Table 1.

Design Features and Results of Studies of Non-Urgent Visits (n=6)

Reference	Study Design	Non-urgent Definition	Sample Description and Setting	Sample Size
Brim (2008)⁴¹	Cross-sectional survey	Determined prospectively at triage (based on vital signs and expectations of procedures and treatments)	Convenience sample of adults presenting during business hours to one ED in Washington State	64 ED patients
Butler (1998)³⁶	Cross-sectional survey and review of health plan administrative data	Determined retrospectively from review of medical record (based on diagnosis). Also used alternate definitions from the literature to test the sensitivity of the logistic regression model	Enrollees of one Medicaid HMO in Colorado who had a non-urgent visit to an ED or PCP	581 patients with 1943 visits (outcome of interest was whether a particular nonemergency visit was to the ED or primary care provider)
Gill (1996)⁵²	Cross-sectional survey and medical records review	Determined prospectively at triage (based on ability to wait several hours or more for an evaluation)	Convenience sample in one ED in an unspecified location	268 ED patients
Northington (2005)⁵³	Cross-sectional survey	Determined prospectively at triage (based on vital signs, responsiveness, level of distress, and expectations of testing)	Convenience sample of adult self-referred patients in one ED in North Carolina	279 ED patients
Redstone (2008)⁵⁴	Cross-sectional survey	Determined prospectively at triage (based on symptoms, vital signs and expectations of resource use)	Convenience sample of adults with an established primary care provider presenting with a non-urgent condition to one ED in Colorado.	240 ED patients
Schwartz (1995)³⁸	Cross-sectional survey	Not clearly defined: Patients with conditions that were not life threatening such as flu, cold, or sprains	Patients who had a non-urgent visit to either one ED in Georgia or to a family practice clinic (FPC)	52 ED patients and 42 FPC patients