Search

Study objective Emergency department (ED) crowding is a barrier to timely care. Several crowding estimation tools have been developed to facilitate early identification of and intervention for crowding. Nevertheless, the ideal frequency is unclear for measuring ED crowding by using these tools. Short intervals may be resource intensive, whereas long ones may not be suitable for early identification. Therefore, we aim to assess whether outcomes vary by measurement interval for 4 crowding estimation tools. Methods Our eligible population included all patients between July 1, 2015, and June 30, 2016, who were admitted to the JPS Health Network ED, which serves an urban population. We generated 1-, 2-, 3-, and 4-hour ED crowding scores for each patient, using 4 crowding estimation tools (National Emergency Department Overcrowding Scale [NEDOCS], Severely Overcrowded, Overcrowded, and Not Overcrowded Estimation Tool [SONET], Emergency Department Work Index [EDWIN], and ED Occupancy Rate). Our outcomes of interest included ED length of stay (minutes) and left without being seen or eloped within 4 hours. We used accelerated failure time models to estimate interval-specific time ratios and corresponding 95% confidence limits for length of stay, in which the 1-hour interval was the reference. In addition, we used binomial regression with a log link to estimate risk ratios (RRs) and corresponding confidence limit for left without being seen. Results Our study population comprised 117,442 patients. The time ratios for length of stay were similar across intervals for each crowding estimation tool (time ratio=1.37 to 1.30 for NEDOCS, 1.44 to 1.37 for SONET, 1.32 to 1.27 for EDWIN, and 1.28 to 1.23 for ED Occupancy Rate). The RRs of left without being seen differences were also similar across intervals for each tool (RR=2.92 to 2.56 for NEDOCS, 3.61 to 3.36 for SONET, 2.65 to 2.40 for EDWIN, and 2.44 to 2.14 for ED Occupancy Rate). Conclusion Our findings suggest limited variation in length of stay or left without being seen between intervals (1 to 4 hours) regardless of which of the 4 crowding estimation tools were used. Consequently, 4 hours may be a reasonable interval for assessing crowding with these tools, which could substantially reduce the burden on ED personnel by requiring less frequent assessment of crowding.

Study Objective: We sought to describe the feasibility of an emergency department observation unit (EDOU) treatment protocol for the management of uncomplicated acute-onset atrial fibrillation (AAF). Methods: This descriptive case series took place at a major suburban, university-affiliated teaching hospital. Patients were prospectively enrolled in an EDOU treatment protocol if they had uncomplicated AAF that failed initial ED attempts to convert to sinus rhythm. In the EDOU, patients underwent ECG monitoring, serial creatine kianse MB measurements, and further rate control with optional electrical cardioversion. Primary outcomes measured were EDOU rate of conversion to sinus rhythm, rate of discharge home, length of stay, positive diagnostic outcomes, complications of AAF, and 7-day return visits. Results: Sixty-seven patients were studied. Patients were symptomatic for a median of 4.0 hours, had mean initial ED pulse rates of 137±23 beats/min, and spent 4.7±2.2 hours in the ED before transfer to the EDOU. While in the EDOU, 55 (82%) patients converted to sinus rhythm. Five (7%) patients were admitted because of positive test results: 2 for myocardial infarction, 2 for fever, and 1 for ventricular tachycardia. Twelve (18%) patients remained in atrial fibrillation, with 9 admitted and 3 discharged. Overall, 81% of patients were discharged in 11.8±7.0 hours, and 19% were admitted after 17.6±9.5 hours of observation. Three discharged patients returned within 7 days, 2 for uncomplicated recurrent AAF and 1 for chest pain subsequently found to be noncardiac in origin. There were no major complications attributable to the EDOU protocol. Conclusion: Selected patients with AAF for whom initial ED management fails can subsequently be managed in an EDOU with a high short-term conversion and discharge rate. [Koenig BO, Ross MA, Jackson RE. An emergency department observation unit protocol for acute-onset atrial fibrillation is feasible. Ann Emerg Med. April 2002;39:374-381.]

