four-year study period. There was a significant difference between CC categories from year-to-year (p < 0.001). Attendings with 1-4 years of experience had the majority of cases (46.3%), while those with 5-9 years had the fewest total cases (15.1%, p < 0.001). CONCLUSIONS: There was a persistent significant difference across CC categories of M&M cases from year-to-year, with down-trending and up-trending of specific CCs suggesting that M&M presentation may prevent repetitive errors. Newer attendings show increased rates of M&M cases relative to more experienced attendings. There may be a distinctive educational benefit of participation at M&M for attendings with fewer than five years of clinical experience.

Neural networks as a tool to predict syncope risk in the Emergency Department.

AIMS: There is no universally accepted tool for the risk stratification of syncope patients in the Emergency Department. The aim of this study was to investigate the short-term predictive accuracy of an artificial neural network (ANN) in stratifying the risk in this patient group. METHODS AND RESULTS: We analysed individual level data from three prospective studies, with a cumulative sample size of 1844 subjects. Each dataset was reanalysed to reduce the heterogeneity among studies defining abnormal electrocardiogram (ECG) and serious outcomes according to a previous consensus. Ten variables from patient history, ECG, and the circumstances of syncope were used to train and test the neural network. Given the exploratory nature of this work, we adopted two approaches to train and validate the tool. One approach used 4/5 of the data for the training set and 1/5 for the validation set, and the other approach used 9/10 for the training set and 1/10 for the validation set. The sensitivity, specificity, and area under the receiver operating characteristic curve of ANNs in identifying short-term adverse events after syncope were 95% [95% confidence interval (CI) 80-98%], 67% (95% CI 62-72%), 0.69 with the 1/5 approach and 100% (95% CI 84-100%), 79% (95% CI 72-85%), 0.78 with the 1/10 approach. CONCLUSION: The results of our study suggest that ANNs are effective in predicting the short-term risk of patients with syncope. Prospective studies are needed in order to compare ANNs' predictive capability with existing rules and clinical judgment.

Hospital ward transfer to intensive care unit as a quality marker in emergency medicine.

BACKGROUND: Minimizing and preventing adverse events and medical errors in the emergency department (ED) is an ongoing area of quality improvement. Identifying these events remains challenging. OBJECTIVE: To investigate the utility of tracking patients transferred to the ICU within 24h of admission from the ED as a marker of preventable errors and adverse events. METHODS: From November 2011 through June 2016, we prospectively collected data for all patients presenting to an urban, tertiary care academic ED. We utilized an automated electronic tracking system to identify ED patients who were admitted to a hospital ward and then transferred to the ICU within 24h. Reviewers screened for possible error or adverse event and if discovered the case was referred to the departmental Quality Assurance (QA) committee for deliberations and consensus agreement. RESULTS: Of 96,377 ward admissions, 921 (1%) patients were subsequently transferred to the ICU within 24h of ED presentation. Of these 165 (19%) were then referred to the QA committee for review. Total rate of adverse events regardless of whether or not an error occurred was 2.1%, 19/921 (95% CI 1.4% to 3.0%). Medical error on the part of the ED was 2.2%, 20/921 (95% CI 1.5% to 3.1%) and ED Preventable Error in 1.1%, 10/921 (95% CI 0.6% to 1.8%). CONCLUSION: Tracking patients admitted to the hospital from the ED who are transferred to the ICU <24h after admission may be a valuable marker for adverse events and preventable errors in the ED.

Geriatric Syncope and Cardiovascular Risk in the Emergency Department.

