A Method for Improving Arrival-to- electrocardiogram Time in by lindayy


A Method for Improving Arrival-to- electrocardiogram Time in ...

More Info

A Method for Improving Arrival-to-
electrocardiogram Time in Emergency
Department Chest Pain Patients and the Effect
on Door-to-balloon Time for ST-segment
Elevation Myocardial Infarction
Kevin M. Takakuwa, MD, Gregory A. Burek, Adrian T. Estepa, and Frances S. Shofer, PhD

           Objectives: The objectives were to determine if an emergency department (ED) could improve the
           adherence to a door-to-electrocardiogram (ECG) time goal of 10 minutes or less for patients who
           presented to an ED with chest pain and the effect of this adherence on door-to-balloon (DTB) time for
           ST-segment elevation myocardial infarction (STEMI) cardiac catheterization (cath) alert patients.
           Methods: This was a planned 1-month before-and-after interventional study design for implementing a
           new process for obtaining ECGs in patients presenting to the study ED with chest pain. Prior to the
           change, patients were registered and triaged before an ECG was obtained. The new procedure required
           registration clerks to identify those with chest pain and directly overhead page or call a designated ECG
           technician. This technician had other ED duties, but prioritized performing ECGs and delivering them to
           attending physicians. A full registration process occurred after the clinical staff performed their initial
           assessment. The primary outcome was the total percentage of patients with chest pain who received an
           ECG within 10 minutes of ED arrival. The secondary outcome was DTB time for patients with STEMI
           who were emergently cath alerted. Data were analyzed using mean differences, 95% confidence intervals
           (CIs), and relative risk (RR) regression to adjust for possible confounders.
           Results: A total of 719 patients were studied: 313 before and 405 after the intervention. The mean
           (±standard deviation [SD]) age was 50 (±16) years, 54% were women, 57% were African American, and
           36% were white. Patients walked in 89% of the time; 11% arrived by ambulance. Thirty-nine percent
           were triaged as emergent and 61% as nonemergent. Patients presented during daytime 68% of the time,
           and 32% presented during the night. Before the intervention, 16% received an ECG at 10 minutes or
           less. After the intervention, 64% met the time requirement, for a mean difference of 47.3% (95%
           CI = 40.8% to 53.3%, p < 0.0001). Results were not affected by age, sex, race, mode of arrival, triage clas-
           sification, or time of arrival. For patients with STEMI cath alerts, four were seen before and seven after
           the intervention. No patients before the intervention had ECG time within 10 minutes, and one of four
           had DTB time of <90 minutes. After the intervention, all seven patients had ECG time within 10 minutes;
           the three arriving during weekday hours when the cath team was on site had DTB times of <90 minutes,
           but the four arriving at night and on weekends when the cath team was off site had DTB times of
           >90 minutes.
           Conclusions: The overall percentage of patients with a door-to-ECG time within 10 minutes improved
           without increasing staffing. An ECG was performed within 10 minutes of arrival for all patients who
           were STEMI cath alerted, but DTB time under 90 minutes was achieved only when the cath team was on
           ACADEMIC EMERGENCY MEDICINE 2009; 16:921–927 ª 2009 by the Society for Academic Emergency
           Keywords: chest pain, electrocardiogram, ECG, door-to-ECG time, door-to-balloon time

From the Department of Emergency Medicine, Thomas Jefferson University Hospital (KMT, GAB, ATE); and the Department of
Emergency Medicine, University of Pennsylvania Health System (FSS), Philadelphia, PA.
Received January 27, 2009; revisions received April 22, May 1, and May 4, 2009; accepted May 10, 2009.
Presented at the 12th International Conference on Emergency Medicine, San Francisco, CA, April 3, 2008.
Address for reprints or correspondence: Kevin M. Takakuwa, MD; e-mail: Kevin.Takakuwa@jefferson.edu.

ª 2009 by the Society for Academic Emergency Medicine                                                    ISSN 1069-6563
doi: 10.1111/j.1553-2712.2009.00493.x                                                             PII ISSN 1069-6563583   921
      Copyright Agency Limited (CAL) licenced copy
922                                                                                           Takakuwa et al.   •   ECG TIME