We derive a decision rule to partition emergency department patients with suspected pulmonary embolism (PE) into a small, high-risk group (40% pretest probability) that is unsafe for D -dimer testing and a larger group that is safe to have PE ruled out with either a whole-blood D -dimer plus alveolar deadspace measurement or a quantitative D -dimer assay.Nine hundred thirty-four patients with suspected PE were studied at 7 urban EDs in the United States. Patients were prospectively interviewed and examined for recognized symptoms, signs, and risk factors associated with PE. These data were collected before standard objective imaging for PE. Selected variables were analyzed by multivariate logistic analysis to determine factors associated with PE (P.05). A decision rule was then constructed to categorize approximately 80% of ED patients as safe for D -dimer testing.Pretest prevalence of PE was 19.4% (181/934; 95% confidence interval [CI] 16.3% to 21.7%). Six variables found to be significant on multivariate analysis were used to construct the decision rule. Unsafe patients had either a shock index (heart rate/systolic blood pressure) more than 1.0 or age older than 50 years, together with any one of the following conditions: unexplained hypoxemia (SaO (2)95%; no prior lung disease), unilateral leg swelling, recent major surgery, or hemoptysis. These criteria were met by 197 (21.0%) of 934 patients, and 83 of 197 (42.1%; 95% CI 35.3% to 49.6%) patients had PE. Exclusion of these 197 unsafe patients significantly decreased the probability of PE in the remaining 737 (79.0%) safe patients to 13.3% (95% CI 10.9% to 15.9%).Simple clinical criteria can permit safe D -dimer testing in the majority of ED patients with suspected PE. These criteria warrant prospective validation.

Study Objective: Emergency physicians frequently rely on normal findings from a lower extremity venous ultrasound examination as a method to decrease the probability of pulmonary embolism (PE) in outpatients with a nondiagnostic ventilation-perfusion lung scan (V/Q scan). The objective of this study was to evaluate the diagnostic utility of bilateral lower extremity venous ultrasound scanning in the diagnosis of PE in emergency department patients with a low-, moderate-, or indeterminate-probability (nondiagnostic) V/Q scan. Methods: This prospective, 2-center, descriptive study was conducted at the EDs of 2 large teaching hospitals. From an initial cohort of 570 nonreferred outpatients, a convenience sample of 156 patients who had both a nondiagnostic V/Q scan and a lower extremity venous ultrasound scan performed was selected as the study population. The sensitivity and specificity for a single lower extremity venous ultrasound scan and the posttest probability of PE were determined for the study population. Results: In the study population, the best-case sensitivity of the lower extremity venous ultrasound scan for PE was 54% (95% confidence interval [CI] 37% to 71%) and the specificity was 98% (95% CI 94% to 100%). The likelihood ratio of a positive test result was 27. The likelihood ratio of a negative test result was 0.49, yielding a lowest possible posttest probability of PE of 12% (95% CI 6% to 17%). Conclusion: This study demonstrates that the combination of a nondiagnostic (low, moderate, or indeterminate) V/Q scan plus a single negative result from lower extremity venous ultrasound examination, even in a best-case scenario, does not exclude the diagnosis of PE. [Daniel KR, Jackson RE, Kline JA. Utility of lower extremity venous ultrasound scanning in the diagnosis and exclusion of pulmonary embolism in outpatients. Ann Emerg Med. June 2000;35:547-554.]

