Health policy/original research
Frequent Users of Emergency Departments: Developing Standard Definitions and Defining Prominent Risk Factors

https://doi.org/10.1016/j.annemergmed.2011.11.036Get rights and content

Study objective

We identify factors that define frequent and highly frequent emergency department (ED) users.

Methods

Administrative health care records were used to define less frequent (1 to 6 visits), frequent (7 to 17 visits), and highly frequent (≥18 visits) ED users. Analyses were conducted to determine the most unique demographic, disease, and health care use features of these groups.

Results

Frequent users composed 9.9% of all ED visits, whereas highly frequent users composed 3.6% of visits. Compared with less frequent users, frequent users were defined most strongly by their substance abuse challenges and by their many visits to primary care and specialist physicians. Substance abuse also distinguished highly frequent from frequent ED users strongly; 67.3% versus 35.9% of these patient groups were substance abusers, respectively. Also, 70% of highly frequent versus only 17.8% of frequent users had a long history of frequent ED use. Last, highly frequent users did not use other health care services proportionally more than their frequent user counterparts, suggesting that these former patients use EDs as a main source of care.

Conclusion

This research develops objective thresholds of frequent and highly frequent ED use. Although substance abuse is prominent in both groups, only highly frequent users seem to visit EDs in place of other health care services. Future analyses can investigate these patterns of health care use more closely, including how timely access to primary care affects ED use. Cluster analysis also has value for defining frequent user subgroups who may benefit from different yet equally effective treatment options.

Introduction

Between 18% and 23% of urban North American adults visit an emergency department (ED) annually1, 2, 3; however, a small portion of these patients account for many visits. Although early literature posits that frequent users visit EDs unnecessarily,4, 5, 6 recent evidence shows that these patients have many health needs. Several authors report that frequent versus less frequent ED users have more chronic diseases such as asthma, renal failure, chronic pulmonary disease, and hypertension7, 8, 9 and have more mental health issues such as depression and substance abuse.8, 10, 11, 12 Frequent ED patients are also reported to use many other health care services10, 13, 14 and to have episodic ED visit patterns,7, 11, 12 possibly reflecting health care emergencies, including “flare-ups” of chronic disease.

Despite this evidence, 3 challenges limit the development of alternate treatments for frequent ED patients. First, generalized findings are compromised by the lack of a standard “frequent ED use” definition, which ranges from patients with 21, 15, 16 to more than 123, 17 visits annually. Although the definition of 4 or more visits is used most often,2, 7, 10, 12, 14, 18 authors have generally not provided a clear rationale for this choice, other than stating that interventions should affect a sufficient number of ED visits to offset ED crowding.2, 9, 19, 20 With one exception,21 to our knowledge no researchers have developed an objective threshold of frequent ED use. This same criticism can be applied to the definition of highly frequent ED users.

As a second challenge, with exceptions noted,2, 3, 13, 14, 19 most frequent ED use studies have been conducted at single or few ED sites.8, 9, 10, 20, 22, 23, 24 Findings from this literature have limited generalizability because frequent users in both Canada3, 25 and the United States11, 19 have been shown to visit multiple EDs. Also, frequent ED use risk factors are captured often with self-report techniques,2, 8, 9, 22, 24 which are prone to nonresponse and patient recall biases. Administrative health care records have significant value for extending this literature26, 27 because current and past ED use patterns are captured accurately across multiple ED sites and on patient populations. For each person, these data can be linked to patterns of other health care use and to patient disease, without recall and loss to follow-up biases.

Although earlier studies are descriptive,28, 29 most literature now uses multivariable techniques to define how risk factors uniquely discern ED user groups.22, 23, 24 However, as a third limitation, no studies rank-order the importance of these risk factors. The most prominent features of frequent (versus less frequent) ED users are therefore unclear, limiting our ability to develop effective treatment options. This same criticism can be applied to frequent user subgroups to define what, if anything, makes them unique.

The goals of this article are 2-fold. First, patient characteristics were compared across incrementally frequent ED users. Breakpoints in these data were used to create frequent and highly frequent ED use thresholds. Second, multivariable analyses were used to determine demographic, illness, and health care use risk factors that discern most strongly these frequent user groups.

Section snippets

Study Design

This research was conducted with health care use records housed at Manitoba Centre for Health Policy, Faculty of Medicine, University of Manitoba. These records capture the dates and types of health care use made by Manitoba residents, including primary care and specialist physician visits, hospitalizations, home care services, and nursing homes. Physician diagnostic (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM]) codes are reported for physician

Results

Analyses were conducted on 105,687 patients with 200,810 ED visits (Table 1). Single users composed 59.2% of all patients but only 31.2% of ED visits. According to the often-used threshold of 4 plus visits to define frequent users, these patients composed 9.7% of all users and 31.3% of all ED visits.

