Characteristics of occasional and frequent emergency department users: do insurance coverage and access to care matter?

Med Care. 2004 Feb;42(2):176-82. doi: 10.1097/01.mlr.0000108747.51198.41.


Objective: The objective of this study was to explore how insurance coverage, access to care, and other individual characteristics are related to the large differences in emergency department (ED) use among the general population.

Materials and methods: We used the 1997 and 1999 National Survey of America's Families, a nationally representative sample. People were classified into 3 ED use levels based on the number of visits over the 12 months before the survey: non-ED users (zero visits), occasional users (1 or 2 visits), or frequent users (3 or more visits). We used a multinomial logit model to estimate the effect of insurance status and other factors on levels of ED use, and to compute the odds ratios of being occasional and frequent users as opposed to nonusers among various subpopulations.

Results: People in fair/poor health are 3.64 times more likely than others to be frequent ED users as compared with nonusers. The uninsured and the privately insured adults have the same risk of being frequent users, but publicly insured adults are 2.08 times more likely to be frequent users. Adults who made 3 or more visits to doctors are 5.29 times more likely to be frequent ED users than those who made no such visits.

Conclusion: The uninsured do not use more ED visits than the insured population as is sometimes argued. Instead, the publicly insured are overrepresented among ED users. Frequent ED users do not appear to use the ED as a substitute for their primary care but, in fact, are a less healthy population who need and use more care overall.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Emergency Service, Hospital / statistics & numerical data*
  • Ethnicity / statistics & numerical data
  • Family Characteristics
  • Health Care Surveys
  • Health Services Accessibility / statistics & numerical data*
  • Health Status
  • Humans
  • Insurance Coverage / statistics & numerical data*
  • Logistic Models
  • Odds Ratio
  • Racial Groups / statistics & numerical data
  • Socioeconomic Factors
  • United States