The Impact of a Concurrent Trauma Alert
Evaluation on Time to Head Computed
Tomography in Patients with Suspected
Esther H. Chen, MD, Angela M. Mills, MD, Bruce Y. Lee, MD, MBA, Jennifer L. Robey, RN, BSN,
Kara E. Zogby, RN, BSN, Frances S. Shofer, PhD, Patrick M. Reilly, MD, Judd E. Hollander, MD
Background: Emergency department (ED) overcrowding threatens quality of care by delaying the time to
diagnosis and treatment of patients with time-sensitive diseases, such as acute stroke.
Objective: The authors hypothesized that the presence of a trauma alert evaluation would impede the time
to head computed tomography (hCT) in patients with stroke-like symptoms.
Methods: This was a secondary analysis of prospectively collected data on patients with potential stroke
who received an hCT in an urban trauma center ED from January 1, 2004, to November 30, 2004. Struc-
tured data collection included historical and examination items, National Institutes of Health (NIH) stroke
scale score, laboratory and radiographic results, and final diagnosis. Admitted patients were followed in
hospital. Patients who presented within one hour following a trauma evaluation were compared with
patients who presented without concurrent trauma for triage time until completion of hCT. Chi-square,
t-tests, and 95% confidence intervals (95% CIs) were used for comparisons.
Results: The 171 patients enrolled had a mean (? standard deviation) age of 60.7 (?7) years; 60% were
female; and 58% were African American. Of these, 72 patients had a significant cerebrovascular event
(38 [22%] ischemic stroke, 25 [15%] transient ischemic attack, seven [4%] intracranial hemorrhage, one
[0.6%] subarachnoid hemorrhage, and one [0.6%] subdural hematoma). The remaining diagnoses included
4.6% migraine, 2.3% seizure, 2.9% syncope, 2.3% Bell’s palsy, and 2.9% vertigo. There was no significant
difference in time to hCT in patients who presented during a trauma activation and those who did not
(99 minutes [interquartile range (IQR) = 24–156] vs. 101 minutes [IQR = 43–151.5]; p = 0.537). In subgroup
analysis of patients with a significant cerebrovascular event, times to hCT were also similar (24 minutes
[IQR = 12–99] vs. 61 minutes [IQR = 15–126]; p = 0.26).
Conclusions: In the authors’ institution, the presence of concurrent trauma evaluation does not delay CT
imaging of patients with potential stroke.
ACADEMIC EMERGENCY MEDICINE 2006; 13:349–352 ª 2006 by the Society for Academic Emergency
Keywords: crowding, emergency service, cerebrovascular accident, trauma alert, head computed
reported their EDs were operating at or above capacity,
and more than 50% of urban hospitals reported times
when they were forced to divert all or specific types of
patients from their EDs.1
Overcrowding threatens patient safety and quality of
care by delaying time to diagnosis and treatment of
critical patients with acute coronary syndromes, stroke,
surgical emergencies, and sepsis.2–4For example, priori-
tizing the care of acutely injured patients compromises
mergency department (ED) overcrowding is a
growing crisis and a burden on the health care
system. In 2002, 90% of Level 1 trauma centers
From the Departments of Emergency Medicine (EHC, AMM,
JLR, KEZ, FSS, JEH) and Trauma and Surgical Critical Care
(PMR), University of Pennsylvania, Philadelphia, PA; and the
Division of General Internal Medicine, University of Pittsburgh
(BYL), Pittsburgh, PA.
Received August 10, 2005; revision received October 14, 2005;
accepted October 14, 2005.
Address for correspondence and reprints: Esther H. Chen, MD,
Department of Emergency Medicine, Ground Floor, Ravdin
Building, Hospital of the University of Pennsylvania, 3400
Spruce Street, Philadelphia, PA 19104-4283. Fax: 215-662-3953;
ª 2006 by the Society for Academic Emergency Medicine
PII ISSN 1069-6563583
the care of concurrent ED patients with acute chest
pain.5Moreover, since trauma evaluations have become
increasingly reliant on computed tomography (CT),
trauma patients had longer ED lengths of stay in 2003
compared with five years earlier.6Increasing CT use by
trauma patients may delay CT scan on other critical ED
patients. Our objective was to determine whether a con-
current trauma evaluation would delay the time to head
CT (hCT) in patients with potential stroke and compro-
mise their care.
This was a secondary analysis of a prospectively col-
lected data set on patients with stroke-like symptoms
from January 1, 2004, to November 31, 2004. The study
was approved by the University of Pennsylvania Commit-
tee on Research Involving Human Subjects.
Study Setting and Population
Patients were enrolled at an urban trauma center ED
with an annual census of approximately 51,000 visits. Pa-
tients were enrolled into the primary study if they were
over 21 years of age with stroke-like symptoms and con-
sented to phlebotomy. Patients from the study population
were included in our analysis if they received an hCT
during their ED evaluation.
