Full text of DOI:10.1080/10903120190939463

A COMPARISON OF EMERGENCY MEDICAL SERVICES
TIMES FOR STROKE AND MYOCARDIAL INFARCTION
Kelly R. Evenson, PhD, Emily B. Schroeder, MSPH, Trent B. Legare, BA,
Jane H. Brice, MD, MPH, Wayne D. Rosamond, PhD, Dexter L. Morris, PhD, MD
ABSTRACT
The American Heart Association emphasizes the
importance of treating both stroke and myocardial
infarction (MI) as emergencies because of the time-
dependent nature of medical therapies for both condi-
tions.1 Goals set by the American Heart Association
include increasing the percentage of stroke patients
arriving at the hospital within three hours from symp-
tom onset to 20% and increasing the percentage of MI
patients arriving at the hospital within one hour of
symptom onset to 20% and within six hours of symp-
tom onset to 90%.1
The chain of events from the onset of symptoms to
treatment involves several steps, each contributing to
overall delay time. The emergency medical services
(EMS) system plays a key role in helping patients
reach medical care quickly. Since stroke symptoms are
often vague, and acute therapies for stroke are more
recently available, stroke patients may not be consid-
ered by EMS with the same urgency as MI patients. In
some circumstances, dispatchers may use a nonemer-
gent response mode because the symptoms communi-
cated to them may not seem urgent or may not be rec-
ognized.2 A better understanding of how EMS
responds to calls from patients with stroke and MI
could lead to professional and public interventions
aimed at making progress towards the American
Heart Association goals. The objective of this study
was to compare EMS transport times for stroke and
MI patients who used a county-based EMS system
during 1999.
Objective. Since stroke symptoms are often vague, and
acute therapies for stroke are more recently available, it has
been hypothesized that stroke patients may not be treated
with the same urgency as myocardial infarction (MI)
patients by emergency medical services (EMS). To examine
this hypothesis, EMS transport times were examined for
both stroke and MI patients who used a paramedic-level,
county-based EMS system for transportation to a single hos-
pital during 1999. Methods. Patients were first identified by
their hospital discharge diagnosis as stroke (ICD-9 430–436,
n = 50) or MI (ICD-9 410, n = 55). Trip sheets with corre-
sponding transport times were retrospectively obtained
from the 911 center. A separate analysis was performed on
patients identified by dispatchers with a chief complaint of
stroke (n = 85) or MI (n = 372). Results. Comparing stroke
and MI patients identified by ICD-9 codes, mean EMS trans-
port times in minutes did not meaningfully differ with
respect to dispatch to scene arrival time (8.3 vs 8.9, p = 0.61),
scene time (19.5 vs 21.4, p = 0.23), and transport time (13.7 vs
16.2, p = 0.10). Mean total call times in minutes from dis-
patch to hospital arrival were similar between stroke and MI
patients (41.5 vs 46.4, p = 0.22). Results were similar when
comparing patients identified by dispatchers with a chief
complaint indicative of stroke or MI. Conclusion. In this sin-
gle county, EMS response times were not different between
stroke and MI patients. Replication in other EMS settings is
needed to confirm these findings. Key words: cerebrovascu-
lar disorders; emergency medical services; myocardial
infarction; stroke; time factors.
PREHOSPITAL EMERGENCY CARE 2001;5:335–339
Received March 7, 2001, from the Department of Epidemiology,
School of Public Health (KRE, EBS, TBL, WDR), and the
Department of Emergency Medicine, School of Medicine (JHB,
DLM), University of North Carolina–Chapel Hill, Chapel Hill,
North Carolina. Revision received April 25, 2001; accepted for pub-
lication April 25, 2001.
METHODS
The University of North Carolina (UNC) School of
Medicine Institutional Review Board approved the
protocol for this study. Patients discharged from the
UNC Hospitals with a stroke or MI diagnosis during
1999 and who were transported to the hospital by
Orange County, North Carolina, EMS were identified
from hospital records. Eligible cases were defined
using hospital discharge diagnosis codes (Inter-
national Classification of Diseases version 9, or ICD-
9). Discharge diagnosis codes 430–436 were used to
define stroke events, and 410 for MI events. If the
patient had more than one stroke discharge diagnosis
code, then the lowest numeric ICD-9 code was
assigned.
