Transcript Emergency Department Triage and Evaluation of the Patient with

Emergency Department Triage
and Evaluation of the Patient
with Chest Pain
Department of Emergency Medicine
University of Pennsylvania Health System
Judd E. Hollander, MD
Professor, Clinical Research Director
Department of Emergency Medicine
University of Pennsylvania
ED Visits
130,000,000 visits annually
3,000,000
likely noncardiac
sent home
40,000 MIs
8,000,000 chest pain
5,000,000
suspected or
actual cardiac
Goals of Triage



Identify patients with AMI
Identify patients with unstable angina
Identify patients at high risk of
cardiovascular complications
– resource utilization in hospital
• CCU vs. monitored vs. floor beds

Identify patients safe for ED release
– need for treatment
Your Risk Tolerance…



5%
2%
1%
Why Do More?

The missed AMI rate is inversely proportional to
the admission rate for ED chest pain patients
Kontos MC & Jesse RL. Am J Cardiol 2000;85:32B-39B
Outline







Gut Impression
Clinical Parameters
Electrocardiography
Cardiac Markers
Disposition with or without Telemetry
Prior Testing
Acute Cardiac Imaging
Initial Impression = “Noncardiac Pain”
Patients with initial emergency
impression of “noncardiac chest pain”


2,992

85
itrACS
17,737 patients enrolled
Conclusion: Even patients
thought to have noncardiac pain
can suffer adverse cardiac
events, especially if risk factors
are present
2.8% had adverse cardiac events (infarction,
revascularization, or death) within 30 days
Miller CD, et al. Ann Emerg Med. 2004;44:565.
Clinical Parameters
Department of Emergency Medicine
University of Pennsylvania Health System
Clinical Parameters

Identifying low risk patients
Lee et al. 1985 Arch IM
1985;145:65.

596 ED patients
MI
USA
Other
Clinical Parameters
MI USA Other
Lee et al. Arch IM 1985;145:65.
Clinical Parameters
MI USA Other
Lee et al. Arch IM 1985;145:65.
Clinical Parameters: Risk Factors

Risk factors do not affect likelihood of AMI
– 1700 patients
• Cholesterol
• Hypertension
• Family history

Slight increase in risk in men only
– Diabetes mellitus
• 2.4 (1.2 - 4.8)
– Family history
• 2.1 (1.4 - 3.3)
Jayes et al. J Clin Epidemiol 1992;45:621.
Clinical Parameters: Risk Factors
CRF Burden and ACS (AUC=0.591)
Han et al. Ann Emerg Med 2007;49:145.
Costochondritis

122 patients evaluated for ARA definition
of costochondritis
– pain caused by pressure enough to blanch
skin
– whether or not it precisely reproduced CC

6% of patients had AMI
Disla et al. Arch Intern Med. 1994;154:2466.
“Clear Cut Alt Diagnosis”


Of 1995 pts, 599 pts had an Alt Dx
Presence of an Alternative Diagnosis
– Reduced the likelihood of 30 day death, MI,
revascularization
• 8.8 to 4.0%
• Risk ratio, 0.45 (95% confidence interval, 0.29-0.69)
4% risk of 30 death, MI, revascularization is not
low enough to allow safe release from the ED
Hollander et al. Acad Emerg Med., 2007:14:215
Clinical Parameters

History and physical examination are not real
helpful in identifying patients with AMI.
Electrocardiograms
Lee et al. 1985 Arch IM 1985;145:65.
Electrocardiograms


Patients admitted to CCU
Morbidity and mortality related to ECG
Slater et al. Am J Cardiol 1987;60:766.
Electrocardiograms

%
Patients admitted to CCU (n=469)
25
20
46
Neg ECG n=167
15
Pos ECG n=302
10
18
18
5
4
0
1
1
0
VF
Sus VT
NS VT
Cond Dist
Brush et al. NEJM 1985;312:1137.
Late Electrocardiograms
Does the NPV of the ECG increase with time?
 Normal ECG over time
Symptom duration
NPV
0-3 hrs
93%
3-6 hrs
93%
6-9 hrs
93%
9-12 hrs
94%

