P1246341508houwx

1
Frequent Emergency Department Utilization In An
Urban Academic Hospital
Redefining Frequent Users Heterogeneously
Jennifer Ruger, PhD Christopher Richter, MD
Lawrence Lewis, MD Gerald Banet, RN MPH
Division of Emergency Medicine Washington
University School of Medicine St. Louis, Missouri
2
Presented at the American College of
Emergency Physicians (ACEP) Meeting
3
Context

• Frequent use of the Emergency Department (ED) is
  an important health care policy as it impacts
  both cost of healthcare and ED overcrowding

4
Context
continued

• Previous studies of frequent users have treated
  them as homogeneous group with large variations
  in the number of ED visits which defines a
  frequent user (2 ³ 12)
• Knowledge of factors that contribute to the
  frequent utilization of the ED for healthcare is
  limited

5
Objectives

• To define ED users heterogeneously based on their
  number of visits
• To determine patient mix differences and factors
  associated with five different groups of ED
  users
• 1 - 4 visits per year
• 5 - 10 visits per year
• 11 - 20 visits per year
• 21 - 30 visits per year
• 30 visits per year

6
Methods

• A retrospective cross-sectional study of data
  collected from hospital clinical, admission and
  financial records for all visits to the ED during
  2001 (N80, 209)
• All patients were assigned to one of five
  mutually exclusive user groups by number of
  visits for that year

7
Methods
continued

• Groups were then analyzed by
• Acuity and DRG severity level scores
• Age and gender
• Time and day of arrival
• Disposition status and payment method

8
Methods
continued

• Compared among five user groups, the main outcome
  measures are
• Triage acuity levels
• DRG severity levels
• Disposition status
• Time and day of arrival
• Payment method and demographic characteristics

9
Methods
continued

• These outcome measures are assessed in terms of
  their independent association, respectively, with
  each of the five user groups

10
Results

• In 2001, there were 80,209 visits to the
  Emergency Department. Visits ranged by category
• Group 1 (1 - 4) 68,466 (85.4)
• Group 2 (5 - 10) 9,274 (11.7)
• Group 3 (11 - 20) 1,620 (2.0)
• Group 4 (21 - 30) 411 (0.5)
• Group 5 (30) 438 (0.6)

11
Results
continued

• Group 1 had the largest number and percentage of
  total visits they were more likely to present
  with acuity level A (highest) severity than
  patients in all other groups combined (1.98 vs.
  1.15, P
• Group 4 (2.8 vs. 1.5, Pvs. 1.5, Ppresent with acuity level E (lowest) severity as
  compared to all other groups combined

12
Results
continued

• Admission and death rates were higher in Group 2
  (26 vs. 23 PP
• ICU/OR admission and death rates were higher in
  Group 1 (2.2 vs. 1.4 P4 (0.73, Psuggesting a lower level of severity in those
  groups

13
Results
continued

• The percentage of patients who had HMO, PPO,
  Commercial or Workers Compensation (private
  insurance) plans dropped off from Group 1 to
  Group 2
• There were virtually no patients with these plans
  in Groups 4 or 5

14
Results
continued

• Additionally, the Medicare Risk and Medicaid Risk
  plans (managed care programs) had virtually no
  patients in Groups 4 and 5 and a lower percentage
  of visits in Groups 2 and 3 than Traditional
  Medicare and Traditional Medicaid (indemnity)
  plans

15
Results
continued

• Patients in Group 5 had the largest percentage of
  self pay or uninsured patients (49 vs. 16
  Pall other groups combined

16
Discussion

• The majority (85) of ED visits were made by
  patients seen ? four times in 2001, and 45 of
  all ED visits were made by patients visiting the
  ED only once

17
Discussion
continued

• Among the 15 of visits by people who came ? 5
  times, our results revealed several different
  sets of independent associations between patient
  characteristics and ED utilization, suggesting
  that people who frequently use the ED comprise
  several heterogeneous groups

18
Discussion
continued

• Those using the ED up to twenty times (Groups 1 -
  3) were more acutely ill or severely injured than
  those using the ED 21 or more times (Groups 4 and
  5). Based on three indicators
• Triage acuity
• DRG severity level
• Composite hospital admission / death rate

19
Discussion
continued

• On the lower end of the severity spectrum, Groups
  4 and 5 were more likely to present with DRG
  severity level 0 (least severe) conditions
• Admission and death rates were higher in Groups 2
  and 3 and ICU/OR admission and death rates were
  highest in Group 1 and were virtually
  non-existent in Groups 4 or 5
• The likelihood of eloping or leaving AMA was
  greater in Groups 4 and 5

20
Discussion
continued

• Payment method comparisons in this study show two
  significant trends
• Patients with HMO, PPO, Commercial or Workers
  Compensation were less likely to be in the higher
  frequency groups, with virtually no such patients
  in Group 4 of 5 during this study period

21
Discussion
continued

• Payment method comparisons in this study show two
  significant trends continued -
• Patients with Medicaid Risk or Medicare Risk
  (managed care) had virtually no visits in Groups
  4 and 5 and a lower percentage of visits in
  Groups 2 and 3 than Traditional Medicare and
  Traditional Medicaid (indemnity) plans

22
Discussion
continued

• Patients in Group 5 had the largest percentage of
  self-pay or uninsured patients (3X that of all
  other groups combined)

23
Conclusion

• However, user groups with more than 21 visits
  /year tend to be less ill or injured than the
  general ED population
• Targeted policies to reduce ED utilization by the
  highest frequency groups may be pursued without
  altering the seemingly appropriate ED demand of
  other user categories

24
Conclusion
continued

• Frequent users of the ED comprise a heterogeneous
  group of patients
• Certain user groups 5-20 visits / year
  previously thought to be over-utilizing the ED
  for seemingly insignificant medical concerns or
  socio-economic problems may be as sick or sicker
  than the general ED population