Transcript EMS Controversies in Prehospital Care

EMS
Controversies in Prehospital Care
Aric Storck PGY1
Dr. Andy Anton
June 26, 2003
Objectives
 Public Access
–
Better or worse than targeted first responder
programs?
 Defibrillation
–

Defibrillation
+/- CPR
Shock first or CPR first?
Prehospital Triage
–
Is it accurate and is it safe?
Background

Cardiovascular disease is a major cause
of death in North America
–
–
–
1 million deaths per year in US
~450,000 sudden and unexpected
Most occur outside of hospital
 Poor
survival rates (1-5% in some major centres)
The target of PAD
Incidence of Unexpected Cardiac Arrest (NEJM 2002:347(16);1223-1224)
Very, very good evidence that
survival is intimately linked to
CPR
 Timeliness of defibrillation

NEJM 2002:347(16);1223-1224
Public Access Defibrillation

What is it?
–

PAD is a strategy to improve survival from VF arrest
by placing AED’s in the hands of trained (and
untrained) laypersons
What is the rationale?
–
Survival decreases by 7-10% for every passing
minute of VF. PAD allows for earlier defibrillation.
What is an Automatic External
Defibrillator?
JAMA 2001 285(9)1193
Sequence of AED voice prompts during mock cardiac arrest.
Circulation 1999;100(16):1703-1707
Electrocardiograms From a Person in Cardiac Arrest
JAMA 2001;285(9):1196
What about EMS and AED’s?

Many studies (which will not be reviewed here)
show AED’s to be an effective part of EMS
– Some studies suggest AEDs more effective
than manual defibrillators among
paramedics
– Many studies have shown increased survival
with targeted first responders (police,
firefighters, etc.) carrying AEDs
AEDs Outside the Traditional EMS
System


Some (including AED manufacturers …) have
argued for a “fire extinguisher” model
AEDs widely located in public and private
places to achieve defibrillation ASAP
Public Access Defibrillation




VF is the most frequent initial rhythm in
witnessed cardiac arrest
Electrical defibrillation is the most effective
treatment of VF
Probability of successful defibrillation
diminishes rapidly over time
VF tends to converts to asystole after a few
minutes
Time to Defibrillation vs Rates of
Survival to Hospital Discharge
Rates given for 4 years of consecutive patients (1112) initially discovered
in ventricular fibrillation. Response times are known for 942 cases. Both
witnessed and unwitnessed cases in this 4-year period are considered.
The average (SD) response time for first emergency unit was 3.0 (1.5)
minutes and for paramedics was 6.5 (3.2) minutes after dispatch. VF
indicates ventricular fibrillation.
JAMA 2001;285(9):1196
Expected Survival According to the Interval
between Collapse and the Administration of the
First Shock by the Defibrillator.
NEJM 2002;347(16):1223-1224
So What Does the
Literature Say…
Outcomes of Rapid Defibrillation by
Security Officers After Cardiac
Arrest in Casinos
Valenzuela T, et al. NEJM. 2000;343(17):1206-1209.

Casinos are unique environments amenable to
a rapid defibrillation program
–
–
–
Security officer visible from every point at all times
Security cameras closely monitor public areas
Many elderly patrons in poor health
Methods




Prospective study
Prospective target of 100 VF arrests
32 participating casinos in Nevada,
Pennsylvania, and Mississippi
AEDs placed at intervals to ensure <3 minute
elapsed time between collapse and
defibrillation
Training


Security officers required to have basic CPR
6 hour course in AED use
–
–
Theoretical and practical training
75% required on exam to pass
Subjects

Inclusion
–
–
–
–

Unconscious
Unresponsive
No palpable pulse
No spontaneous respiration
Exclusion
–
–
<9 years old
<36 kg
Protocol

When informed of “sick person” the nearest
officer procedes to the patient
–



Begins CPR if needed
Second officer brings nearest AED to patient
AED immediately attached and prompts
followed
Resuscitation continued until EMS arrival or
return of spontaneous circulation and
respiration
Data Colection