Singer et al. Annals EM 1997;29:575.
Combination of Clinical
Parameters and
Electrocardiography
Department of Emergency Medicine
University of Pennsylvania Health System
Chest Pain Study Group
Chest Pain Study Group Risk






Heavily dependent on ECG
No group of patients at less than 1% risk of AMI
Cardiac risk factors not useful
Defined high and low risk as 7% cut-off
May be useful for triage
No patients deemed safe for release from ED
Lee et al. NEJM 1991;324:1239.
Young Patients-Validated





Of 4492 visits for CP, 1023 visits were pts<40 yrs
If no cardiac risk factors and no prior cardiac history (n=436)
– 6 USA (1.4%) initial diagnosis
– 2 AMI (0.5%) during index visit
– 30 days – no death, AMI, PCI or CABG (0.5%, 95% CI, 0-1.1%).
Normal ECG and no prior cardiac history (n=593)
– 6 USA (1%) initial diagnosis
– 1 AMI (0.17%) during index visit
– no AMI, PCI or CV deaths during follow up (0%, 95% CI, 0-0.5%).
– Risk of 30 day adverse events 0.3 (0-0.8%)
No prior history, no risks, normal ECG (n=299)
– 3 USA (1%), no AMI
– No 30 day adverse events (0%; 0-1%)
Add initial marker
– Only 1 ACS, nothing else for any of the groups (0.14%; 0.1-0.2%)
Marsan et al. AEM 2005;128:26.
Clinical Parameters: Risk Factors
.763
.602
.518
CRF Burden and ACS
Han et al. Ann Emerg Med 2007;49:145.
TIMI Risk Score

TIMI Risk Score for UA
–
–
–
–
–
–
–
Age > 65
3 or more CRF’s
Known CAD > 50%
ST segment changes on ECG
2 or more anginal events in past 24 hours
ASA use within 7 days prior
Elevated cardiac markers
TIMI Risk in the ED
# of TIMI Risk Factors
Chase, et al. Ann Emerg Med. 2006:48:252
High Sensitivity
Cardiac
Markers
Department of Emergency Medicine
University of Pennsylvania Health System
TnI-Ultra: 60d AMI/CV Death
371 patients with symptoms suggestive of ACS
cTnI
N
# Events
Rate, %
RR
<0.006
108 (29%)
2
1.9
-
0.006-0.04
>0.04-0.10
174 (47%)
38 (10%)
11
9
6.4
24.1
3.3
13.0
>0.10
51 (14%)
28
55.1
34.9
Apple et al. Clin Chem 2008;54:723
High Sensitivity Troponin
 718 patients with potential AMI; 123 had AMI
Presentation
 Sens = 84-95%
 Spec = 80-84%
Reichlin et al. NEJM 2009;361:858
High Sensitivity Troponin
 1818 patients with potential AMI; 413 had AMI
Presentation
 Sens = 90%
 Spec = 90%
Within 3 hours
 Sens = 100%
Keller et al. NEJM 2009;361:868
hsTnI in UA: Protect TIMI 30
Wilson et al. Am Heart J 2009;158:386
2009
100 potential ACS patients
35% discharged
65% admitted
85% bogus
15% real
The Future
100 potential ACS patients
35% discharged
90% Sens
80% Spec
44
8
35 discharged
71 discharged
65% admitted
85% bogus
15% real
55 not sick (IM) 10 real (cards)
19 (trop FP)
9-10 real
Stuck with
Admissions?
Department of Emergency Medicine
University of Pennsylvania Health System
Evidence Based
Work Arounds
Observation Unit Rationale



Cannot identify a group of clinical and/or
ECG variables that identifies patients at such
low risk for AMI/complications that they can
be safely released from the ED
No single test sufficiently excludes risk of
AMI or complications
Attempts to shorten evaluation
Telemetry

Hollander et al – Prospective study
AJC 1997;110
– 460 CP pts with normal or nonspecific ECG’s
• 4 CV complications (1 VT/VF post op; 1 SVT in CHF pt; 2
sinus pauses of 2.4 and 4 seconds without intervention)