Presented in poster form at the Research Educational Support
Conference, Chapel Hill, North Carolina, July 1999; and in abstract
form at the American Heart Association International Stroke
Conference, Fort Lauderdale, Florida, February 2001.
Supported in part by the North Carolina Department of Health and
Human Services, Cardiovascular Health Data Unit. Dr. Evenson
and Ms. Schroeder were funded by NIH, NHLBI, NRSA Training
Grant #5-T32-HL007055. Mr. Legare was funded by the UNC-CH
Minority Student Development Program (Research Education
Support Program) #5-R25-GM55336.
Address correspondence and reprint requests to: Kelly Evenson,
PhD, Department of Epidemiology, School of Public Health, 137
East Franklin Street, Suite 306, Chapel Hill, NC 27514. e-mail:
<kelly_evenson@unc.edu>.
Orange County EMS is a nontraditional system that
335

336
PREHOSPITAL EMERGENCY CARE OCTOBER/DECEMBER 2001 VOLUME 5 / NUMBER 4
TABLE 1. Chief Complaints Reported to the
Paramedic (in Percentages*) among Patients
Identified by Hospital Discharge Diagnosis
of arrival to the hospital, then the corresponding time
was taken from the ambulance trip sheet. All other
times were taken from the medic trip sheet only. For
patients with one or more missing times, the times
were extracted from the call-for-service reports.
The priority of each call, as determined by the 911
operator, was categorized as either with or without
lights and sirens. In this county, use of lights and
sirens to the scene indicated a high-priority, life-
threatening emergency. The protocol specified that if
the dispatcher suspected a stroke or MI, the para-
medic was dispatched at highest priority (lights and
sirens) and the ambulance at medium priority (no
lights and sirens). The chief complaint was based on
classification made by the 911 operator from caller
information. The chief complaint was communicated
to the paramedic and ambulance en route to the scene.
Once reaching the scene, the paramedic decided
whether an ambulance was needed and whether
transport to the hospital warranted lights and sirens.
Triage was defined by the paramedic as urgent (dead
or imminent mortal condition or need for physician
evaluation within one to four hours) and nonurgent
(need for physician evaluation within 24 hours). The
triage level was used to communicate with emergency
department personnel.
Patients with a dispatch diagnosis of “CVA/stroke”
(CVA = cerebrovascular accident) and “rule out
MI/chest pain” were identified for separate analyses
by obtaining all call-for-service reports for patients
with a dispatch diagnosis indicative of stroke or MI.
These reports reside on a computerized database for
all calls into the EMS system for the county. Arrival,
scene, transport, and total call times were defined
identically for these patients as compared with
patients identified by ICD-9 codes. All 911 calls that
did not result in transport to the hospital were exclud-
ed. Other covariates of interest were not available for
analysis of these patients (e.g., age, gender, race, use of
lights and sirens, triage).
Myocardial
Stroke
Infarction (MI)
(n = 47)
(n = 54)
Cerebrovascular accident/stroke
Sick call
Breathing problems
Unconsciousness
Rule out MI/chest pain
Fall
Unresponsiveness
Shortness of breath
Diabetes-related complaints
Hemorrhage
Abdominal pain
Other†
42.6
10.6
12.8
6.4
4.3
4.3
4.3
2.1
2.1
0
3.7
9.3
16.7
1.9
48.2
0
1.9
11.1
1.9
1.9
1.9
1.9
0
10.6
*Percentages may not add up to 100% due to rounding.
†Other category includes chief complaints identified for only one patient.
sends paramedics in initial response vehicles to every
call for medical assistance. The dispatcher receiving
the call for assistance determines the mode of
response (e.g., lights and sirens) and the number of
additional resources required (e.g., first responder,
transporting basic life support ambulance) based on
emergency medical dispatch cards (Advanced
Medical Priority Dispatch System). The estimated
total population of Orange County as of July 1999 was
109,746, with 82% white race.3 The county covers 400
square miles. From July 1999 to June 2000, the dis-
patch center received 152,357 calls for assistance.