Time of collapse obtained from security
cameras or interview with witness (if not in a
public area)
AED recorded
–
–
Rhythm strip for analysis
Time of shocks
Outcome Variables

Predictor variables
–
–

Time from collapse to CPR
Time from collapse to defibrillation
Primary outcome variable
–
Survival to hospital discharge
Results





Total Cardiac Arrests - N = 148
Initial rhythm VF = 105
Initial rhythm PEA = 17
Initial rhythm asystole = 26
NB: No subjects without VF as initial rhythm
survived
Ventricular Fibrillation


105/148 arrests (71%)
15/105 unobserved collapse
–




3/15 (20%) survival to discharge
4/105 pronounced dead on scene
35/105 pronounced dead at hospital ER
10/105 admitted to hospital but died before d/c
56/105 (53%) survived to hospital discharge
Witnessed VF Arrests
90/105 VF arrests
 54% received CPR prior to arrival of AED
 Collapse to CPR 2.9+/-2.8 minutes
 Collapse to defibrillation 4.4 +/- 2.9
 Collapse to arrival of EMS 9.8+/-4.3
 Survival to hospital discharge 53 (59%)

Valenzuela, et al. NEJM 2000
The importance of time …

Defibrillation < 3 minutes after witnessed arrest
–

Defibrillation > 3minutes after witnessed arrest
–

Survival 74% (26/35)
Survival 49% (27/55)
P=0.02
Implications

Rapid AED use by targeted responders is an
effective and feasible strategy
–


Much improved survival compared with previous
estimates of out of hospital arrest survival
Minimally trained personnel (ie. security
officers) can safely and effectively operate
AEDs
Waiting for paramedics would have added ~5
minutes to time to defibrillation
Limitations

Casinos are atypical environments
–
–

Very intense monitoring
Very high density of cardiac arrests
Survival very dependent on collapse to
defibrillation interval
–
Must be able to ensure very rapid response times
Use of AEDs by a U.S. Airline
Page R, et al. NEJM. 2000;343:1210-6.

Aircraft are unique settings for cardiac arrest
–
–
–

Stress/exertion associated with flying
Disruption of circadian rhythms
Decreased oxygen tension in cabin
VF arrests on airplanes rarely survive due to
the delay obtaining EMS and defibrillation
–
Minimum 20 minutes required for emergency
landing under the best of circumstances
Methods


In 1997 American Airlines began installing
AEDs on all aircraft
All 24,000 flight attendants were trained in AED
use
–
–
–

One hour classroom
Four hour workshop
1.5 hour refresher courses
Use of AEDs monitored for a 2 year period
AED Protocol

When SSx of possible cardiac arrest
–
–
–
–
–
Patient moved to aisle, galley, bulkhead
Clothing covering chest is removed
Electrodes placed on chest
Assistance of medical personnel is solicited
Flight attendants follow instructions on AED
Results

Patient characteristics
–
N=200
 66%
male
 Mean age 58 years
–
–
191 aboard aircraft
9 in terminals
Indications for AED use









LOC - 99 (49.5%)
Chest Pain – 62 patients
Dyspnea 19
Nausea / malaise – 8
Light-headedness – 3
Palpitations – 3
Stroke – 1
Unclear reasons – 5
NB – physician assisted with 139 patients (69.5%)
Electrocardiographic Data

185 electrocardiograms available for review
–

145 – sinus rhythm
–
–




15 lost due to device malfunction, memory erased,
etc
14 sinus bradycardia
21 sinus tachycardia
8 – atrial fibrillation
3 – junctional rhythm
SVT – 1
MAT -1
Agonal rhythm – 13
 Ventricular fibrillation - 14

Shocks and survival

13/14 patients with documented VF defibrillated
–


13/13 – VF terminated with first shock
Recurrent VF in 8 patients
–

defibrillation withheld at request of one patient’s family
7/8 successfully resuscitated with subsequent shocks
NB – two patients defibrillated for presumed VF
but data lost in machine. Both patients died
Outcomes