Schull et al – Retrospective study AEM 2000;7:647
– 8932 pts admitted to tele ward
• 20 cardiac arrest
– 9 detected by monitor
• 3 survival to discharge
– 1 definitely detected by monitor; 1 detected by neighbor when
he fell to floor; 1 no record of when it began on monitor
(?detected)
Telemetry: HUP Data
Total Patients
(n=3686)
ICU/cath lab
Telemetry
Floor
Home
424 (12%)
1748 (47%)
110 (3%)
1383 (38%)
Goldman < 7
Goldman > 7
1157 (66%)
591 (34%)
Markers positive
Markers negative
130 (11%)
1027 (89%)
Hollander et al. Annals EM 2004;43:71.
Telemetry: HUP Data


Sustained VT/VF
Bradydysrrhythmias requiring treatment
0% (95 CI, 0-0.3%)

Preventable CV Death
0% (95 CI, 0-0.3%)
Hollander et al. Annals EM 2004;43:71.
Telemetry: HUP Outcomes
Initial Hospitalization
No. Percent
Myocardial infarction
15
Unstable angina
121
Percutaneous intervention 11
Stent Placement
10
CABG
4
Death
2
1.5%
12%
1.1%
1.0%
0.4%
0.2%
Hollander et al. Annals EM 2004;43:71.
It’s Not My Heart
I Had a Test Already
Department of Emergency Medicine
University of Pennsylvania Health System
Stress Tests and ED Disposition
92%
100
90
72%
Percent
80
67%
70
60
50
40
30
20
10
0
Abnormal
Normal
None
Disposition (% admitted)
Nerenberg et al. AmJEM 2007;25:39.
Stress Tests & 30-Day Outcomes
12
10.1%
Percent
10
8
4.8%
6
5.2%
4
2
0
Abnormal
Normal
None
30-Day Adverse Cardiovascular Outcomes (%)
Nerenberg et al. AmJEM 2007;25:39.
Maybe It Keeps Them Away?

Shaver et al demonstrated that patients
evaluated with stress testing were just as
likely to:
– Return to the ED (39 vs 40%)
– Be admitted to the hospital (29 vs 32%)
– Receive cardiac catheterization (12.5 vs 10.4%)
Shaver et al. Acad EM 2005;11:1272
Better Than Stress Testing

deFillipi et al found that compared with
patients who were evaluated with stress
testing, patients evaluated with coronary
angiography (CA) had:
–
–
–
–
Fewer repeat ED visits
Fewer hospitalizations
Higher satisfaction rates
Better understanding of their disease
deFillipi et al. JACC 2001
Acute Cardiac
Imaging (in the ED)
Department of Emergency Medicine
University of Pennsylvania Health System
Echocardiography

Detects wall motion abnormality
– sensitivity moderate high

Cannot distinguish old from new
– many “false positives”

May miss non-Q wave AMI
– usually small infarcts

Never compared to physician judgment or
cardiac markers to assess incremental value
Sestamibi Imaging

338 ED chest pain patients with normal scans
– None had a cardiac death during 1 year period
– None had an MI
– 7 required coronary revascularization

100 abnormal scans
– 7 AMI
– 30 revascularization within one year
Tatum et al. Annals EM 1997;29:116.
Sestamibi Imaging

Relative risks of abnormal scans
– AMI
50 (2.8-890)
– Revascularization 14.5 (6-34)
– Death by 1 year 30 (1.6-570)

Sensitivity for AMI
– 100% (64-100)

Specificity
– 78% (74-82)
Tatum et al. Annals EM 1997;29:116.
ER Assessment of Sestamibi (ERASE)


RCT of 2475 ED chest pain patients with normal
or nondiagnostic ECGs
– Usual ED evaluation (n=1260)
– Usual evaluation & resting MPI (n= 1215)
Primary outcome
– Appropriateness of initial triage decision
Udelson JE et al. JAMA. 2002;288:2693
ERASE