Information pertaining to each trip is routinely
recorded on standardized forms by both the medic
evaluating the patient and the transporting personnel.
The data were collected retrospectively and EMS
providers were unaware of the study. During the
study period there was no additional training or pub-
lic campaign for emergency stroke and MI treatment,
other than the standard practice in this county. The
EMS trip sheets for all eligible patients identified in
the search of hospital discharge codes were abstracted
by research personnel. In this county, timing of events
is recorded by EMS personnel on the trip sheet with
times provided by the dispatcher using a computer-
aided dispatch system.
Arrival time was defined as the time from EMS dis-
patch until EMS arrival at the scene. Scene time was
defined as the time from arrival of the first EMS unit
until the patient was taken from the scene. Transport
time was defined as the time from when the EMS unit
left the scene until it arrived at the emergency depart-
ment. Total call time was defined as the time from
EMS dispatch until the EMS unit reached the emer-
gency department. Times were taken first from the
medic trip sheet. If the medic trip sheet was missing
either the time of departure from the scene or the time
Statistical Analyses
We excluded patients who walked into the EMS cen-
ter and patients who had both a stroke and MI diag-
nosis according to the ICD-9 codes. We also excluded
one incident involving a disabled EMS vehicle. Means
were compared with the nonparametric Wilcoxon
two-sample test and categorical data were compared
with the Mantel-Haenszel chi-square statistic. Multi-
variable linear regression models were used to exam-
ine whether having a stroke or MI ICD-9 diagnosis
predicted shorter arrival time, scene time, transport
time, and total call time separately. These statistical
models controlled for age, gender, race, triage, and
use of lights and sirens to the scene. In addition, for
the models predicting transport time and total call

Evenson et al. EMS TIMES FOR STROKE AND MI
337
TABLE 3. Mean Prehospital Delay Times (in Minutes;
Interquartile Ranges in Parentheses) by “CVA/Stroke” and
“Rule Out MI/Chest Pain” Dispatch Chief Complaint*
time, use of lights and sirens to the hospital was also
included. All analyses were performed using SAS
Version 6.12 (SAS Institute, Inc., Cary, NC).
“Rule Out MI/
“CVA/Stroke”
Chest Pain”
(n = 372)
RESULTS
(n = 85)
p-value
There were 50 stroke patients included with the fol-
lowing discharge diagnoses: ten hemorrhagic stroke
(ICD-9 430–432), 38 ischemic stroke or transient cere-
bral ischemia (ICD-9 433–435), and two acute but ill-
defined cerebrovascular disease (ICD-9 436). There
were 55 acute MI diagnoses (ICD-9 410). Stroke
patients were on average older than MI patients
(mean 78.2 vs 72.6 years, p = 0.03), but there was no
significant difference by gender (54.0% vs 54.6%
female, p = 0.96) or ethnicity (77.6% vs 68.5% white, p
= 0.31). For stroke and MI patients, uses of lights and
sirens to the scene (83.7% vs 90.7%, p = 0.28) and to the
hospital (18.4% vs 13.0%, p = 0.45) were also similar.
The MI patients were more likely to be triaged as
urgent (89.1%) compared with the stroke patients
(73.5%, p = 0.04). Table 1 shows the dispatch chief
complaints communicated to the paramedic for stroke
and MI patients identified by ICD-9 codes. Among the
stroke patients, 42.6% were dispatched as “CVA/
stroke” and among the MI patients, 48.2% were dis-
patched as “rule out MI/chest pain.”
Mean arrival times, scene times, transport times,
and total call times were similar for the stroke and MI
patients identified by ICD-9 codes (Table 2). The pat-
terns were further examined using multivariable lin-
ear regression models, controlling for age, gender, eth-
nicity, triage, and the use of lights and sirens.
Comparing MI patients with stroke patients, arrival
times, scene times, transport times, and total call times
were not significantly different. Additionally, trans-
port times did not differ by age, gender, or ethnicity.
The use of lights and sirens to the scene was associat-
ed with a shorter arrival time (–4.4 minutes, 95% CI
–7.5, –1.3), scene time (–9.3 minutes, 95% CI –13.7,
–4.9), and total call time (–12.7 minutes, 95% CI –21.9,
–3.5) as compared with not using lights and sirens to
the scene. However, the use of lights and sirens to the
hospital was not significantly associated with trans-
port time or total call time. Omitting triage and use of
lights and sirens from the models did not change the
result for the stroke vs MI time comparisons.