BACKGROUND: Syncope is a transient loss of consciousness that is caused by a brief loss in generalized cerebral blood flow. OBJECTIVE: This article reviews the background, epidemiology, etiologies, evaluation, and disposition considerations of geriatric patients with syncope, with a focus on cardiovascular risk. DISCUSSION: Although syncope is one of the most common symptoms in elderly patients presenting to the emergency department, syncope causes in geriatric patients can present differently than in younger populations, and the underlying etiology is often challenging to discern. History, physical examination, and electrocardiography (ECG) have the greatest utility in evaluating syncope. Additional testing should be guided by history and physical examination. There are multiple scoring tools developed to aid in management and these are reviewed in the article. Common predictors that would indicate a need for further work-up include a history of cardiac or valvular disease (i.e., ventricular dysrhythmia, congestive heart failure), abnormal ECG, anemia or severe volume depletion (i.e., from a gastrointestinal bleed), syncope while supine or with effort, report of palpitations or chest pain, persistent abnormal vital signs, or family history of sudden death. With advancing age, cardiovascular morbidity plays a more frequent and important role in the etiology of syncope. CONCLUSIONS: The syncope work-up should be tailored to the patient's presentation. Disposition should be based on the results of the initial evaluation and risk factors for adverse outcomes.

Can medical record reviewers reliably identify errors and adverse events in the ED?

BACKGROUND: Chart review has been the mainstay of medical quality assurance practices since its introduction more than a century ago. The validity of chart review, however, has been vitiated by a lack of methodological rigor. OBJECTIVES: By measuring the degree of interrater agreement among a 13-member review board of emergency physicians, we sought to validate the reliability of a chart review-based quality assurance process using computerized screening based on explicit case parameters. METHODS: All patients presenting to an urban, tertiary care academic medical center emergency department (annual volume of 57,000 patients) between November 2012 and November 2013 were screened electronically. Cases were programmatically flagged for review according to explicit criteria: return within 72hours, procedural evaluation, floor-to-ICU transfer within 24hours of admission, death within 24hours of admission, physician complaints, and patient complaints. Each case was reviewed independently by a 13-member emergency department quality assurance committee all of whom were board certified in emergency medicine and trained in the use of the tool. None of the reviewers were involved in the care of the specific patients reviewed by them. Reviewers used a previously validated 8-point Likert scale to rate the (1) coordination of patient care, (2) presence and severity of adverse events, (3) degree of medical error, and (4) quality of medical judgment. Agreement among reviewers was assessed with the intraclass correlation coefficient (ICC) for each parameter. RESULTS: Agreement and the degree of significance for each parameter were as follows: coordination of patient care (ICC=0.67; P<.001), presence and severity of adverse events (ICC=0.52; P=.001), degree of medical error (ICC=0.72; P<.001), and quality of medical judgment (ICC=0.67; P<.001). CONCLUSION: Agreement in the chart review process can be achieved among physician-reviewers. The degree of agreement attainable is comparable to or superior to that of similar studies reported to date. These results highlight the potential for the use of computerized screening, explicit criteria, and training of expert reviewers to improve the reliability and validity of chart review-based quality assurance.

Utility of Procedural Sedation as a Marker for Quality Assurance in Emergency Medicine.

BACKGROUND: The Joint Commission requires health care organizations to monitor and evaluate procedural sedation. However, the utility of mandatory review of procedural sedation in evaluating health care quality is unknown. OBJECTIVE: To determine whether procedural sedation is a useful marker for evaluating error in the emergency department (ED). METHODS: We prospectively collected data for patients presenting to an urban, tertiary care, academic medical center ED between October 2013 and June 2015. We used an automated, electronic tracking system to identify patients who underwent procedural sedation. We randomly assigned cases to physician reviewers. Reviewers used a structured tool to determine the presence of error and adverse events. If a reviewer felt that the case had an error or adverse event, it was referred to a quality assurance (QA) committee, which made a final determination as to whether or not an error or adverse event occurred. RESULTS: There were 166 cases of procedural sedation reviewed. Two errors were identified, for an error rate of 1.2% (95% confidence interval [CI] 0.003-0.043). Both errors occurred during the use of propofol to facilitate upper gastrointestinal endoscopy. Neither error resulted in an adverse event. One adverse event was identified that was unrelated to physician error (0.6%; 95% CI 0.001-0.033). CONCLUSION: Routine review of procedural sedation performed in the ED offers little advantage over existing QA markers. Directed review of high-risk cases, such as those involving endoscopy or other longer-duration procedures, may be more useful. Future studies focusing quality review on projected high-risk sedation cases may establish more valuable markers for QA review.