        he guidelines of the American College of Cardiol-
                                                             Study Setting and Population
        ogy and the American Heart Association
                                                             This study was conducted at Thomas Jefferson Univer-
        (ACC ⁄ AHA) specify that a 12-lead electrocardio-
                                                             sity Hospital, an urban Level 1 trauma center and ter-
gram (ECG) should be obtained and interpreted as soon
                                                             tiary care hospital with an ED that has an annual
as possible after a patient arrives to an emergency
                                                             census of more than 60,000 visits. All patients aged
department (ED) with chest discomfort or symptoms sus-
                                                             18 years and older who presented to our ED between
picious for acute coronary syndrome (ACS).1,2 They
                                                             September 1, 2007, and October 31, 2007, with any
specify a time goal of within 10 minutes of arrival, which
                                                             mention or complaint of chest pain were included.
may be an optimistic goal. In contrast, the Veterans
Health Administration measures the percentage of
                                                             Study Protocol
patients with cardiac symptoms who undergo an ECG
                                                             In 2007, facing prolonged ECG and DTB times, our
within 10 minutes and do not measure the duration of
                                                             hospital’s Chest Pain Center Committee, a working
the extra step of ECG interpretation.3 The ACC ⁄ AHA
                                                             group composed of the leaders of emergency medi-
further recommends that patients with ST-segment ele-
                                                             cine, interventional cardiology, the cardiac care unit,
vation myocardial infarction (STEMI) receive percutane-
                                                             ED and hospital nursing, and hospital administrators,
ous coronary intervention (PCI) in 90 minutes or less.4
                                                             recognized the importance of improving door-to-ECG
   A study reporting on 63,478 high-risk non-ST-seg-
                                                             times to within 10 minutes as a way to improve DTB
ment elevation ACS patients revealed that only 33% had
                                                             time, with a DTB goal of 90 minutes or less. We began
ECGs in less than 10 minutes of arrival; the median
                                                             tracking the time to ECG for all patients who com-
ECG time was 15 minutes.5 Another study of 8,885 ED
                                                             plained of chest pain during registration or triage
visits showed that 34% of non-ST-elevation ACS and
                                                             early in 2007. We had already been tracking DTB time
40.9% of STEMI patients had an ECG within 10 minutes
                                                             for all emergency cath alerts for years prior to the
of arrival.6 The study further demonstrated an increased
                                                             new intervention. A cath alert in our hospital occurs
risk of adverse clinical outcomes for STEMI patients
                                                             when an ED patient presents with a STEMI or new
who had ECGs greater than 10 minutes after arrival. A
                                                             left bundle branch block (LBBB) that triggers a rapid
study of undifferentiated chest pain patients showed
                                                             response to move the patient to our cardiac cath labo-
that whites received ECGs faster than nonwhites, and
                                                             ratory for PCI.
men received ECGs faster than women, although these
differences disappeared for ACS patients.7
                                                             Before Intervention. During this time, when a patient
   The benefit of a rapid initial ECG is intuitive for a
                                                             came to our ED, he or she would be directed to a regis-
number of reasons. First, the ECG results direct subse-
                                                             tration desk for a brief registration process that
quent care down different pathways. Logically, ECG
                                                             included name, date of birth, and chief complaint. If the
delays can lengthen time to other therapies such as
                                                             patient mentioned chest pain, the registration clerk
thrombolysis and cardiac catheterization (cath). ECG
                                                             would send the patient to triage. The triage nurse
changes (ST-segment elevation and new Q waves)
                                                             would then assess the patient and order an ECG if ACS
increase the likelihood of ACS, while normal ECG
                                                             was suspected. The ECG would be taken by any avail-
results decrease the chances of ACS.8 The National
                                                             able technician and could be delayed if the technician
Heart Attack Alert Program Coordinating Committee
                                                             decided to insert an intravenous line, draw up blood
has made suggestions for improving ECG time, includ-
                                                             tests, and deliver the blood to the nurse assigned to the
ing standing orders to perform ECGs, housing ECG
                                                             patient. The technician was responsible for delivering
machines in EDs, and having ECG technicians available
                                                             the ECG to an attending physician for interpretation,
within 5 minutes of paging.9 However, neither they nor
                                                             and those ECGs were logged into a separate database
the ACC ⁄ AHA provide specific recommendations on
                                                             that was accessible to the Chest Pain Center Commit-
how to implement adherence to a 10-minute goal, if it
                                                             tee. It was this method that comprised the preinterven-
is even possible.
                                                             tional group of patients.
   The primary purpose of our study was to determine
if we could improve adherence to our door-to-ECG
                                                             After Intervention.    Nursing, ED technicians, and
time goal of 10 minutes or less for patients who pre-
                                                             registration staff devised a plan to improve door-to-
sented to our ED with chest pain by implementing a
                                                             ECG time without the need for an increase in staffing.
new process for obtaining ECGs, without increasing
                                                             No other interventions were scheduled or implemented
staffing. The secondary goal was to describe how the
                                                             during the time period. All nonclinical staff members in
new process affected the door-to-balloon (DTB) time
                                                             the ED waiting room would direct patients to the regis-
for STEMI cath alert patients.
                                                             tration desk. The registration clerks were trained to ask
                                                             patients for their chief complaint. If a patient com-
METHODS                                                      plained of chest pain or affirmed chest pain after being
                                                             prompted for complaints that were not trauma-related
Study Design
                                                             but could be ACS-related (e.g., symptoms such as
This was a planned before-and-after interventional
                                                             weakness, shortness of breath, or epigastric pain), a
study design of ED arrival-to-ECG time. We tested the
                                                             quick registration would be taken that comprised name
null hypothesis that we could not improve the percent-
                                                             and date of birth. Next, the registration staff would
age of patients who received an ECG within 10 min-
                                                             immediately overhead page or directly call a designated
utes. The study was approved by our institutional
                                                             ECG technician to the triage station.
review board.
       Copyright Agency Limited (CAL) licenced copy
ACAD EMERG MED • October 2009, Vol. 16, No. 10   •   www.aemj.org                                                          923

