Characterizing emergency departments to improve understanding of emergency care systems

The task of characterizing different emergency departments (EDs) is complicated by the fact that a wide array of entities function as EDs. This is particularly the case when studying EDs in different countries; yet, as the ACEP Section on International Emergency Medicine (EM) has emphasized, increased globalization trends both facilitate and require the exchange of knowledge and ideas within the international EM community in order to benefit global public health and health policy [1]. Since 2002, the Emergency Medicine Network (EMNet) has made such an effort by collecting information about emergency care in countries around the world as part of the National ED Inventories (NEDI) project. Countries studied, to date, include the United States (US, including more detailed work in 9 states), China (Beijing), Denmark, Nigeria (Abuja), Paraguay (Asuncion), Singapore, and Slovenia. (Much of the data provided in this paper comes from projects summarized on the NEDI website: http://www.emnet-nedi.org webcite.). In conducting these studies, we expected to find international diversity among EDs [2,3]. We were surprised, however, by the ED diversity even within the US [4,5].By repeating NEDI studies in multiple countries, it became clear that one can learn a great deal about a system of emergency care by examining its constituent EDs. Though this is not the only means of understanding systems of emergency care and emergency care can exist without EDs, examining emergency care systems via EDs yields a particularly rich portrait of local emergency care delivery. For instance, during the NEDI-Slovenia project, we found an unusual amount of variation in ED visit volume across the country. Upon examining the layout of EDs, we discovered that many Slovenian EDs are located within other specialty units and, therefore, may exist in multiple areas of a hospital. As such, complete visit volume data reflecting all emergency visits were not always obtainable, resulting in the observed inco