Comparison of 1-Day Emergency Department Observation and Inpatient Ward for 1-Day Admissions in Syncope Patients.

BACKGROUND: In an era of increasing health care costs, the need for hospitalization is being scrutinized. In particular, 1-day hospitalizations are thought to be especially costly and unnecessary, and, increasingly, emergency department observation units (EDOUs) are being used as alternatives. OBJECTIVE: Our aim was to determine the differences in outcomes and diagnoses between 1-day inpatient and EDOU stays for syncope. METHODS: We retrospectively reviewed a cohort of patients with syncope who were seen in an urban ED with 1-day admission to an inpatient ward, EDOU, or full hospitalization. Etiology of syncope was classified as benign (vasovagal, dehydration), serious (dysrhythmia, sepsis, stroke/intracranial bleed, hemorrhage, valvular, ischemia, pulmonary embolism), or unknown. Data were analyzed using Fisher's exact test and t-test. RESULTS: One hundred and seventy-two of 351 patients were >1-day admissions, 152 (85%) were admitted for 1 day, and 27 (15%) were admitted to EDOU. The mean (standard deviation [SD]) age when admitted to the hospital was significantly higher at 72 (18.4) years for > 1-day admissions and 68.8 (19.6) years for 1-day admissions vs. 53.0 (18.9) years for EDOU patients (p < 0.01). For fully admitted patients, 36% had benign etiologies of syncope and 38% had serious causes of syncope; in 1-day admitted patients, 48% had benign etiologies and 14% had serious causes. Among EDOU patients, 44% had benign etiologies and none were serious. One-day patients were more likely to have unknown causes of syncope at discharge (36%; 95% confidence interval 0.28 to 0.43) when compared with admitted patients (26%; 95% CI 0.2 to 0.33); similarly, observation patients were more likely to be discharged without a diagnosis (56%; 95% CI 0.37 to 0.74; p ≤ 0.05). CONCLUSIONS: EDOU patients were less likely than patients admitted to the hospital to be discharged with an etiology of their syncope. Future EDOU protocols can benefit from set admission criteria and standardized evaluation protocols to facilitate maximal use of EDOU for syncope.

Assessing the rates of error and adverse events in the ED.

BACKGROUND: The incidence of errors and adverse events in emergency medicine is poorly characterized. OBJECTIVE: The objective was to systematically determine the rates and types of errors and adverse events in an academic, tertiary care emergency department (ED). METHODS: Prospective data were collected on all patients presenting to a tertiary-care academic medical center ED with an annual census of 55,000 patients between January 2009 and November 2012. Cases of patients meeting predetermined criteria were systematically identified by an electronic medical record system. Criteria for review included patients who (1) returned to the ED within 72 hours and were admitted on their second visit, (2) were admitted from the ED to the floor and then transferred to the intensive care unit (ICU) within 24 hours, (3) expired within 24 hours of ED arrival, (4) required airway management, or (5) were referred to the QA committee as the result of complaints. Cases were randomly assigned to individual physicians not involved with the care. All cases were reviewed using a structured electronic tool that assessed the occurrence of error and adverse events. Institutional review board jurisdiction was waived by the Beth Israel Deaconess Medical Center IRB. RESULTS: During the study period, 152,214 cases were screened and 2131 cases (1.4%) met prespecified criteria for review. The incidence of error in these cases was 9.5% (95% confidence interval [CI], 8.3%-10.8%), representing an overall incidence of 0.13% among all ED patients. In cases that involved error, 50.5% occurred among patients who returned to the ED within 72 hours; 17.3% occurred among floor-to-ICU transfers; 5.4% occurred among mortality cases; 2.0% occurred among airway cases; and 24.8% occurred among cases referred as the result of complaints. The incidence of adverse events in the reviewed cohort was 8.3% (CI, 7.2%-9.6%), representing an overall incidence of 0.11% among all ED patients. In cases that involved adverse events, 48.6% occurred among patients who returned to the ED within 72 hours; 16.4% occurred among floor-to-ICU transfers; 9.0% occurred among mortality cases; 1.1% occurred among airway cases; and 24.9% occurred among cases referred as the result of complaints. CONCLUSION: Although the overall incidence of error and adverse events in EDs is low, the likelihood of such events is markedly increased among patients who return to the ED within 72 hours, among patients who require floor-to-ICU transfer within 24 hours, and among those whose cases come to attention as the result of complaints.