We sought to determine whether a patient's sex independently influences the interval from emergency department arrival to the initiation of thrombolytic therapy in acute myocardial infarction (AMI).We conducted a retrospective cohort study in two suburban EDs, one at a 929-bed tertiary care teaching hospital and the other at a 189-bed community hospital. Only patients found to be having an ST-segment-elevated AMI on their first ECG who were treated with a thrombolytic agent in the ED were eligible. We excluded patients who arrived at the ED after cardiac arrest or with a known AMI. We used as the main outcome measure the interval from ED arrival to initiation of thrombolytic therapy. Secondary outcome variables included time elapsed before ECG, interval between ECG and treatment, and 1-year mortality.Entry criteria were satisfied by 328 patients. The 88 women experienced a mean 23-minute delay to treatment initiation compared with men (P.01). This observation is not accounted for by age, race, time of day, medical history, sex of the physician, type of thrombolytic agent, hospital, or triage category. The longest delays were found in women treated by female physicians, although female physicians also waited longer than male physicians to administer thrombolytic therapy to men. The mean time elapsed before the first ECG was also 6 minutes longer for women (P.01) Women had an increased 1-year mortality rate that was fully explained by their advanced age at the time of AMI.We infer that a patient's sex may play a significant role in the observed delay in treatment for women. Our data, coupled with previously published work, strongly suggest a systematic negative effect for women in their interaction with the health care system during AMI. We suggest that variables other than systems issues affect the time elapsed before thrombolytic therapy.

Study objective: To validate criteria predicting ankle and midfoot fractures with 100% sensitivity. Design: Prospective validation study Setting: A 929-bed community teaching hospital with an annual census of 76,488 ED visits. Participants: Convenience sample of patients older than 18 years with acute ankle or midfoot injury. Interventions: Radiography was performed in each patient received after pertinent history and physical examination findings were recorded. Results: Five hundred seventy radiographs were obtained in 484 patients. Four hundred twenty-one were of the ankle, and 149 were of the foot. There were 93 ankle fractures and 29 midfoot fractures, giving a fracture yield of 22.1% for ankle films and 19.5% for foot films. Decision rules had sensitivity of 94.6% and specificity of 15.5% for ankle fractures and sensitivity of 93.1% and specificity of 11.5% for midfoot fractures. Prospective criteria failed to predict fracture in five of the ankle group and two of the midfoot group. Physicians predicting fracture solely on the basis of clinical suspicion had a sensitivity of 69% in ankle injuries and 76% in midfoot injuries. Conclusion: We were unable to validate with 100% sensitivity the Ottawa rules predicting ankle and midfoot fractures. However, the Ottawa rules were more sensitive than clinical suspicion alone. [Lucchesi GM, Jackson RE, Peacock WF, Cerasani C, Swor RA: Sensitivity of the Ottawa rules. Ann Emerg Med July 1995;26:1-5.]

Study objective: To compare the levels of work-related stress and depression reported by practicing emergency physicians in three survey sites and to determine the effects of gender and marital status on the stress and depression experienced by these physicians. Design: Cross-sectional mail surveys. Setting and participants: Seven hundred sixty-four practicing emergency physicians from the United States, 91 fellows in full-time practice from Australasia, and 154 consultants and 47 senior registrars from the United Kingdom. Intervention: Administration of questionnaires requesting demographic information and including an inventory to assess work-related stress and a scale to measure depressive symptomatology. Measurements and main results: A 3×2×2 multivariate analysis of variance performed to compare scores on the stress inventory and depression scale simultaneously by survey site, gender, and marital status revealed significant differences in stress and depression by survey site and marital status. Univariate analyses of variance revealed significant differences in both stress and depression among the three survey sites and in depression by marital status. Adjusted means indicated that physicians from the United Kingdom reported higher levels of stress and depression than physicians from the United States and Australasia. Physicians from the United States and Australasia did not differ with respect to stress or depression. Physicians who were not married reported higher levels of depression than married physicians. No large mean differences, actual or adjusted, were found for any of the grouping factors. Conclusion: Statistical differences among practicing emergency physicians from the United States, Australasia, and the United Kingdom were observed, but the actual levels of work-related stress and depression were similar and did not appear severe. Marriage was associated with lower levels of depressive symptomatology.