   At the start of every shift, one ED technician would
                                                                    Outcome Measures
be assigned to carry the new ECG cell phone. That
                                                                    The primary outcome was the interval from ED arrival
technician’s primary responsibility was performing
                                                                    time to ECG completion time. With a sample size of 300
ECGs and immediately delivering them to attending
                                                                    in each group, the study was powered to detect a mini-
physicians, although the technician had other ED
                                                                    mum difference of 2 minutes between groups with
duties. Emergency physicians were responsible for
                                                                    power set at 90% and alpha at 0.05. Arrival time was
recording the time of ECG interpretation on a log
                                                                    taken from our computerized tracking system and
using computerized time. A backup ECG technician
                                                                    began once the quick registration time was initiated,
whose primary assignment was in the ED was also
                                                                    which we refer to as door time. ECG completion time
identified who would be handed the cell phone by
                                                                    was taken from computerized ECG logs that were syn-
the ED charge nurse when the primary ECG techni-
                                                                    chronized to our computerized tracking system. The
cian was on a break or otherwise unavailable. He or
                                                                    difference between ECG completion time and comput-
she also served as a second ECG technician during
                                                                    erized arrival time was calculated as door-to-ECG time.
times of high patient volume, when a second triage
                                                                    The secondary outcome, DTB time, was taken as the
and ECG area was used if a second chest pain pre-
                                                                    difference between the time of registration and the time
                                                                    the coronary catheter guidewire passed the culprit
   After the ED technician verified with the ED attend-
                                                                    lesion as reported by the interventional cardiologist.
ing that there was not a STEMI or new LBBB, he
or she would return to the patient to draw blood. A
                                                                    Data Analysis
triage nurse would then assess the patient and assign
                                                                    To determine whether patients differed before and after
the acuity by our hospital’s modified Emergency
                                                                    intervention with regard to age, sex, race, triage classi-
Severity Index (ESI) criteria.10 In our hospital, we
                                                                    fication (ESI 1–2, which we dichotomized to as ‘‘emer-
assign as follows: 1 = immediate resuscitation, 2 =
                                                                    gent’’ vs. 3–4, which we refer to as ‘‘nonemergent’’),
unstable vital signs ⁄ emergent, 3 = urgent, 4 = nonur-
                                                                    mode of arrival, and time of arrival, data were analyzed
gent, and 5 = routine. Chest pain patients with a history
                                                                    using chi-square or Fisher’s exact tests. Door-to-ECG
of coronary artery disease or an abnormal ECG are
                                                                    times are reported as medians ± interquartile ranges
assigned as emergent, and those with no history of cor-
                                                                    (IQRs). For analysis purposes, these times were dichot-
onary artery disease and a normal ECG are assigned as
                                                                    omized into £10 minutes or >10 minutes. Differences in
urgent. Patients can also be assigned as nonurgent if
                                                                    the proportion of patients receiving ECG in £10 min-
deemed to be very low risk for ACS at triage. No
                                                                    utes before and after the intervention were calculated
patients included in this study were triaged as routine.
                                                                    with 95% confidence intervals (CIs). To adjust for possi-
A full registration by registration staff would occur
                                                                    ble confounders chosen a priori (age, sex, race, triage
after the clinical nurse or physician performed his or
                                                                    classification, mode of arrival, and time of arrival), rela-
her initial assessment.
                                                                    tive risk (RR) regression using the Gaussian estimating
   For patients who arrived by ambulance, the process
                                                                    equation was performed.11,12 A post hoc analysis to
remained unchanged due to geography and patient vol-
                                                                    examine door-to-ECG time and DTB time for STEMI
ume. Patients who receive their ECG initiated from reg-
                                                                    alert patients was performed using the Wilcoxon rank
istration have their ECG taken at the front of the
                                                                    sum test. All analyses were performed using SAS statis-
department near registration. Patients who arrive via
                                                                    tical software (Version 9.1, SAS Institute, Cary, NC). A
the ambulance bay enter through a door at the back of
                                                                    probability <0.05 was considered statistically significant.
the ED and are placed in a treatment room, triaged by
the clinical nurse, and registered at the bedside.
Because the primary ECG technician spent much time                  RESULTS
performing ECGs on the walk-in patients, we had the
                                                                    A total of 718 consecutive patients with a complaint of
nurse caring for the ambulance patient, or the ED tech-
                                                                    chest pain at triage were studied over 2 months: 313
nician assigned to that geographical area, do the ECG
                                                                    before and 405 after the intervention. The mean
rather than take the ambulance stretcher to the front of
                                                                    (±standard deviation [SD]) age was 50 (±16) years; 54%
the department when no registration information had
                                                                    were women, 57% were African American, 36% were
been taken.
                                                                    white, and 7% were other races. Patients arrived on
   All data were retrospectively collected by trained
                                                                    their own (walked in) 89% of the time, and 11%
research assistants and taken from our computerized
                                                                    arrived by ambulance. They were triaged as emergent
tracking system in a preplanned analysis. We collected
                                                                    39% of the time and nonemergent 61% of the time.
data for 1 month before and 1 month after the imple-
                                                                    Patients presented during daytime hours 68% of the
mentation of the new ECG process. There was no train-
                                                                    time and during night time hours 32% of the time.
ing phase for this project. While the nursing and ECG
                                                                    Patients before and after the intervention were similar
technicians were aware of the change, the ED physi-
                                                                    with regard to age, sex, race, mode of arrival, and
cians were not. No staff members were notified that
                                                                    time of arrival (Table 1). Patients before the interven-
ECG times were being formally analyzed for this study.
                                                                    tion were more likely to be classified as emergent
Demographic information included age, sex, and race.
                                                                    (44% vs. 34%, difference = 10%, 95% CI = 3% to 17%,
Other information collected included mode of arrival
                                                                    p = 0.007).
(walk-in or ambulance), triage classification (ESI), and
                                                                       Before the intervention, 16% of chest pain patients
time of arrival (day time was defined as 07:00–18:59,
                                                                    received ECGs at 10 minutes or less, and the median
and night was 19:00–06:59).
      Copyright Agency Limited (CAL) licenced copy
924                                                                                                               Takakuwa et al.   •   ECG TIME