Observation vs admission in syncope: can we predict short length of stays?

BACKGROUND: Rising health care costs demands justifying prolonged hospitalization for syncope, yet predictors of shorter length of stay (LOS) have not been identified. OBJECTIVE: The objective of this study is to identify independent predictors of LOS for syncope patients presenting to the emergency department (ED). METHODS: We performed an analysis of a prospectively collected cohort of ED syncope patients. We examined risk factors from the patient's ED presentation and workup and used Spearman rank correlation to evaluate the relationship between these risk factors and LOS in a logistic regression model for prediction of hospitalization less than 1 day. RESULTS: Of 568 total syncope patients, 350 (61.6%) were admitted and comprise the study cohort. Mean age was 57 years (SD, ±25 years); 39.6% were male. Length of hospitalization was less than or equal to 1 in 179 patients (51.1%). The number of risk factors was associated with LOS (P < .001). Most hospitalized patients with less than 1 predictor were hospitalized for less than 1 day (152/272; 55.9%). In multivariate logistic regression analysis, clear vasovagal etiology was positively correlated with LOS less than 1 day (odds ratio [OR], 1.92), whereas dysrhythmia (OR, 0.3), coronary artery disease (OR, 0.33), abnormal vital signs (OR, 0.35), implantable defibrillator/pacemaker (OR, 0.29), anemia/gastrointestinal bleeding (OR, 0.34), and central nervous system abnormalities (OR, 0.09; P < .05 for all) were negatively correlated with LOS less than 1 day. Area under the curve for model accuracy was moderate: 0.70 (95% confidence interval, 0.65-0.76). CONCLUSION: Patients with less than 1 predictor of prolonged stay, who present to the ED with syncope, are unlikely to require hospitalization of more than 1 day. A set of predictors may help define which patients are appropriate for 24-hour observation rather than full hospitalization.

Are echocardiography, telemetry, ambulatory electrocardiography monitoring, and cardiac enzymes in emergency department patients presenting with syncope useful tests? A preliminary investigation.

BACKGROUND: Prior studies of admitted geriatric syncope patients suggest that diagnostic tests affect management < 5% of the time; whether this is true among all emergency department (ED) patients with syncope remains unclear. OBJECTIVES: To determine the diagnostic yield of routine testing in the hospital or after ED discharge among patients presenting to an ED with syncope. METHODS: A prospective, observational, cohort study of consecutive ED patients aged ≥ 18 years presenting with syncope was conducted. The four most commonly utilized tests (echocardiography, telemetry, ambulatory electrocardiography monitoring, and troponin) were studied. Interobserver agreement as to whether test results determined the etiology of the syncope was measured using kappa (κ) values. RESULTS: Of 570 patients with syncope, 73 patients (8%; 95% confidence interval 7-10%) had studies that were diagnostic. One hundred fifty (26%) had echocardiography, with 33 (22%) demonstrating a likely etiology of the syncopal event, such as critical valvular disease or significantly depressed left ventricular function (κ = 0.75). On hospitalization, 330 (58%) patients were placed on telemetry, and 19 (3%) had worrisome dysrhythmias (κ = 0.66). There were 317 (55%) patients who had troponin levels drawn, of whom 19 (3%) had positive results (κ = 1); 56 (10%) patients were discharged with monitoring, with significant findings in only 2 (0.4%) patients (κ = 0.65). CONCLUSION: Although routine testing is prevalent in ED patients with syncope, the diagnostic yield is relatively low. Nevertheless, some testing, particularly echocardiography, may yield critical findings. Current efforts to reduce the cost of medical care by eliminating nondiagnostic medical testing and increasing emphasis on practicing evidence-based medicine argue for more discriminate testing when evaluating syncope.