Arrival time
Scene time
Transport time
Total call time
7.9 (5–11)
20.3 (15–24)
17.0 (10–22)
44.7 (36–55)
7.6 (5–10)
19.0 (14–23)
16.6 (9–24)
42.8 (33–51)
0.51
0.19
0.84
0.21
*CVA = cerebrovascular accident; MI = myocardial infarction.
Time intervals for patients with a dispatch chief
complaint of “CVA/stroke” (n = 85) and “rule out
MI/chest pain” (n = 372) were also examined, regard-
less of whether their corresponding discharge diagno-
sis was assigned as a stroke, MI, or an unrelated diag-
nosis. Comparing “CVA/stroke” with “rule out MI/
chest pain,” mean EMS transport times were similar
for arrival time, scene time, transport time, and total
call time (Table 3).
DISCUSSION
Timely access to medical care for acute stroke and MI
patients includes several steps: rapid recognition of
the problem by the patient or bystander, contact with
and access to medical care, identifying the problem at
the highest level of emergency, transportation to the
emergency department, evaluation in the emergency
department, and initiation of appropriate treatment.4
A significant delay in any step results in limiting treat-
ment options for patient care. This study found that
the time it takes for EMS to arrive at the scene after a
911 call is received by dispatchers was short and rep-
resented a small portion of the total prehospital delay
reported in the literature for both stroke5 and MI6
patients. However, for patients with a stroke or MI
discharge diagnosis, the mean total call time was
approximately 44 minutes, which represents a clini-
cally significant portion of the treatment window,
especially for ischemic stroke patients (three hours).
Lights and sirens to the scene decreased arrival, scene,
and total call times for stroke and MI patients, which
may be due to heightened urgency because of the
higher dispatch priority.
The use of EMS is generally associated with earlier
arrival to the hospital for both stroke7–13 and MI14–16
patients. For stroke patients, use of 911 is also associ-
ated with decreased in-hospital times, such as time to
emergency physician evaluation,7,8,11,17,18 time to a
computed tomography (CT) scan,8,17 and time to neu-
rology consultation.8 For MI patients, prolonged pre-
hospital delay is an independent predictor of longer
in-hospital delay.15 Despite the importance of EMS in
identifying, treating, and transporting patients, few
TABLE 2. Mean Prehospital Delay Times (in Minutes;
Interquartile Ranges in Parentheses) by Stroke and
Myocardial Infarction (MI) Hospital Discharge Diagnosis
Stroke
MI
(n = 50)
(n = 55)
p-value
Arrival time
Scene time
Transport time
Total call time
8.3 (4–11)
19.5 (15–23)
13.7 (8–16)
41.5 (30–49)
8.9 (5–12)
21.4 (17–25)
16.2 (9–23)
46.4 (33–52)
0.61
0.23
0.10
0.22

338
PREHOSPITAL EMERGENCY CARE OCTOBER/DECEMBER 2001 VOLUME 5 / NUMBER 4
data exist on how well the system meets these needs.18
The EMS designation for stroke has historically not
been placed at the highest response level, although an
acute life support response for stroke is appropriate.19
This study found that transport times did not signif-
icantly differ between stroke and MI patients.
Although the differences between stroke and MI are
numerous, we chose to compare their EMS times due
to the similarities of the populations at risk and the
availability of emergent treatment.20 Several points
must be considered when making comparisons across
these diagnoses. First, the tissue plasminogen activa-
tor (t-PA) treatment time window for ischemic stroke
is within three hours of symptom onset,21 yet for MI it
may be up to 12 hours.22 Although the earlier throm-
bolytics are administered in the case of MI, the greater
the reduction in infarct size.23 Second, the delays asso-
ciated with the clinical diagnosis of acute ischemic
stroke (e.g., CT scan required) are often greater than
those associated with the diagnosis of acute MI.24
Third, patients with MI typically present with more
focal or localized symptoms, which often include
pain, while stroke patients can present with vague
symptoms, often with impaired communication and
perception without pain.4
half of MI patients were dispatched as “rule out
MI/chest pain.” The EMS response times were not dif-
ferent comparing stroke and MI patients identified
either by dispatch chief complaint or by discharge
diagnosis. However, any delay in elapsed EMS time
represents a greater proportion of the thrombolytic
treatment window for ischemic stroke as compared
with MI patients. Approximately one fourth of the
calls for both stroke and MI patients were treated
nonurgently. Additionally, the scene time represented
the largest EMS time interval for both stroke and MI
patients and may offer the greatest opportunity for
reducing total call time. The EMS system is a crucial
part of the chain of recovery for both stroke and MI.