6/15 (40%) patients defibrillated were
discharged home from hospital with full
neurologic recovery
Use of defibrillator as a monitor




Used and recorded ECG on 171 patients
without VF
101/200 patients did not have LOC
In no case was a shock inappropriately
administered
Specificity for algorithm = 100%



No shocks recommended or delivered
inappropriately
Sensitivity 100%
Specificity 100%
Page R, et al. NEJM 2000;343:1210-6.
Frequency of defibrillator use

200 uses on
–
–
–


627,956 flights
896,000,000 miles
70,801,874 passengers
AED used once per 3288 flights
Death or resuscitation once per 21,654 flights
Conclusions




AEDs 100% sensitive and 100% specific
VF terminated after first shock 100% of time
Rate of survival to discharge 40%
AEDs safe for use as monitor
–
Allows further assessment of need to divert aircraft
or utilize medications in the emergency medical kit
What about Real Public
Access Defibrillation?
Public Use of Automatic External
Defibrillators
Caffrey S, et al. NEJM. 2002;347(16):1242-1247.

AEDs shown to save lives when used by
designated personnel in certain public places

Prospective study to determine whether
random bystanders could successfully retrieve
and use AEDs for cardiac arrest
Methods



AEDs placed at 60-90 second intervals in three
Chicago area airports (100 million passengers
per year)
Use of AEDs promoted in media, pamphlets,
and videos in waiting areas
Outcomes
–
–
–
–
Time from notification of dispatchers to defibrillation
Survival at 72 hours and one year
Neurologic status
Characteristics of rescuers
Study design

Study sites
–
O’Hare – 80,000,000 passengers/year
 59 AED’s
–
Midway – 20,000,000 passengers/year
 10 AED’s
–
Meigs field – 77,000 passengers/year
 1 AED

CPR/AED training provided to 450 employees
–
–
–

Airport police
Security
Public safety dispatchers
Voluntary training of 3,000 / 40,000 other
airport employees
AEDs

Housed in glass cabinet 60-90sec apart

Highly visible signs indicating location of
nearest AED

Cabinets equipped with alarms / dispatcher
alerts
Data collected
Dispatch time
 Defibrillation time
 Recovery time (return of consciousness)
 ECG tracing
 Audio recording

Complications
Defibrillator tampering
 Inappropriate delivery of shocks
 Failure to deliver shock in VF
 Injury to rescuers

Map of O'Hare International Airport, Showing the
Locations of Automated External Defibrillators in
Public Areas and the Locations of 20 Patients with
Witnessed Cardiac Arrest.
Results



2 year study period
AED used
– O’Hare – 21 patients
– Midway – 5 patients
– Meigs – none
4 patients did not have cardiac arrest
– 2 seizures
– 1 SOB
– 1 feigned LOC to avoid immigration official

4 additional patients defibrillated with AED’s
from outside the study
–
–

Three from nearby airplanes
One from a nearby paramedic
Total of 22 cardiac arrests in study group
–
–
One traumatic arrest (33 y.o. fall)
One patient found dead on transit train

21 non-traumatic arrests
–
–
–
–

19 men / 2 women
19 travelers
1 employee
1 visitor
20 witnessed non-traumatic arrests
–
–
PEA x 2
VF x 18
Characteristics of 18
Patients with VF Who
Received Defibrillator
Shocks from AEDs
Installed at
Chicago Airport
Terminals between
June 1, 1999, and May
31, 2001.
Outcomes

Long-term survival
–
–
10/18 (56%) of all non-traumatic, witnessed VF
arrests
8/12 (67%) among those defibrillated within 5
minutes
Rescuer characteristics

All but two were good Samaritans (passengers
or employees) with no duty to act

6/11 successful resuscitations performed by
rescuers who had never been trained in AED
use (although 3 were MDs)
Implications



With public education untrained laypeople can
effectively use AEDs
Much increased survival rate (56% vs ~5% with
conventional EMS)
Cost effective
–
–
–
Estimated cost $35,000 per year
$3,000 per patient
$7,000 per life saved
Limitations

Airports are unique environments
–
–
–

Intensive monitoring of public areas
Almost always trained personnel nearby (MD, etc.)
Very busy with a high density of cardiac arrests
makes PAD cost effective
Cost calculations did not include training,
retraining, time away from job, promotion, etc.
Okay, so people in airports can
use AEDs. Can other people?
Comparison of Naive Sixth-Grade
Children With Trained
Professionals in the Use of an AED
Gundry J, et al. Circulation. 1999;100:1703-1707.