Study objectives: This study tested the hypothesis that serial creatine phosphokinase (CK)-MB sampling in the emergency department can identify acute myocardial infarction (AMI) in patients presenting to the ED with chest pain and nondiagnostic ECGs. Design: Patients more than 30 years old who were evaluated initially in the ED and hospitalized for chest pain were studied. Serial CK-MB levels were analyzed prospectively using a rapid serum immunochemical assay for identification of AMI patients in the ED. Presenting ECGs showing new, greater than 1-mm ST elevation in two or more contiguous leads were considered diagnostic for AMI. All other ECGs were considered nondiagnostic ECGs. CK-MB levels were determined at ED presentation and hourly for three hours (total of four levels). Patients with at least one level of more than 7 ng/mL were considered to have a positive enzyme study. The in-hospital diagnosis of AMI was determined by the development of typical serial ECG changes or separate standard cardiac enzyme changes after admission. Setting: Eight tertiary-care medical center hospitals. Methods and main results: Of the 616 study patients, 108 (17.5%) were diagnosed in the hospital as AMI; 69 of these AMI patients (63.9%) had nondiagnostic ECGs in the ED. Of the patients with nondiagnostic ECGs, 55 (sensitivity, 79.7%) had a positive ED serial CK-MB enzyme study within three hours after presentation. Combining serial ED CK-MB assay results with diagnostic ECGs yielded an 88.4% sensitivity for AMI detection within three hours of ED presentation. The predictive value of a negative serial ED enzyme study for no AMI was 96.2% (specificity, 93.7%). Conclusion: Serial CK-MB determination in the ED can help identify AMI patients with initial nondiagnostic ECGs. Use of serial CK-MB analysis may facilitate optimal in-hospital disposition and help guide therapeutic interventions in patients with suspected AMI despite a nondiagnostic ECG.

There are currently 20 autonomous departments of emergency medicine in United States medical schools. EDs seeking autonomous status should institute a faculty development program to channel faculty energy into worthwhile research projects; establish protected time for clinical faculty to increase research productivity; develop expertise to compete for extramural funding; initiate an intramural research program so that faculty can learn the basics of grantsmanship; teach health care issues in ambulatory medicine; become involved in interdisciplinary teaching programs and curriculum development; maintain the present faculty commitment to 24-hour attending coverage; and develop university-based programs that originate from the ED. Program directors should establish liaisons with the medical school dean to acquaint him with the advantages of an autonomous department of emergency medicine; attempt to assess other relationships within the medical school to determine support for emergency medicine and to uncover and address opposition to autonomous departmental status; attempt to serve on medical school committees to meet other faculty, solve problems with them and develop trusting relationships; and develop broad-based support for autonomous departmental status both within and outside of the university. By devising and following a deliberate approach to attaining departmental status, emergency medicine will be assured of continued growth in the important decade ahead.

Study objective Point-of-care testing reduces time to cardiac marker results in patients evaluated for acute coronary syndromes, yet evidence this translates to a decreased length of stay is lacking. We hypothesized that point-of-care testing decreases length of stay in patients being evaluated for acute coronary syndromes in the emergency department (ED). Methods Patients being evaluated for possible acute coronary syndromes at 4 EDs in the United States were randomized to having point-of-care markers as well as central laboratory markers, or central laboratory markers only (laboratory arm). Point-of-care markers were obtained using early serial testing at presentation and at 90, 180, and 360 minutes as required by the treating physician. Evaluation, treatment, and disposition decisions were at the treating physician's discretion. Length of stay was from presentation to the time of departure from the ED, either to an inpatient setting or to home. Results There were 1,000 patients in each study arm. There were 520 patients discharged home from the ED. Median (interquartile range) time to discharge home was 4.6 hours (3.5 to 6.1 hours) in laboratory patients and 4.5 hours (3.5 to 6.1 hours) in point-of-care patients. Median (interquartile range) time to transfer to an inpatient setting for admitted patients was 5.5 hours (4.2 to 7.5 hours) in laboratory patients, and 5.4 hours (4.1 to 7.3 hours) in point-of-care patients. At one site, time to transfer to the floor was reduced in the point-of-care arm compared with the laboratory arm (difference in medians 0.45 hours; 95% confidence interval [CI] –0.14 to 1.04 hours). At one site, time to ED departure for discharged patients was higher in the point-of-care arm than the laboratory arm (difference in medians 1.25 hours; 95% CI 0.13 to 2.36 hours). Conclusion The effect of point-of-care testing on length of stay in the ED varies between settings. At one site, point-of-care testing decreased time to admission, whereas at another, point-of-care testing increased time to discharge. Potential effects of point-of-care testing on patient throughput should be considered in the full context of ED operations.

