ORIGINAL ARTICLE
Utility of the Emergency Department Observation Unit in
Ensuring Stress Testing in Low-Risk Chest Pain Patients
Troy Madsen, MD, Michael Mallin, MD, Joseph Bledsoe, MD, Philip Bossart, MD, Virgil Davis, MD,
Christopher Gee, MD, and Erik Barton, MD, MS
Background: Recent research has noted low rates of compliance among ED
chest pain patients referred for outpatient stress testing. The practice at our
institution, a 39,000 visits per year emergency department (ED), is to
place chest pain patients considered low risk for acute coronary syndrome
in an observation unit for serial biomarker testing and provocative cardiac
testing. Our objective was to determine the rates of positive stress tests
among this group and to extrapolate from this the potential missed
positive stress tests if these patients were referred instead for outpatient
stress testing.
Methods: This was a retrospective chart review of all chest pain patients
admitted to the ED observation unit between April 2006 and June 2007.
Baseline information, including a history of coronary disease, was recorded.
Patients underwent a treadmill stress test, nuclear stress test, or coronary CT
scan at the discretion of the attending emergency physician and/or the
consulting cardiologist. Rates of positive stress test or coronary CT and
patient disposition (admission to an inpatient unit versus discharge for
outpatient follow-up) were noted.
Results: A total of 353 patients underwent stress testing or coronary CT
during the study period: 257 (72.8%) patients had an exercise treadmill
echocardiogram, 61 (17.3%) patients underwent nuclear stress testing, and
35 (9.9%) patients had a coronary CT. Seventy patients (19.8%) had a history
of coronary disease but had been considered appropriate for observation by
the attending emergency physician. Thirty-nine stress tests were positive
(11%) and 11 were indeterminate (3.1%). Among patients with no history of
coronary disease, 20 stress tests were positive (7.1%), and 10 were indeter-
minate (3.5%). Of all patients with a positive stress test, 19 (48.7%)
underwent cardiac catheterization and 1 (2.6%) had coronary artery bypass
graft. Twenty-one of 39 patients with a positive stress test (54%) were
ultimately admitted to an inpatient unit per the recommendation of the
consulting cardiologist. Assuming a best-case scenario in which 70% of
patients referred for outpatient stress testing actually have the testing done
(based on a recent report of outpatient compliance), physicians would miss
approximately 3.3% of patients with a positive stress test if these patients
were discharged directly from the ED.
Conclusion: Among chest pain patients admitted to an ED observation unit,
the rate of positive stress tests was 11%. Approximately 3.3% of patients
with positive stress tests may have been missed if these patients were instead
referred for outpatient testing.
Key Words: observation unit, chest pain, stress test
(Crit Pathways in Cardiol 2009;8: 122–124)
O
bservation units for chest pain patients are commonly used to
expedite the diagnosis and treatment of patients with the
potential for acute coronary syndrome and unstable angina.
1–6
Earlier research has focused on the use of chest pain units or
observation units to expedite the care of patients presenting with
chest pain, and many studies have focused on the cost effectiveness
of these units.
7–9
One advantage of an observation unit for these
chest pain patients is the performance of stress testing once an acute
myocardial infarction has been ruled out. An in-hospital stress test
thus removes the variable of patient compliance to outpatient stress
testing in addition to facilitating more urgent additional testing such
as cardiac catheterization.
A recent study has noted the low rates of compliance among
patients who present to the emergency department (ED) with chest
pain and are subsequently referred for outpatient stress testing.
1
Our
objective was to determine the rates of positive stress testing among
low-risk chest pain patients admitted to our observation unit and to
extrapolate the potential missed rates of positive stress tests if these
patients were to be referred instead for outpatient stress testing.
METHODS
We performed a retrospective chart review of all patients
admitted to the University of Utah observation unit under the chest
pain protocol during the 14-month period between April 2006 and
June 2007. The University of Utah Emergency Department is an
urban emergency department with 39,000 patient visits per year. The
observation unit is a 10-bed unit under the direction of the emer-
gency physicians and midlevel providers. The observation unit uses
complaint-driven protocols with strict inclusion and exclusion cri-
teria. The inclusion criteria for a chest pain patient who is deemed
low risk for acute coronary syndrome is a normal or nondiagnostic
EKG, normal or indeterminate troponin I on initial ED testing,
hemodynamic stability, and no ongoing chest pain. A history of
coronary artery disease is not an exclusion criterion (Table 1).
Emergency physicians admitting patients to the observation
unit write the initial order set for testing (ie, serial troponin, stress
testing). Cardiologists evaluate these chest pain patients in the
morning after their observation unit admission and make further
decisions on additional testing (nuclear versus treadmill stress test
versus coronary CTA), cardiac catheterization, and inpatient admis-
sion based on patient history, characteristics, and testing results.
Six fourth-year medical students who had worked in the
emergency department and were familiar with the observation unit
performed the chart review. The reviewers were blinded to the goals
of this study. Reviewers were given a template form and a guide
with definitions of categories for data entry. Data entry was re-
viewed on 20% of patients by one of the study’s lead investigators
(J.B.) to ensure consistent data acquisition between reviewers.
We recorded baseline characteristics, including age, past
medical history, gender, and cardiac history on all patients admitted
to the observation unit. Rates of positive stress test or coronary CT,
in addition to patient disposition (admission to an inpatient unit
versus discharge for outpatient follow-up), were noted. As stress
testing and coronary CT were often used interchangeably as a testing
From the University of Utah, Salt Lake City, UT.
This study was presented at the 2008 Scientific Assembly for the American
College of Emergency Physicians; October 28 –31, 2008; Chicago, IL.
Reprints: Troy E. Madsen, MD, Division of Emergency Medicine, University of
Utah School of Medicine, 30 North 1900 East, 1C26, Salt Lake City, UT
84132. E-mail: troy.madsen@hsc.utah.edu.
Copyright © 2009 by Lippincott Williams & Wilkins
ISSN: 1535-282X/09/0803-0122
DOI: 10.1097/HPC.0b013e3181b00782
Critical Pathways in Cardiology • Volume 8, Number 3, September 2009 122 | www.critpathcardio.com