Primary end point – time to defibrillation
Secondary end points – pad placement and
safety

Subjects

–
–
15 sixth graders - only shown the pacing pads and
told they had to peel the sticker off
22 EMTs - extensive training and experience
Results

Time to defibrillation
–
–
–

Children – 90 +/- 14 seconds
EMTs – 67 +/- 10 seconds
P < 0.0001
Electrode pad placement and safety
–
100% in both groups
Conclusions

Faster time to defibrillation among EMT’s (22
seconds)
–

Clinical significance of difference is unclear
Minimal training required to educate the public
in AED use
It’s not possible to put
AEDs everywhere …. so
where should they be?
Public Locations of Cardiac Arrest
Implications for Public Access Defibrillation
Becker L, et al. Circulation. 1998;97:2106-2109.
Studied the incidence of annual cardiac
arrest in 71,000 public locations in
Seattle
 Identified sites with >0.03 cardiac arrests
per year (>1 per 30 years) and postulated
that an AED should be located at each of
these

Results



7185 non-traumatic arrests in 5 year period
1130 (16%) were in public places
Calculated that placement of 276 AED’s in 172
areas would have treated 134 cardiac arrests
over 5 years
–

Estimated 8-32 lives saved in 5 years
To cover the remaining 347 arrests occuring in
public places (excluding vehicles and public
outdoors) would require AEDs at 71,000 sites
Incidence of Cardiac Arrest per Site:
Higher-Incidence Location Categories
Circulation 1997;97(21):2106-2109
Implications



PAD has the potential to reduce the time to
defibrillation and increase survival
It is not cost effective to place AEDs at every
site where a cardiac arrest occurs
Placement of AEDs at targeted public locations
is a feasible way of reducing mortality
Defibrillation First vs CPR First In
Delayed Resuscitation
1992 American Heart
Association Guideline
“Because the most important
determinant of survival in out-ofhospital VF is rapid defibrillation,
shocks should be given as soon as a
defibrillator arrives.”
JAMA 1992;268:2212
Time-sensitive defibrillation
the controversy

Present ACLS Treatment Algorithm
–
VF = defibrillation regardless of downtime
–
3 unsuccessful shocks = CPR + drugs +
more defibrillation attempts
New evidence suggests …

Present ACLS approach not optimal in all
patients

Present approach may be harmful in delayed
defibrillation
Niemann J, et al. Treatment of prolonged VF:
immediate countershock vs high-dose
epinephrine and CPR preceding countershock.
Circulation. 1992;85:281-287.



An animal model
VF induced in dogs
After 7.5 minutes animals randomized to
1)
2)
Immediate defibrillation
5 minutes CPR plus epinephrine then defibrillation
Results

Immediate defibrillation group
–
–

Countershock rhythm of asystole in 10/14 dogs
Resuscitation successful in 3/14
CPR grounp
–
–
Countershock rhythm of asystol in 3/14 (p<0.01)
Resuscitation successful in 9/14 (p=0.01)
Resuscitation After Cardac Arrest
A 3-Phase Time-Sensitive Model
Weisfeldt M, et al. JAMA 2002:288(23);3035-3038.