Table 1
Variables Before and After the Intervention

 Variable                                         Before, n (%)                            After, n (%)                                 p-value
 Age, yr                                           50.5 ± 16.1                                48.8 ± 16.5                                0.17
 Male sex                                           145 (46)                                   188 (46)                                  1.00
   African American                                   185 (60)                                 224 (56)                                  0.47
   White                                              106 (34)                                 156 (39)
   Other                                               17 (6)                                   21 (5)
 Triage classification
   Emergent                                           139 (44)                                 138 (34)                                  0.007
   Nonemergent                                        174 (56)                                 265 (66)
 Mode of arrival
   Walk-in                                            279 (89)                                 356 (88)                                  0.64
   Ambulance                                           34 (11)                                  49 (12)
 Time of arrival
   Day (07:00–18:59)                                  212 (68)                                 273 (67)                                  0.94
   Night (19:00–06:59)                                101 (32)                                 132 (33)

                                                                           When examining time to ECG for walk-in versus
                                                                        ambulance patients, times significantly improved after
                                                                        the intervention regardless of mode of arrival (Table 2).
                                                                        The largest improvement was for walk-in patients: 67%
                                                                        had an ECG performed within 10 minutes, compared to
                                                                        15% before the intervention (difference = 52%, 95%
                                                                        CI = 46% to 58%).
                                                                           There were a total of 11 STEMI cath alerts during the
                                                                        study: four before (all were walk-in patients) and seven
                                                                        (four walk-ins) after the intervention (Table 3). There
                                                                        were no cath alerts for new LBBB. Prior to the inter-
                                                                        vention, none of the four patients had ECG time within
                                                                        10 minutes. However, two of the STEMIs did not com-
                                                                        plain of chest pain: one patient’s chief complaint was
                                                                        dehydration, and the other was shortness of breath.
                                                                        The average DTB time for these four patients was 154
Figure 1. Time to ECG          before   and   after    intervention.    (SD±72) minutes. Three of the four came during off-
ECG = electrocardiogram.                                                peak hours when the cath team had to be called into
                                                                        the hospital.
                                                                           All seven STEMI cath alerts that occurred after the
door-to-ECG time was 16 minutes (IQR = 12–24). After                    intervention had ECG times within 10 minutes. All had
the intervention, 64% of such patients met the time                     chest pain as their chief complaint, and the median time
requirement of 10 minutes or less, and the median                       to ECG was 7 minutes (compared to 22 minutes in the
door-to-ECG time was 9 minutes (IQR = 8–12 minutes).                    preintervention group, p < 0.05). The median DTB time
The difference between the groups was 47.3% (95%                        for this group was 91 minutes (compared to 159 min-
CI = 40.8% to 53.3%, p < 0.0001; Figure 1). After adjust-               utes, p = NS). The three STEMI alerts that occurred
ing for age, sex, race, mode of arrival, triage classifica-              during weekday hours when the cath team was on site
tion, and time of arrival, patients after the intervention              all had DTB times of <90 minutes, while the four STEMI
were 3.9 times more likely to receive an ECG in 10 min-                 alerts that occurred during off peak hours had DTB
utes or less (RR = 3.9, 95% CI = 3.0 to 5.1).                           times of >90 minutes.