One of the value statements of the American College of Emergency Physicians states that, "Quality Emergency Medicine is best practiced by qualified, credentialed emergency physicians." 1 To address this value ACEP has established the following goal: "The number of board-certified physicians will be sufficient to meet the manpower needs of the public." 1 It is the position of ACEP that there is currently a severe shortage of appropriately trained and certified emergency physicians and, moreover, that the shortage will continue well into the next century. We discuss how ACEP arrived at this position and the role of academic emergency medicine in addressing this shortage. For many years, there has been a public debate as to whether there is a physician shortage or surplus. The Graduate Medical Education National Advisory Commission report of 1980 estimated that there would be 630,000 US physicians by 1990, with a surplus of 70,000. 2 This report also identified emergency medicine as a shortage specialty, indicating there would be a need for 14,000 emergency physicians in 1990, with a supply of only 8,000. Schwartz included such factors as increased provision of administrative and research activity by physicians and concluded that there would be a shortage of 7,000 physicians by the year 2,000. 3

Emergency physicians frequently rely on normal findings from a lower extremity venous ultrasound examination as a method to decrease the probability of pulmonary embolism (PE) in outpatients with a nondiagnostic ventilation-perfusion lung scan (V/Q scan). The objective of this study was to evaluate the diagnostic utility of bilateral lower extremity venous ultrasound scanning in the diagnosis of PE in emergency department patients with a low-, moderate-, or indeterminate-probability (nondiagnostic) V/Q scan.This prospective, 2-center, descriptive study was conducted at the EDs of 2 large teaching hospitals. From an initial cohort of 570 nonreferred outpatients, a convenience sample of 156 patients who had both a nondiagnostic V/Q scan and a lower extremity venous ultrasound scan performed was selected as the study population. The sensitivity and specificity for a single lower extremity venous ultrasound scan and the posttest probability of PE were determined for the study population.In the study population, the best-case sensitivity of the lower extremity venous ultrasound scan for PE was 54% (95% confidence interval [CI] 37% to 71%) and the specificity was 98% (95% CI 94% to 100%). The likelihood ratio of a positive test result was 27. The likelihood ratio of a negative test result was 0.49, yielding a lowest possible posttest probability of PE of 12% (95% CI 6% to 17%).This study demonstrates that the combination of a nondiagnostic (low, moderate, or indeterminate) V/Q scan plus a single negative result from lower extremity venous ultrasound examination, even in a best-case scenario, does not exclude the diagnosis of PE.

To determine whether race, when controlled for income, is an independent predictor of survival to hospital discharge after out-of-hospital cardiac arrest (OHCA).Prospective OHCA data were collected over 4 years (1991-1994) from a convenience sample of OHCA patients transported to nine hospitals in three suburban counties. Race was determined from hospital and vital statistics records. The average household income was identified from ZIP codes and used as a marker of socioeconomic status. Demographic data and known predictors of survival were compared between blacks and whites. A logistic regression analysis was used to assess the association between race, income, and survival.Of the 1,690 patients, 223 (13%) were blacks and 1,467 (87%) were whites. Average household income was less for blacks than for whites ($40,225 versus $46,193; P.001), but both populations were affluent by national standards (national percentile ranks were 73% and 88%, respectively). The populations were no different in percentage of witnessed arrests (57% versus 61%; P = .465). Blacks were younger (mean +/- SD, 62 +/- 16 versus 68 +/- 15 years; P.001); less frequently received bystander CPR (11% versus 20%; P = .002); less often had ventricular tachycardia or ventricular fibrillation (37% versus 50%; P.001); and had a shorter advanced life support call-response interval (median, 4 versus 6 minutes; P.001). The odds ratio for survival (white/black) was .931 (95% confidence interval, .446 to 1.945).Race was not found to predict adverse OHCA outcomes in this affluent population.