At the Hospital of the University of Pennsylvania, a
trauma alert is activated by out-of-hospital personnel.
That alert mobilizes the trauma care team to the trauma
area, and includes an ED nurse assigned to cover the
trauma area, an ED nurse who records all events that oc-
cur during the initial resuscitation phase, radiology and
CT technicians, a respiratory therapist, an ED attending
physician and senior resident, and a trauma attending
physician and trauma residents. The ED staff assesses
and stabilizes the patient’s airway during this initial
resuscitation phase. After the initial assessment, an un-
stable patient or one who requires a surgical intervention
is promptly transported to the operating room by the
trauma team, an ED nurse, and a respiratory therapist
for intubated patients. A stable patient is further evalu-
ated by CT imaging with a dedicated ED scanner while
continuously being monitored by an ED nurse. All trauma
evaluations are entered into a trauma registry, including
time of prenotification, time of patient arrival, triage
and primary survey results, CT evaluation results, and
For patients with potential stroke, an ED triage nurse
and physician promptly perform an initial neurologic as-
sessment and stabilization, and then they directly call the
CT technician to prioritize the patient’s hCT over all other
ED patients. In addition, the stroke team may be notified
if the patient’s symptoms occurred within three hours of
Trained research assistants enrolled ED patients 16
hours per day (8 AM to 12 AM), seven days per week,
which captures 85% to 95% of eligible patients. Informa-
tion collected for each patient included demographics,
historical and physical examination items, National Insti-
tutes of Health (NIH) stroke scale score, laboratory and
radiographic results, and final diagnosis. Hospitalized
patients were followed daily for complications and inter-
ventions. Clinical information was obtained from the
treating physician. Determination of final diagnosis was
made by communication between the investigators and
the health care team. Postdischarge medical record re-
view was not utilized.
All CTs were performed by standard protocol using
a Siemens Sensation 16 CT scanner (Siemens Medical
Solutions USA, Inc., Malvern, PA). A standard hCT may
be completed in 20 seconds, and a standard trauma CT
(head, cervical spine, chest, abdomen, and pelvis) may
be completed in approximately 15 minutes. However,
the actual trauma CT evaluation prevents CT use by
stroke patients for at least 30 minutes because additional
time is required to assemble multiple personnel to trans-
port the patient to the scanner, place and remove the
patient from the table, and clear the blood/body fluids
from the room.
The main outcome measure was the time to hCT, mea-
sured from triage time as recorded in the patient’s elec-
tronic medical record by the triage nurse, to completion
of the hCT as recorded on the patient’s radiology acqui-
sition form entered by the CT technician. Concurrent
trauma evaluation was defined as a case in which a
patient presented within one hour of a trauma patient’s
ED arrival time as recorded from the trauma registry.
Main outcome data are presented as median time with
interquartile ranges (IQRs). Descriptive data are pre-
sented as means (? standard deviations), frequencies,
and percentages, with 95% confidence intervals where
appropriate. To test for differences between patients
who presented during concurrent trauma and those
who did not, the t-test was used for continuous data
and Fisher’s exact test was used for categorical data.
Data were analyzed using SAS statistical software (Ver-
sion 8.2, SAS Institute, Cary, NC).
One hundred seventy-one patients with potential stroke
were enrolled into the study. Patients had a mean age
of 60.8 (?17) years (range 22–95 years) and were more
likely to be female (60%) and African American (59%).
More than half of the study population (56%) had a his-
tory of hypertension, 22% had diabetes mellitus, and
16% had had a prior stroke (Table 1).
Of the study population, 72 patients had a significant
cerebrovascular event, with 38 (22.2%) suffering an
ischemic stroke, 25 (14.6%) a transient ischemic attack,
and nine (5.3%) cerebrovascular hemorrhage or hema-
toma. The remaining patients had a less emergent neuro-
logic diagnosis or other medical diagnosis (Table 1).
Patients who presented during concurrent trauma
activation (n = 39) had a similar median time to hCT
compared with patients evaluated without concurrent
trauma (n = 132) (99 min [IQR = 24–156] vs. 101 min
Chen et al.?TRAUMA ALERT AND CT FOR STROKE
[IQR = 43–151.5]; p = 0.537). There was also no significant
difference in time to hCT in the subset of patients (n = 72)
diagnosed as having a serious cerebrovascular disease
(24 min [IQR = 12–99] vs. 61 min [IQR = 15–126]; p =
0.26). However, compared with patients who had a less
serious neurologic or medical condition (n = 99), patients
with a significant cerebrovascular disease received their
hCT more quickly (58.5 min [IQR = 14–120] vs. 122 min
[IQR = 55–188]; p < 0.001). One patient left before ED eval-
uation was complete.