Table 2
Time to ECG Before and After Intervention for Walk-in Versus Ambulance Arrival

                            Before Intervention                               After Intervention
                  %<10 Min        Time to ECG           IQR       %<10 Min         Time to ECG              IQR     Difference           95% CI
 Walk-in             14.7               16             12–24           67.0               9                8–11        52.3%              46–58
 Ambulance           17.7               17             12–26           38.3              12               12–18        20.6%             0.4–39

 ECG = electrocardiogram; IQR = interquartile range.

        Copyright Agency Limited (CAL) licenced copy
ACAD EMERG MED • October 2009, Vol. 16, No. 10    •   www.aemj.org                                                             925

Table 3
Pre- Versus Post-intervention STEMI Cath Alert Patients

                                                                                 Arrival       Time to      DTB
                     Age, yr   Sex      Race            Chief Complaint    ESI   Mode         ECG, min   Time, min   Time of Day
 Pre-intervention      49      Female   White           Chest pain          2    Walk-in         14         196      Weekday
  patients             53      Female   White           Chest pain          2    Walk-in         20          70      Night
                       45      Female   African         Shortness of        2    Walk-in         24         122      Weekday
                                         American        breath
                       41      Male     White           Dehydration         3    Walk-in         56         229      Night
 Post-intervention     72      Male     White           Chest pain          2    Ambulance        3          37      Weekday
  patients             42      Male     African         Chest pain          2    Ambulance        4          91      Night
                       60      Male     White           Chest   pain        1    Walk-in          5          60      Weekday
                       81      Male     White           Chest   pain        1    Ambulance        7         101      Weekend
                       41      Male     White           Chest   pain        2    Walk-in          7         115      Night
                       67      Male     White           Chest   pain        2    Walk-in          9          75      Weekday
                       66      Female   White           Chest   pain        2    Walk-in         10         134      Night

 cath = catheterization; ECG = electrocardiogram; ESI = Emergency Severity Index; STEMI = ST-segment elevation myocardial