Sensitivity for MI and acute
ischemia were not
significantly different
Patients in the acute MPI arm
had a significantly lower
hospitalization rate
Costs reduced in the MPI arm
by an average of $70/patient
100
NS
97 96
NS
83 81
MPI
Standard
P<.01
75
52
50
42
25
0
MI
ACS
Admit
Udelson JE et al. JAMA. 2002;288:2693
Coronary CTA Accuracy
Correlation with cardiac catheterization
Study
Scanner
Sensitivity
Specificity
Janne d’Othee
All
95%
85%
Janne d’Othee
64 slice
98%
91%
Heuschmid
Dual source
96%
87%
Weustnik
Dual source
99%
87%
Scheffel
Dual source
96%
98%
Coronary CTA Prognosis
 Meta-analysis
 9592 patients
 Median f/u
 20 months
 MACE
 Sensitivity = 99%
 LR - = 0.008
Hulton et al. JACC 2011:57:1237
No / noncritical disease
 Hollander et al.
– 100% NPV for D/AMI/revasc in 525 patients at 30
days
 Hoffman et al.
– 100% NPV for ACS in 73 pts over 5 months
 Rubinshtein et al.
– 100% NPV for 35 pts over 15 months
 Pundziate et al.
– 100% NPV for 20 pts over 13 months
No / noncritical disease
 Goldstein et al.
– 100% NPV for D/AMI/revasc in 67 patients at 30
days
 Hoffman et al. (ROMICAT)
– Any plaque
• 100% NPV for ACS/events in 183 pts over 6 months
– Stenosis < 50%
• 98% NPV for ACS/events in 300 pts over 6 months
CT Coronary Angiography

Largest cohort study
– 525 of 568 patients with negative CTA
– 30 day follow-up



No cardiac deaths (95% CI, 0-0.8%)
No AMI (95% CI, 0-0.8%)
No revascularization (95% CI, 0-0.8%)
Hollander et al Ann EM 2009;53:295.
All CTA (n=568)
Ca>400, no contrast
injection (n=6)
CTA with contrast
injection (n=562)
No stenosis or maximal
stenosis < 50%
(n=508)
Maximal stenosis
50-69%
(n=41)
Maximal stenosis
>70%
(n=13)
5
None
Stress
Cath
None
Stress
473
32
3
18
21
3
-
+
-
+
32
0
16
5
-
+
-
3
0
0
Cath
None
Stress
5
3
5
2
50-69%
3*
Cath
1
6
+
3
+
-
50-69%
+
1
0
4
2
CT Coronary Angiography

RCT of CCTA v MPI post CDU (n=197)
– Normal CCTA discharged home (75%)
– 9 with severe disease to catheterization
– Intermediate disease to stress test