Although out-of-hospital delay significantly affects the
management of stroke patients,7emergency physicians
may mitigate its effects by facilitating the ED care of
these patients, including stroke team evaluation and
immediate CT imaging, and preventing additional treat-
ment delay. Individual physicians have little control over
out-of-hospital delays, but they can potentially have
an impact on radiographic imaging delays. Because
this stroke evaluation depends on the availability of re-
sources (i.e., stroke team, CT scanner, triage nurse, and
transporter) at the time of patient arrival, our study
examined the impact of allocating the same resources
to concurrent critically ill ED patients. Based on previous
studies showing the negative impact of expedited trauma
care on other critical patients,2,5we hypothesized that
prioritizing a 30-minute trauma CT evaluation might
delay the time to hCT, thereby placing stroke patients
outside the three-hour window for thrombolysis.
However, in our institution, concurrent trauma evalua-
tions did not delay the time to hCT in potential stroke pa-
tients. In fact, subgroup analysis of the 72 (42%) patients
with a significant cerebrovascular injury showed that a
concurrent trauma evaluation actually decreased their
time to hCT from 61 minutes to 24 minutes. They also
received their hCTs more quickly than did patients with
less serious neurologic or medical diagnoses. Similar
findings were shown in a previous chest pain study in
which trauma evaluations did not delay time to electro-
cardiogram or door-to-needle time, but instead occurred
in later phases, such as final disposition.5
There are several possible explanations for these find-
ings. Because emergency physicians are trained to triage
and risk-stratify patients quickly, they are more likely to
expedite the hCT in those patients for which they have
a high clinical suspicion for an acute stroke, especially
if there has been adequate prenotification about an
incoming trauma patient. Depending on the stability of
the respective patients, the trauma and ED attending
will then determine how to best allocate resources.
Moreover, new-generation CT scanners allow hCTs to
be completed in less than 5 minutes, including transport
time to and from the CT suite. For most stable trauma
patients, delaying their CT evaluation for a few minutes
may not significantly affect their clinical outcomes.
Additionally, access to a single ED scanner may not
cause a bottleneck in the assessment of potential stroke
patients, because much of the patient’s time to hCT
and transporting the patient, activities not necessarily
affected by the presence of concurrent trauma patients.
This study was conducted in one institution with a
specific trauma system structure with a single ED CT
scanner and may not be applicable to other EDs. In
addition, although we used broad inclusion criteria to
ensure generalizability of the data, our sample size was
small, and patients with a significant cerebrovascular
disease who presented during a trauma evaluation were
Competition for a single CT scanner for trauma patients
did not appear to delay the CT evaluation of potential
stroke patients, as the emergency physicians and ED
system appeared to effectively coordinate use of the CT.
Therefore, other steps in ED stroke evaluations such as
triage, assessment, and transport should be evaluated
as possible causes of delays in patient care. Furthermore,
future studies could determine whether concurrent
trauma evaluations affect the time to CT for diagnoses
less urgent than potential stroke. Such studies would
help not only optimize the use of limited resources, but
also reduce ED overcrowding, while expediting and
improving patient care.
Age—mean ? SD
60.8 ? 17 yr
Serious cerebrovascular event
Transient ischemic attack
Other neurologic disease
Altered mental status
Adverse drug reaction
Other medical disease
ACAD EMERG MED?March 2006, Vol. 13, No. 3?www.aemj.org
1. Lewin Group (for the American Hospital Association).
Emergency Department Overload: A Growing Crisis.
The Results of the AHA Survey of Emergency Depart-
ment (ED) and Hospital Capacity. Falls Church, VA:
Lewin Group, 2002.
2. Derlet RW, Richards JR. Emergency department over-
crowding in Florida, New York, and Texas. South Med
J. 2002; 95:846–9.
3. Derlet RW, Richards JR. Overcrowding in the nation’s
emergency departments: complex causes and disturb-
ing effects. Ann Emerg Med. 2000; 35:63–8.
4. Schull MJ, Vermeulen M, Slaughter G, Morrison L,
Daly P. Emergency department crowding and throm-
bolysis delays in acute myocardial infarction. Ann
Emerg Med. 2004; 44:577–85.
5. Boutros F, Redelmeier DA. Effects of trauma cases on
the care of patients who have chest pain in an emer-
gency department. J Trauma. 2000; 48:649–53.
6. Reilly PM, Schwab CW, Kauder DR, et al. The invisible
trauma patient: emergency department discharges.
J Trauma. 2005; 58:675–83; discussion 683–5.
7. Evenson KR, Rosamond WD, Morris DL. Prehospital
and in-hospital delays in acute stroke care. Neuroepi-
demiology. 2001; 20(2):65–76.
Chen et al.?TRAUMA ALERT AND CT FOR STROKE