DISCUSSION                                                             to meet the 10-minute door-to-ECG goal was to
                                                                       immediately identify eligible patients at the door. The
Our data demonstrate that we were able to improve                      addition of trained personnel at the door was deemed
adherence to the 10-minute goal of door-to-ECG time                    to be too costly.
from 16% to 64%. We were able to accomplish this by                       Because of the relatively small number of STEMI acti-
retraining our registration staff, eliminating the reassess-           vations due to our patient population and demograph-
ment by a triage nurse for chest pain, and prioritizing                ics, we did not have the statistical power to perform a
ECGs to an assigned ED technician. This initiative did                 thorough analysis of our STEMIs. We did capture all
not utilize additional staff.                                          patients who underwent a cath alert with the new pro-
   We were able to improve ECG time without regard                     cess. It appears that if a patient presents with chest
to sex or race, which is important because there is                    pain during daytime hours, a short ECG time helps to
abundant literature showing disparities in cardiac care                improve DTB time, but due to the small number of
for women and minorities.7,13–20 We were also able to                  patients, we were unable to show statistical signifi-
do it at a minimal cost (one new cell phone).                          cance. In contrast, during weekends and nights when
   Previously, two small studies that examined the initi-              the cath team had to come to the hospital, DTB time
ation of a multistrategy approach to reducing DTB time                 was relatively long. To meet a DTB time goal of
showed reduced door-to-ECG times. In the first, ECG                     <90 minutes, not only do we need to work on improv-
time was reduced by performing routine ECGs in men                     ing ECG time, we need to improve the response time of
>35 years or women >40 years old with nontraumatic                     the cath team. A study by Bradley et al.24 showed that
chest pain.21 This strategy required an ED tech                        expecting cardiac cath team members to arrive within
assigned to triage using a new dedicated ECG machine,                  20 minutes of being paged saved a mean of 19.3 min-
but the study does not detail how patients were identi-                utes off of DTB time.
fied and whether a new tech was needed. The second                         Our initiative increased the total number of patients
study of this type showed similar reduced door-to-ECG                  who received ECGs in triage between the pre- and
time, but provides no details on how staff identified                   post-intervention period by almost 30%. During this
patients with a ‘‘complaint consistent with angina.’’22                same time period, our census increased by 5%. This
Another study showed that adding an ED triage                          increase in ECGs was undoubtedly due to the prompt-
greeter improved ECG time.23 In this small study, an                   ing of chest pain in patients at registration. While it can
extra triage staff member had the sole job of assessing                be argued that this increase was for people who did
the need for an ECG, then performing it if necessary;                  not need emergent ECGs, we believe these people
this reduced overall ECG time. We believe that our ED                  would have received ECGs sometime during their ED
environment reflects the actual clinical environment of                 visit. From the pre- to post-intervention period, the
most busy EDs. We have limited resources and needed                    total number of ECGs performed in the ED decreased
to address an important clinical issue without adding                  by 21%, suggesting that our initiative consolidated the
personnel.                                                             performance of ECGs to earlier in the patient ED visit.
   The method we chose of identifying patients for                     Given that our patient population is heterogeneous with
rapid ECGs by registration clerks was controversial. It                respect to race and sex, and many of our younger
was made with the knowledge that we would be                           patients with atypical symptoms may have coronary
relying on untrained medical personnel. We suspected                   artery disease, we believe that it is better to obtain
that this would probably increase the number of ECGs                   more ECGs earlier in the workup than to risk delaying
recorded early in a patient visit. However, one way
      Copyright Agency Limited (CAL) licenced copy
926                                                                                              Takakuwa et al.   •   ECG TIME