While there appears to be little difference between
stroke and MI transport times, these times were a sig-
nificant part of the therapeutic window, leaving open
important areas for improvement through interven-
tion in the care of these patients. A comprehensive
intervention for EMS personnel should include train-
ing in recognition of stroke and MI diagnoses, as well
as acting with speed and urgency in the treatment and
transport of these patients.
The authors acknowledge Orange County Emergency Management
for their assistance.
LIMITATIONS
References
Several limitations to this study should be acknowl-
edged. The dispatcher processing time was not col-
lected, but is likely to be less than 2 minutes. Triage
algorithms were used to identify and prioritize
patients by dispatchers, which may provide more
timely and efficient quality of care. It is not known
how response times may have changed if this system
was not in place. In the multivariable models we were
unable to control for distance (i.e., mileage) to the
scene and to the hospital, which probably varied
widely due to the geographic area covered by this sin-
gle EMS system. Also, severity was not directly con-
trolled for as a potential confounder in the analyses,
although triage and use of lights and sirens may have
served as a proxy for it. Additionally, using hospital
discharge codes to define stroke or MI will likely mis-
classify some cases, more often for stroke.25,26 than for
MI patients.27 Our definition of stroke combined both
hemorrhagic and ischemic stroke events, due to limit-
ed statistical power to examine them separately. The
generalizability of these findings is limited because it
included only one county EMS system with a relative-
ly advanced dispatch system. These findings should
be confirmed in other EMS settings.
1.American Heart Association. American Heart Association
strategic plan. Am Heart Assoc News (Dallas, TX). 1999:1,3.
2.Smith M, Doliszny K, Shahar E, McGovern P, Arnett D, Luepker
R. Delayed hospital arrival for acute stroke: The Minnesota
Stroke Survey. Ann Intern Med. 1998;129:190-6.
3.North Carolina Office of State Planning, 2000. Website: http://
osbpm.state.nc.us/
4.Broderick J. Practical considerations in the early treatment of
ischemic stroke. Am Fam Physician. 1998;57:73-80.
5.Evenson K, Rosamond W, Morris D. Prehospital and in-hospital
delays in acute stroke care. Neuroepidemiology. 2001;20:65-76.
6.Simons-Morton D, Goff D, Osganian S, et al. Rapid Early Action
for Coronary Treatment: rationale, design, and baseline charac-
teristics. Acad Emerg Med. 1998;5:726-38.
7.Lacy C, Suh D, Bueno M, Kostis J, for the STROKE Collaborative
Study Group. Delay in presentation and evaluation for acute
stroke: Stroke Time Registry for Outcomes Knowledge and
Epidemiology (STROKE). Stroke. 2001;32:63-9.
8.Schroeder E, Rosamond W, Morris D, Evenson K, Hinn A.
Determinants of emergency medical services use in a popula-
tion with stroke symptoms: The Second Delay in Accessing
Stroke Healthcare (DASH II) Study. Stroke. 2000;31:2591-6.
9.Kothari R, Jauch E, Broderick J, et al. Acute stroke: delays to
presentation and emergency department evaluation. Ann
Emerg Med. 1999;33:3-8.
10.Rosamond W, Gorton R, Hinn A, Hohenhaus S, Morris D. Rapid
response to stroke symptoms: The Delay in Accessing Stroke
Healthcare (DASH) Study. Acad Emerg Med. 1998;5:45-51.
11.Menon SC, Pandey DK, Morganstern LB. Critical factors in
determining access to acute stroke care. Neurology. 1998;51:427-
32.
