Until Emergency Severity Index (ESI) triage became available in the Unites States, emergency department (ED) triage was most commonly performed by ED nurses using a poorly designed, nonstandardized, three-level triage method. Most often the three-level method classified incoming patients as emergent, urgent, or nonurgent; or as a level 1, 2 or 3 patient. These three-level methods hinged greatly on an assessment by the triage nurse of “How long do you think this patient can wait?” The reliability of the triage assignments produced by these threelevel methods, asking this type of question, was very poor. That is, the interrater (between triage nurses) and test-retest (same nurse doing triage again) assignments that these three-level methods produced were very inconsistent. In contradistinction, the definitions used to differentiate patients with ESI five-level triage are explicit and are easily understood-by clinicians and nonclinicians (such as hospital administrators) alike. In ESI triage the ED triage nurse asks two questions: Who should be seen first (levels 1 and 2)? and What do you think the patient will need in terms of resources (where resources are specified) to reach an ED disposition (levels 5, 4, and 3)? The ESI Triage Research Team believes that a principal goal of ED triage should be to determine who should be seen first. But a second major goal of ED triage should be not to just “sort” but to “stream.” This second goal of ED triage is about getting the right patient to the

right resources in the right place and at the right time. ESI triage is fundamentally rooted in an industrial engineering way of thinking. ESI triage is about quick sort and stream; indeed, it is often called the “Quick Sort” step in an emergency department flow mapping or simulation modeling exercise.