To develop a frequent ED use threshold, select patient characteristics were viewed by ED visit frequency, and “breakpoints” in these trends were sought (Figure 2). No breakpoints were observed at 4

Limitations

This study has 5 limitations. First, we restricted analyses to adult users only, and our findings cannot be generalized to the pediatric population. Second, because administrative ED records are available only in Winnipeg Health Region, results cannot be generalized to rural EDs. Third, physician diagnostic (ICD-9-CM) codes are not provided for ED visits, which means we can identify patients with substance abuse (for example) according to their previous physician visits and hospitalizations,

Discussion

This article extends the ED literature in 3 important ways. First, our research provides objective thresholds of frequent (7 to 17 visits) and highly frequent (≥18) ED use. Standard definitions are lacking in literature18 but are important for comparing results across future studies. Our ability to create these definitions is likely attributed to the data systems we used and to our methodological design. Research on single ED sites cannot provide an overall picture of frequent ED use because

References (57)

  • F.R. Purdie et al.

    The chronic emergency department patient

    Ann Emerg Med

    (1981)
  • L.L. Roos et al.

    Policy analysis in an information-rich environment

    Soc Sci Med

    (2004)
  • L. Roos et al.

    A research registry: use, development and accuracy

    J Clin Epidemiol

    (1999)
  • J.A. Huang et al.

    Exploring medical utilization patterns of emergency department users

    J Formos Med Assoc

    (2008)
  • H.J. Ovens et al.

    Heavy users of emergency services: a population-based review

    CMAJ

    (2001)
  • R.E. Malone

    Whither the almshouse?overutilization and the role of the emergency department

    J Health Polit Policy Law

    (1998)
  • L.L. Spillane et al.

    Frequent users of the emergency department: can we intervene?

    Acad Emerg Med

    (1997)
  • J.H. Mandelberg et al.

    Epidemiologic analysis of an urban, public emergency department's frequent users

    Acad Emerg Med

    (2000)
  • B.C. Sun et al.

    Predictors and outcomes of frequent emergency department users

    Acad Emerg Med

    (2003)
  • K.K. Fuda et al.

    Frequent users of Massachusetts emergency departments: a statewide analysis

    Ann Emerg Med

    (2006)
  • B.T. Chan et al.

    Frequent users of emergency departmentsDo they also use family physicians' services?

    Can Fam Physician

    (2002)
  • J.M. Pines et al.

    Predictors of frequent emergency department utilization in southeastern Pennsylvania

    J Asthma

    (2006)
  • S. Zuckerman et al.

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

    Med Care

    (2004)
  • L.J. Cook et al.

    Repeat patients to the emergency department in a statewide database

    Acad Emerg Med

    (2004)
  • K.S. Okuyemi et al.

    Describing and predicting frequent users of an emergency department

    J Assoc Acad Minor Phys

    (2001)
  • T.E. Locker et al.

    Defining frequent use of an urban emergency department

    Emerg Med J

    (2007)
  • Y. Ledoux et al.

    Occasional and frequent repeaters in a psychiatric emergency room

    Soc Psychiatry Psychiatr Epidemiol

    (2006)
  • C. Naughton et al.

    The role of health and non-health-related factors in repeat emergency department visits in an elderly urban population

    Emerg Med J

    (2010)
  • Cited by (193)

    • Profiles of high emergency department users with mental disorders

      2022, American Journal of Emergency Medicine
    View all citing articles on Scopus

    Provide feedback on this article at the journal's Web site, www.annemergmed.com.

    A podcast for this article is available at www.annemergmed.com.

    Supervising editors: Brendan G. Carr, MD, MS; Donald M. Yealy, MD

    Author contributions: MBD, WP, DC, and R-AS planned the study. MBD and S. Day drafted the article. MBD supervised data analyses and approved the final article. WP, DC, CB, and S. Derksen revised the article. S. Day assisted with writing the final version of the article. DC provided detailed input into statistical analysis. R-AS, CB, and S. Derksen provided data analysis. MBD takes responsibility for the paper as a whole.

    Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). The authors have stated that no such relationships exist. This research was supported by contractual agreement with Manitoba Health (HIPC 2005/2006-15).

    Publication date: Available online February 2, 2012.

    Please see page 25 for the Editor's Capsule Summary of this article.

    View full text