CONCLUSION
In this single county, almost half of stroke patients
were dispatched as having “CVA/stroke” and almost
12.Wester P, Radberg J, Lundgren B, Peltonen M, for the Seek-
Medical-Attention-in-Time Study Group. Factors associated
with delay admission to hospital and in-hospital delays in acute

Evenson et al. EMS TIMES FOR STROKE AND MI
339
stroke and TIA. Stroke. 1999;30:40-8.
(eds). Proceedings of
a
National Symposium on Rapid
13.Williams L, Bruno A, Rouch D, Marriott D. Stroke patients’ Identification and Treatment of Acute Stroke. Bethesda, MD:
knowledge of stroke: influence on time to presentation. Stroke.
1997;28:912-5.
National Institutes of Health, 1997, pp 119-25.
21.National Institute of Neurological Disorders and Stroke rt-PA
14.Meischke H, Ho M, Eisenberg M, Schaeffer S, Larsen M. Stroke Study Group. Tissue plasminogen activator for acute
Reasons patients with chest pain delay or do not call 911. Ann
Emerg Med. 1995;25:193-7.
ischemic stroke. N Engl J Med. 1995;333:1581-7.
22.Committee on Management of Acute Myocardial Infarction.
15.Berglin Blohm M, Hartford M, Karlsson T, Herlitz J. Factors 1999 update: ACC/AHA guidelines for the management of
associated with pre-hospital and in-hospital delay time in acute
myocardial infarction: a 6-year experience. J Intern Med.
1998;243:243-50.
patients with acute myocardial infarction: executive summary
and recommendations. Circulation. 1999;100:1016-30.
23.Newby L, Rutsch W, Califf R, et al. Time from symptom onset
16.Hartford M, Herlitz J, Karlson B, Risenfors M. Components of to treatment and outcomes after thrombolytic therapy. J Am
delay time in suspected acute myocardial infarction with partic-
ular emphasis on patient delay. J Intern Med. 1990;228:519-23.
Coll Cardiol. 1996;27:1646-55.
24.Kasner S, Grotta J. Emergency identification and treatment of
17.Morris D, Rosamond W, Hinn A, Gorton R. Time delays in acute ischemic stroke. Ann Emerg Med. 1997;30:642-53.
accessing stroke care in the emergency department. Acad
Emerg Med. 1999;6:218-23.
25.Leibson C, Naessens J, Brown R, Whisnant J. Accuracy of hos-
pital discharge abstracts for identifying stroke. Stroke.
18.Bratina P, Greenberg L, Pasteur W, Grotta J. Current emergency 1994;25:2348-55.
department management of stroke in Houston, Texas. Stroke.
1995;26:409-14.
26.Rosamond W, Folsom A, Chambless L, et al. Stroke incidence
and survival among middle aged adults: 9-year follow-up of the
19.Porteous G, Cory M, Smith W. Emergency medical services dis- Atherosclerosis Risk in Communities (ARIC) cohort. Stroke.
patcher identification of stroke and transient ischemic attack.
Prehosp Emerg Care. 1999;3:211-6.
1999;30:736-43.
27.Petersen L, Wright S, Normand S, Daley J. Positive predictive
20.Spilker J. The importance of patient and public education in value of the diagnosis of acute myocardial infarction in an
acute ischemic stroke. In: Marler J, Winters Jones P, Emr M administrative database. J Gen Intern Med. 1999;14:555-8.
≡≡≡
SIRENS
Winter’s Depression
Jingle Blues. Jingle Blues, Jingling
all the way. Mostly night,
sometimes day.
No what fun it is to ride Dr. Rosen’s one-horse
open chaise, jingling jingling all the way.
I pay Dr. Rosen all my dough. Let it snow.
Let it snow. Let it snow.
Up and down I go, a Bi-Polar bear
slip-sliding the snow.
Jingle Blues. Jingle Blues.
So Dr. Rosen can pay AMA dues.
Dr. Rosen, with your mouth hung open,
it makes you look so frozen.
Dr. Rosen! Give me a pill. Send me the bill.
Dr. Rosen! Dr. Rosen!
DR. ROSEN!
DON SHIELDS, PhD
Maquoketa, Iowa