288 Evaluation of Syncope in a Dedicated Observation Unit

Abstract

The purpose of this study is to 1) determine the clinical background of patients evaluated for syncope in the emergency department (ED) observation unit (EDOU) and 2) determine which characteristics predict appropriateness of an EDOU level of care versus direct inpatient admission. This is a retrospective analysis of the 171 patients placed in the EDOU between March 2013 and July 2013 following evaluation in the ED for syncope. The study site is located in a diverse urban setting with a volume of 130,000 visits per year. The EDOU receives an average of 450 patients per month for a variety of chief complaints with an admission conversion rate of approximately 25%. Each patient file was reviewed by two trained evaluators to ensure accuracy. Patients were assigned to two different groups: those admitted during the EDOU evaluation (admitted group) and those discharged home from the EDOU (not admitted group). Descriptive statistics were computed for study variables to ensure that all data values were within expected ranges and to eliminate any data errors that may have occurred. SPSS version 21 was used to conduct the analysis. Chi-square and independent sample t-test were used to determine significant differences between groups. The sample consisted of 171 patients. The demographics of the patients evaluated are representative of the patient population of the study site. Of the 171 patients, 142 were discharged from the EDOU while 29 were converted to inpatient admissions from the EDOU. There were no significant differences between groups based on age (P=.756), sex (P=.449) or race/ethnicity (P=.310). There was no significant difference in the average number of past medical diagnoses for the admitted group and those discharged home (P=.796). However, admitted patients had significantly higher rates of reported medical history of syncope (27.6% versus 11.3%), P=.021, diabetes mellitus (DM) (37.9% versus 19.7%), P=.033, and an abnormal result in initial cardiac enzymes (20.7% versus 5.7%), P=.007. Of the 29 admitted patients, 15 (8.8% of the overall sample) developed unstable arrhythmias or non-ST elevation myocardial infarction (NSTEMI) with a median time to admission of 24 hours. With the exception of cardiac enzyme testing, no objective lab data, ECG finding, nor diagnostic testing, obtainable in the initial ED evaluation predicted the need for inpatient evaluation. The only statistically significant independent predictors of inpatient evaluation are a positive initial cardiac enzyme result, previous episode of syncope, and a past medical history that includes diabetes mellitus (DM). These factors may be used to create the inclusion and exclusion criteria for transfer to an EDOU. After initial evaluation, a finding of positive cardiac enzymes or a history of DM or previous syncope may be deemed grounds for exclusion from the EDOU. Using these exclusion criteria may help expedite evaluation and allocation of resources appropriately thereby positively impacting patient safety. This study was limited by the retrospective analysis of data. Future research will evaluate the differences between patients directly admitted for syncope versus those transferred to the EDOU first and then subsequently admitted. Applying the data learned from this study can help guide the appropriate care and ED disposition of patients thereby improving patient safety, patient flow, and hospital reimbursement.