Identifying patients with cellulitis who are likely to require inpatient admission after a stay in an ED observation unit.

BACKGROUND: Emergency department observation units (EDOU) are often used for patients with cellulitis to provide intravenous antibiotics followed by a transition to an oral regimen for discharge. Because institutional regulations typically limit EDOU stays to 24 hours, patients lacking a clinical response within this period will often be subsequently admitted to the hospital for further treatment. OBJECTIVE: The aim of this study was to determine the rate of hospital admission and characteristics predictive of admission in patients with cellulitis who are initially placed in an ED observation unit. METHODS: A retrospective cohort study of patients placed into EDOU with a diagnosis of skin infection was conducted. Age, sex, history of diabetes mellitus, immunosuppression, intravenous drug use, location of cellulitis, presence of abscess, laboratory infectious markers, vital signs, and outpatient antibiotic treatment were recorded. The primary outcome was a hospital admission due to failure to respond to treatment within the 24-hour observation time window. Significant variables on univariate analysis were used to create a multivariate analysis, which identified predictive characteristics. RESULTS: Four hundred six patient charts were reviewed, with 377 meeting inclusion criteria; the inpatient admission rate from EDOU was 29.2%. Using logistic regression techniques, we created a model of independent predictors for need of admission after 24 hours: cellulitis of the hand (odds ratio [OR], 2.9; 95% confidence interval [CI], 1.8-4.9), measured temperature higher than 100.4°F (OR, 2.5; 95% CI, 1.1-5.5), and lactate greater than 2 (OR, 3.1; 95% CI, 1.3-7.3) were predictive of failure of ED observation. CONCLUSIONS: Patients with cellulitis placed into ED observation status were more likely to fail an observation trial if they had an objective fever in the ED, an elevated lactate, or a cellulitis that involved the hand.

The Association between Emergency Department Length of Stay and In-Hospital Mortality in Older Patients Using Machine Learning: An Observational Cohort Study.

The association between emergency department (ED) length of stay (EDLOS) with in-hospital mortality (IHM) in older patients remains unclear. This retrospective study aims to delineate the relationship between EDLOS and IHM in elderly patients. From the ED patients (n = 383,586) who visited an urban academic tertiary care medical center from January 2010 to December 2016, 78,478 older patients (age ≥60 years) were identified and stratified into three age subgroups: 60-74 (early elderly), 75-89 (late elderly), and ≥90 years (longevous elderly). We applied multiple machine learning approaches to identify the risk correlation trends between EDLOS and IHM, as well as boarding time (BT) and IHM. The incidence of IHM increased with age: 60-74 (2.7%), 75-89 (4.5%), and ≥90 years (6.3%). The best area under the receiver operating characteristic curve was obtained by Light Gradient Boosting Machine model for age groups 60-74, 75-89, and ≥90 years, which were 0.892 (95% CI, 0.870-0.916), 0.886 (95% CI, 0.861-0.911), and 0.838 (95% CI, 0.782-0.887), respectively. Our study showed that EDLOS and BT were statistically correlated with IHM (p < 0.001), and a significantly higher risk of IHM was found in low EDLOS and high BT. The flagged rate of quality assurance issues was higher in lower EDLOS ≤1 h (9.96%) vs. higher EDLOS 7 h