• T Liu, M R Sayre, and S C Carleton.
• Department of Emergency Medicine, University of Cincinnati, OH 45267-0769, USA. tiepu.liu@uc.edu
• Acad Emerg Med. 1999 Nov 1;6(11):1147-52.

ObjectivesTo describe and compare national trends in ED use by statistical analyses on data from the 1992 to 1996 National Hospital Ambulatory Medical Care Survey (NHAMCS) with a special interest in factors related to nonurgent visits.MethodsThe NHAMCS collects data for ED visits using a four-stage national probability sample. Data from 135,723 ED visits in 1992-1996 were analyzed using the chi-square test for proportions with logistic regression modeling for multivariate analysis.ResultsMore than half of the ED visits were considered nonurgent. There was a decreasing trend for nonurgent ED visits over the first three years of the sample (54.0% to 52.1%, p < 0.05). The proportion of ED visits for nonurgent care bounced back in 1995 (54.7%) and 1996 (54.1%). Significant variation existed in the proportion of nonurgent care visit based on disease category, age, race, and insurance coverage status. Marked variation in nonurgent visits also existed among geographic regions and types of hospital ownership.ConclusionsAnalyses of data from the NHAMCS identify trends in ED use. The study of nonurgent ED visits with this database has inherent methodologic problems such as retrospective coding and geographic coding inconsistency. Since the nonurgent visit is clearly linked to certain social-demographic factors, addressing these underlying issues by establishing a comprehensive health care system is a priority.

21 keywords...

hide…