Study objective: To develop criteria that optimize clinical decisionmaking in the use of radiography after isolated knee trauma in adults. Design: A prospective survey of emergency department patients over a 7-month period. Standardized data forms were completed by emergency physicians, residents, and certified physician assistants. Setting: A large suburban community teaching hospital. Participants: Two hundred forty-two patients older than 17 years with isolated knee injuries sustained less than 24 hours previously. Results: We constructed a clinical decision model, calculating sensitivity, specificity, and odds ratios. Twenty-eight patients (11.6%) had fractures, with the patella the most commonly fractured osseous structure. Patients able to walk without limping had not experienced a fracture, nor had patients with twist injuries without effusion. Sensitivity of this model for detecting fracture was 1.0 (99% confidence interval, .97 to 1.0), and specificity was .337 (99% confidence interval, .26 to .42). Conclusion: Clinical decision rules are effective in detecting knee fractures with 100% sensitivity and with sufficient specificity to eliminate 29% of knee radiographs in the ED. These findings require prospective validation. [Weber JE, Jackson RE, Peacock WF, Swor RA, Carley R, Larkin GL: Clinical decision rules discriminate between fractures and nonfractures in acute isolated knee trauma. Ann Emerg Med October 1995;26:429-433.]

Study objective: To compare the levels of work-related stress and depression reported by physicians-in-training in emergency medicine in three survey sites and to determine the effects of gender and marital status on stress and depression among these physicians. Design: Cross-sectional mail surveys. Setting and type of participants: Physicians-in-training in the United States, United Kingdom, and Australasia. Intervention: Questionnaires requesting demographic information and including scales assessing work-related stress and depression were administered. Measurements and results: A 3 × 2 × 2 multivariate analysis of variance in which survey site, gender, and marital status were independent variables and stress and depression scale scores were dependent variables revealed significant differences when stress and depression were analyzed simultaneously. Univariate analyses of variance revealed significant differences in stress by survey site and gender and in depression for all three independent variables. Comparison of adjusted means revealed that respondents from the United Kingdom reported significantly higher levels of stress than did respondents from the United States and that women reported significantly higher levels than men. Respondents from the United States reported significantly higher levels of depression than did respondents from the other countries, women reported higher levels than men, and unmarried respondents reported higher levels than married respondents. Conclusion: Despite limitations resulting from self-report bias, cross-sectional survey methodology, sampling error, and differences in training among the three survey sites, the respondents experienced similar levels of stress and depression attributable to anticipated sources.

The case of an 18-year-old man with sigmoid volvulus and recurrent abdominal pain is presented. He was seen in the emergency department three times in a 4-month period, each time complaining of cramping left lower quadrant pain of one to two hours duration without vomiting or diarrhea. Physicial examination on each occasion revealed left lower quadrant tenderness without mass, guarding, or rebound. Radiologic evaluation on the first visit revealed sigmoid volvulus, which was reduced by barium enema. Despite identical clinical presentation on two subsequent occasions, radiologic studies showed no evidence of recurrent volvulus. During the ensuing two years, the patient has had no further symptoms.

The current recommendation of the American Heart Association is to give 0.5 to 1.0 mg (7.5 to 15 micrograms/kg in a 70-kg man) of epinephrine intravenously every five minutes during cardiac arrest. The optimal dose of epinephrine to augment the aortic diastolic pressure (ADP) is not known. The effect of various doses of central bolus epinephrine on the ADP during closed-chest massage was studied. A group of 25 large dogs was divided equally into five groups: control and 15, 45, 75, and 150 micrograms/kg. After three minutes of cardiac arrest, closed-chest massage was initiated, and the study drug was given two minutes later. The ADP and right atrial pressures were monitored for 15 minutes. Changes in ADP peaked at two minutes after injection in all groups receiving epinephrine, and the drop in ADP over time noted in the control group was prevented by increasing doses of epinephrine. Among the groups receiving epinephrine, however, there was no difference in the absolute ADP and diastolic coronary perfusion pressure.