a possible STEMI with a more atypical ACS presenta-            10-minute-or-less door-to-electrocardiogram time stan-
tion.                                                          dard for chest pain patients presenting to a busy urban
   One arena that our initiative did not target was            academic ED. This appeared to help achieve quicker
patients who arrive by ambulance, who comprised 11%            door-to-balloon time during hours when the catheteri-
of our patients. We decided to keep ambulance patients         zation team was on site, but not during nights and
in the analysis even though they were not targeted for         weekends.
ECG time improvements. A post hoc analysis showed
that eliminating ambulance arrivals had little effect on       References
the total results. We suspect that the significant
improvement for ambulance patients reflected a general           1. Anderson JL, Adams CD, Antman EM, et al.
effort from our ECG technicians to improve ECG times.              ACC ⁄ AHA 2007 guidelines for the management of
It is possible to replicate the same procedure as for the          patients with unstable angina ⁄ non-ST-elevation
walk-in patients by having the ED clinical nurse page              myocardial infarction: a report of the American Col-
the ECG technician to the ambulance receiving area,                lege of Cardiology ⁄ American Heart Association
but this was not done. An alternative strategy would be            Task Force on Practice Guidelines (Writing Com-
to initiate prehospital wireless 12-lead ECG transmis-             mittee to Revise the 2002 Guidelines for the Man-
sion to the ED25 or a cardiologist’s hand-held device26            agement of Patients With Unstable Angina ⁄ Non ST-
or to empower the paramedics to activate a cath alert              Elevation Myocardial Infarction). Circulation. 2007;
based on the ECG computer algorithm27,28 for all                   116:e148–304.
ambulance patients. Several studies have shown that             2. Antman EM, Anbe DT, Armstrong PW, et al.
this strategy reduces door-to-reperfusion and DTB time             ACC ⁄ AHA guidelines for the management of
for STEMI patients.29,30 However, this was not prac-               patients with ST-elevation myocardial infarction: a
ticed in our city at the time of the study.                        report of the American College of Cardiol-
                                                                   ogy ⁄ American Heart Association Task Force on
                                                                   Practice Guidelines (Committee to Revise the 1999
LIMITATIONS                                                        Guidelines for the Management of Patients with
Our study is limited by its single location in an urban            Acute Myocardial Infarction). Circulation. 2004;
academic environment and the institutional idiosyncra-             110:e82–292.
sies. The study was performed for a relatively short            3. Office of Quality and Performance (10Q). FY 2008,
duration and the improvements seen may gradually les-              Q1 technical manual for the VHA performance mea-
sen over time. We only included those patients who                 surement system. Washington, DC: Oct 31, 2007, p
mentioned chest pain to the registration clerk. It is pos-         315. Available at: http://www.qualitymeasures.
sible that patients who mentioned ACS equivalents                  ahrq.gov/summary/pdf.aspx?doc_id=9142&stat=1&-
such as shortness of breath, weakness, and neck ⁄ shoul-           string=. Accessed Jun 19, 2009.
der ⁄ back pain and who did not have chest pain but had         4. Krumholz HM, Anderson JL, Bachelder BL, et al.
ACS were missed for rapid ECGs. In our pre-interven-               ACC ⁄ AHA 2008 performance measures for adults
tion group, the two patients who had chief complaints              with ST-elevation and non-ST-elevation myocardial
of shortness of breath and dehydration respectively                infarction: a report of the American College of Car-
would not have been captured for rapid ECGs with the               diology ⁄ American Heart Association Task Force on
new intervention, because neither of them had chest                Performance Measures (Writing Committee to
pain. As the next step in our ECG process, we should               develop performance measures for ST-elevation
include all patients with ACS symptoms for rapid                   and non-ST-elevation myocardial infarction). Circu-
ECGs. While this would have enabled us to obtain an                lation. 2008; 118:2596–648.
ECG on the patient with shortness of breath, we still           5. Diercks DB, Peacock WF, Hiestand BC, et al. Fre-
would have missed the patient who complained of                    quency and consequences of recording an electro-
dehydration. We recognize the importance of identify-              cardiogram >10 minutes after arrival in an
ing ACS equivalents, but decided to change the process             emergency room in non-ST-segment elevation acute
for obtaining ECGs in a stepwise fashion. It is also pos-          coronary syndromes (from the CRUSADE Initia-
sible that a patient may not have mentioned chest pain             tive). Am J Cardiol. 2006; 97:437–42.
to the registration clerk, but then went on to tell the tri-    6. Diercks DB, Kirk JD, Lindsell CJ, et al. Door-to-ECG
age nurse, which would have delayed obtaining a rapid              time in patients with chest pain presenting to the
ECG and altered our results. While our triage system               ED. Am J Emerg Med. 2006; 24:1–7.
may overtriage possible ACS patients into higher cate-          7. Takakuwa KM, Shofer FS, Hollander JE. The influ-
gories, the benefits of identifying all possible STEMI              ence of race and gender on time to initial electro-
patients without adding skilled personnel may outweigh             cardiogram for patients with chest pain. Acad
the risks of overtriage. Finally, our results represent            Emerg Med. 2006; 13:867–72.
time from arrival to ECG completion and do not                  8. Panju AA, Hemmelgarn BR, Guyatt GH, Simel DL. The
account for ED attending interpretation time.                      rational clinical examination. Is this patient having a
                                                                   myocardial infarction? JAMA. 1998; 280:1256–63.
                                                                9. Anonymous. Emergency department: rapid identifi-
CONCLUSIONS                                                        cation and treatment of patients with acute
These data show that a cooperative effort in the ED can            myocardial infarction. National Heart Attack Alert
result in a significantly higher rate of adherence to the           Program Coordinating Committee, 60 minutes to
        Copyright Agency Limited (CAL) licenced copy
ACAD EMERG MED • October 2009, Vol. 16, No. 10     •   www.aemj.org                                                       927