CCTA reduced LOS (3.4 v 15.0 hours)
CCTA reduced costs ($1586 v $1872)
Re-evaluation of chest pain (2% v 7%)
Goldstein et al JACC 2007;49:863-871
CT STAT
CCTA (n=361)
MPI (n=338)
Time to Diagnosis
7.9 hrs
6.2 hrs
ED Cost
$2137
$3458
Death within 6 months
0
0
AMI within 6 months
1 (0.3%)
5 (1.5%)
ED revisit, cardiac
2 (0.6%)
4 (1.3%)
0
0
82%
90%
2/268 (0.8%)
1/266 (0.4%)
Rehospitalization, cards
Normal test
MACE with nl test
Goldstein et al . JACC 2011;58:1414-22
Main Outcomes: Efficacy
CTA
(N=98)
Obs/CTA
(N=102)
Obs/Stress
(N=154)
Tele
Testing
(N=289)
$4154
Total
Facility
Cost $
(IQR)
1240
(723-1943)
2318
(2000-3041)
4024
(3322-4751)
2913
(1713-5592)
LOS hours
(IQR)
8.1
(5.9-13.7)
20.9
(15.1-26.5)
26.2
(21.3-32.1)
30.2
(24.0-73.1)
Chang et al. AEM 2008;15:649.
Main Outcomes: Safety
% CAD
% Death/MI
% Rehosp
CTA
(N=98)
Obs/CTA
(N=102)
Obs/Stress
(N=154)
Tele
(N=289)
5.1
(1.7-11.5)
5.9
(2.2-12.4)
5.8
(2.7-10.8)
6.6
(4.0-10.1)
0
(0-3.7)
0
(0-3.6)
0.7
(0.1-3.6)
3.1
(1.4-5.8)
0
(0.0-4.0)
3.2
(0.7-9.0)
2.3
(.06-12.0)
12.2
(8.5-16.7)
Chang et al. AEM 2008;15:649.
Fagan’s Nomogram for MACE
Hulton et al. JACC 2011:57:1237
CT Coronary Angiography
Volume
rendered (VR)
LAO view:
Normal LAD
and diagonal
branches
VR images
provide an
overview of the
coronary
arteries but can
not be used on
their own to
exclude
stenosis.
CT Coronary Angiography
Thin-slab MIP
(maximum
intensity
projection):
No stenosis in
proximal LAD,
circumflex and
ramus medianus
(RM) arteries.
CT Coronary Angiography
L Main: Calcified
plaque with 50%
stenosis of the left
main
LAD: Mixed
calcified and
noncalcified plaque
resulting in 70%
stenosis
Diagonal: mild
stenosis
LCx: Patent
CT Coronary Angiography
Low density
noncalcified
plaque (arrow)
causing >50%
stenosis of the
proximal right
coronary
artery.
CT Coronary Angiography
Severe RCA lesion
ACRIN
 Randomized 2:1 to Coronary CTA
 Coronary CTA group




Coronary CTA
Clinical bloods (c/w guidelines) at time 0 & 90-180 minutes
Banked bloods at T0, 90-180 and 6 hours
Dispo per physicians
 Traditional care group
 Anything but coronary CTA
 Banked bloods at T0, 90-180 and 6 hours
 Dispo per physicians

Results March 26 at ACC
 ROMICAT 2 – March 27
Putting It Together
Department of Emergency Medicine
University of Pennsylvania Health System
Triage



Clinical Presentation
ECG
Past history
– CAD


Available technology
Required medications
– Fibrinolytics
– IV nitrates
– Heparin
Triage

Risk stratification
–
–
–
–
–

TIMI Risk Score
HEART Score
GRACE or PURSUIT
Lee and Goldman algorithm
Clinical impression
Adjunctive Testing
– Markers
– Imaging
Triage

High risk patients
–
–
–
–
–
–
–

ECG abnormalities
Heart failure
Dysrrhythmias
Unstable vital signs
Need for IV drips
Positive markers or MPI scans in the ED
Positive CCTA with good story
Admit to Cardiac Care Unit
Triage

Lowest risk patients
–
–
–
–
–

Young patients
Normal electrocardiograms
Low risk story
TIMI Score <3
Normal markers and sestamibi scans, if done
Triage to
– Observation unit
– Nonmonitored beds
– Home if lowest possible risk
• Normal CCTA goes home
• individual and institutional cut-off for misses
Triage

Intermediate Risk Patients
–
–
–
–
–

Equivocal stories
Abnormal but not diagnostic ECG’s
TIMI Score > 3
Markers normal or slightly elevated
Scans with old abnormalities (CAD)
Most should be admitted to monitored beds
Unstable Angina


Distinguish real unstable angina from
need to “rule out” AMI
Single atypical episode of chest pain
– “rule out MI”
– unstable angina?
Summary




ACS versus anything else for dispo/triage
CTA to allow discharge
AMI is ANY elevation in markers above
normal
STEMI or NSTEMI drive treatment
Words of wisdom? (without evidence)

Short cuts to “r/o MI”
– 90 minute to 3 hour “rule outs”
• Rising or delta cardiac markers

Incidental abnormal ECG’s
– Always make referral
– QTc intervals

Admission diagnosis also should include
– “rule out life threatening conditions”

Stable angina
– Whatever it is – it is stable for outpt evaluation

The ROS curse