316 Emergency Department Benchmarking Alliance Survey: Anatomy and Physiology of an Emergency Department Observation Unit

Abstract

Health care entities can deliver observation services to determine whether or not a patient needs an inpatient admission. Many studies have shown that relative to care in a traditional inpatient bed, managing observation patients in an emergency department observation unit (EDOU), using condition-specific protocols, is associated with shorter hospital length of stays, lower costs, and decreased inpatient admission rates. Because of these advantages, health care entities have been opening these units more frequently. A 2003, survey found that 1/3 of hospitals in the US had EDOUs, and more were planning on building them. This statistic has been the standard talking point in the growth of observation for more than a decade. Since this original survey, there has been very little published data on the number of EDOUs as well as their operating characteristics. We surveyed the members of the Emergency Department Benchmarking Alliance (EDBA) as well as the American College of Emergency Physicians (ACEP) Observation Services Section as to the presence of an Observation Unit at their hospitals. Surveys were sent via email which contained a SurveyMonkey link to the questions. The account is securely maintained by EDBA. The responses were de-identified and entered into an excel spreadsheet for analysis. Respondents: Ninety sites responded to the survey representing 28 states plus the District of Columbia. Five responses answered only the state of practice and an additional 2 respondents answered only who admits to the OU. Unit Characteristics: Most of the OUs in this cohort were closed units. The median number of beds (IQR) for the respondents of the survey was 12 (10-18). Only 7 units reported not using patient care protocols. Unit Staffing: Emergency medicine admitted patients to 61 (67%) of the OUs, Hospitalist/Internal medicine were the admitting physicians in 5 units and Cardiology in 1 unit. Two sites reported they had hybrid units where EM or IM could admit to the OU. Fifteen of the units (16%) were “open,” were any service could admit to the OU. Advanced Practice Providers (APPs) are utilized in the vast majority of sites (69 as opposed to 13 that don’t). In terms of physician rounding in the unit, 49 of the sites (54%) utilize an emergency physician who is concurrently working a shift in the ED. Eighteen sites have a dedicated OU physician. One site reported that their NPs function independently with no physician oversight. The most common RN: Patient ratio is 1:5, used in 42 (46%) of the surveyed sites. The highest and lowest RN:Patient ratios are 1:2 and 1:8, respectively. These were each reported by one site. Unit Throughput Median annual volume (IQR) for this cohort is 4,015 (2,676 - 5,475) patients. Median length of stay in the unit (IQR) is 20 hours (17-23). Average conversion to inpatient status is 16%. Patient characteristics: The most common chief complaint seen in OUs in this cohort is chest pain. Other common reasons for observation include abdominal pain, TIA, syncope, asthma and COPD. Observation units as surveyed in this cohort are mainly emergency department-based and ED managed. Future challenges for these units will be maintaining occupancy in the face of potentially decreasing chest pain cases. Further study needs to be done as to the impact of APPs and role of EM/IM collaboration. Ultimately, a national database of Observation Units and observation patient stays is needed.