      Treatment Working Group. Ann Emerg Med. 1994;                   22. Caputo RP, Kosinski R, Walford G, et al. Effect of
      23:311–29.                                                          continuous quality improvement analysis on the
10.   Gilboy N, Tanabe P, Travers DA, Rosenau AM, Eitel                   delivery of primary percutaneous revascularization
      DR. Emergency Severity Index, Version 4: Imple-                     for acute myocardial infarction: a community hospi-
      mentation Handbook. AHRQ Publication No.                            tal experience. Catheter Cardiovasc Interv. 2005;
      05-0046-2, May 2005. Agency for Healthcare                          64:428–33.
      Research and Quality, Rockville, MD. Available at:              23. Purim-Shem-Tov YA, Rumoro DP, Veloso J, Zettin-
      http://www.ahrq.gov/research/esi/. Accessed Jun                     ger K. Emergency department greeters reduce
      19, 2009.                                                           door-to-ECG time. Crit Pathways Cardiol. 2007;
11.   Lumley T, Kronmal R, Ma S. Relative risk regres-                    6:165–8.
      sion in medical research: models, contrasts, estima-            24. Bradley EH, Herrin J, Wang Y, et al. Strategies for
      tors, and algorithms. UW Biostatistics Working                      reducing the door-to-balloon time in acute myocar-
      Paper Series 293. July 19, 2006.                                    dial infarction. N Engl J Med. 2006; 355:2308–20.
12.   McNutt LA, Wu C, Xue X, Hafner JP. Estimating                   25. Garvey JL, MacLeod BA, Sopko G, Hand MM,
      the relative risk in cohort studies and clinical trials             National Heart Attack Alert Program (NHAAP)
      of common outcomes. Am J Epidemiol. 2003;                           Coordinating Committee. National Heart, Lung, and
      157:940–3.                                                          Blood Institute (NHLBI). National Institutes of
13.   Stone PH, Thompson B, Anderson HV, et al.                           Health. Pre-hospital 12-lead electrocardiography
      Influence of race, sex, and age on management of                     programs: a call for implementation by emergency
      unstable angina and non-Q-wave myocardial                           medical services systems providing advanced life
      infarction: The TIMI III registry. JAMA. 1996; 275:                 support–National Heart Attack Alert Program
      1104–12.                                                            (NHAAP) Coordinating Committee; National Heart,
14.   Jackson RE, Anderson W, Peacock WF, et al. Effect                   Lung, and Blood Institute (NHLBI); National Insti-
      of a patient’s sex on the timing of thrombolytic                    tutes of Health. J Am Coll Cardiol. 2006; 47:485–91.
      therapy. Ann Emerg Med. 1996; 27:8–15.                          26. Adams GL, Campbell PT, Adams JM, et al. Effec-
15.   Lambrew CT, Bowlby LJ, Rogers WJ, Chandra NC,                       tiveness of prehospital wireless transmission of
      Weaver WD. Factors influencing the time to throm-                    electrocardiograms to a cardiologist via hand-held
      bolysis in acute myocardial infarction. Time to                     device for patients with acute myocardial infarction
      Thrombolysis Substudy of the National Registry of                   (from the Timely Intervention in Myocardial Emer-
      Myocardial Infarction-1. Arch Intern Med. 1997;                     gency, NorthEast Experience [TIME-NE]). Am J
      157:2577–82.                                                        Cardiol. 2006; 98:1160–4.
16.   Weitzman S, Cooper L, Chambless L, et al. Gender,               27. Bachour FA, Smith SW, Hildebradt D, Simegn M,
      racial, and geographic differences in the perfor-                   Asinger R. Paramedic prehospital cath lab activa-
      mance of cardiac diagnostic and therapeutic proce-                  tion for STEMI, without ECG transmission, dramat-
      dures for hospitalized acute myocardial infarction                  ically reduces door to balloon time [Abstract 1991].
      in four states. Am J Cardiol. 1997; 79:722–6.                       Circulation. 2007; 116(Suppl II):II_430.
17.   Rathore SS, Berger AK, Weinfurt KP, et al. Race,                28. Bachour FA, Smith SW, Hildebrandt D, Simegn M,
      sex, poverty, and the medical treatment of acute                    Asinger R. Effect of prehospital cath lab activation
      myocardial infarction in the elderly. Circulation.                  on door to balloon time of STEMI patients present-
      2000; 102:642–8.                                                    ing during normal workday hours vs. after hours
18.   Barakat K, Wilkinson P, Suliman A, Ranjadayalan K,                  [Abstract 2403]. Circulation. 2007; 116(Suppl
      Timmis A. Acute myocardial infarction in women:                     II):II_527–II_528.
      contribution of treatment variables to adverse out-             29. Curtis JP, Portnay EL, Wang Y, et al. National Reg-
      come. Am Heart J. 2000; 140:740–6.                                  istry of Myocardial Infarction-4. The pre-hospital
19.   Venkat A, Hoekstra J, Lindsell C, et al. The impact                 electrocardiogram and time to reperfusion in
      of race on the acute management of chest pain.                      patients with acute myocardial infarction, 2000-
      Acad Emerg Med. 2003; 10:1199–208.                                  2002: findings from the National Registry of Myo-
20.   Vaccarino V, Rathore SS, Wenger N, et al. Sex and                   cardial Infarction-4. J Am Coll Cardiol. 2006;
      racial differences in the management of acute myo-                  47:1544–52.
      cardial infarction, 1994 through 2002. N Engl J                 30. Brown JP, Mahmud E, Dunford JV, Ben-Yehuda O.
      Med. 2005; 353:671–82.                                              Effect of prehospital 12-lead electrocardiogram on
21.   Zarich SW, Sachdeva R, Fishman R, et al. Effective-                 activation of the cardiac catheterization laboratory
      ness of a multidisciplinary quality improvement ini-                and door-to-balloon time in ST-segment elevation
      tiative in reducing door-to-balloon times in primary                acute myocardial infarction. Am J Cardiol. 2008;
      angioplasty. J Interven Cardiol. 2004; 17:191–5.                    101:158–61.

        Copyright Agency Limited (CAL) licenced copy

To top