Phase One – The Electrical Phase
–
–
–
–
–
First 4 minutes of VF
Excellent data to support immediate defibrillation
(class I recommendation)
Basis of the highly effective Implantable
Cardioverter Defibrillator (ICD)
Survival rates of ~50% shown in several trials
This strategy is presently the standard of care for all
VF regardless of time
Phase 2 – The Circulatory Phase
–
–
–
–
4-10 minutes after arrest
Many animal models have shown improved survival
when CPR for 1-6 minutes provided provided to
defibrillation
Hypothesis: during this time period oxygen delivery
(CPR) takes precedence over defibrillation
Implication – EMS resuscitates most people during
this phase …
Phase 3 – The Metabolic Phase





>10 minutes after arrest
Very poor outcomes with defibrillation +/- CPR
Global whole-body ischemia
Reperfusion can contribute to cell death
independently of ischemia
Promising therapies include
–
–

Therapeutic hypothermia (even hours after ROSC)
Apoptosis inhibitors
No practical, proven therapies for EMS yet …
So …. since we’ve already
figured out phase one, and
have no clue about phase
three, we should work to
improve phase two…
right?
Influence of CPR Prior to
Defibrillation in Patients With Outof-Hospital Ventricular Fibrillation
Cobb L, et al. JAMA 1999:281(13);1182-1188.

Background
–
–
–
Introduction of AE’s in Seattle decreased time to
defibrillation by 3-4 minutes
Survival increased in certain subsets of patients
(short downtime) but overall survival was
unchanged
animal models showed increased survival with
prolonged downtimes if CPR initiated before
defibrillation

Objective
–

Methods
–

To evaluate effect of 90 seconds CPR prior to
delivery of first shock
Observational study of 42 months of preintervention
and 36 months post intervention data
Intervention
–
Provision of 90 seconds CPR prior to delivery of first
shock

Hypothesis
–

Greater survival benefit in cases where the AED
response intervals were longest
Outcomes
–
–
Primary – survival to hospital discharge
Secondary – neurologic status
Results
JAMA 1999;281(13):1182-1188.
JAMA 1999;281(13):1182-1188.
Survival of the 2 Study Periods According to the
Response Intervals of the First Arriving Unit
JAMA 1999;281(13):1182-1188.
JAMA 1999;281(13):1182-1188.
Conclusions




Results are consistent with animal models
Current ACLS guidelines may not provide
optimal treatment for prolonged VF
Patients with delayed EMS response
(>4minutes) were most likely to benefit from
CPR first.
Other patients did not seem harmed by
provision of 90 seconds CPR
Limitations

90 second duration of CPR arbitrary. Further
studies needed to optimize duration.

Prospective RCT’s needed to definitively prove
the superiority of CPR before defibrillation

Did not include Norwegians
Delaying Defibrillation to Give Basic CPR to
Patients with Out-of-Hospital VF
A Randomized Trial
Wik L, et al. JAMA 2003;289(11):1389-1395.

Context
–

A non-randomized study (Cobb, et al.) showed that
90 seconds of CPR prior to defibrillation improved
survival in patients where response time was >4
minutes
Objective
–
Determine effect of CPR prior to defibrillation in
patients with VF and response times more and less
than five minutes
Methods
Prospective, randomized trial of 200 patients
with out of hospital VF
Patients received either


i.
ii.


Standard care with immediate defibrillation (N=96)
3 minutes of CPR prior to defibrillation (N=104)
The remainder of the resuscitation proceeded
identically
A priori subgroup analysis of cases with
response times more and less than five
minutes
Flow Diagram of CPR First and Standard Care
JAMA 2003;289(11)1389-1395
Outcomes

Primary
–

Survival to hospital discharge
Secondary
–
–
–
Hospital admission with ROSC
1 year survival
Neurological outcome
Results
JAMA 2003;289(11)1389-1395
JAMA 2003;289(11)1389-1395
Estimated Probability of Survival to Hospital
Discharge Plotted Against Response Time
JAMA 2003;289(11)1389-1395
Conclusions


No overall differences in survival for patients
with out of hospital VF among two study
groups
Among patients with longer ambulance times
(>5 minutes), those receiving prior CPR had
improved rates of hospital discharge and one
year survival
Further research necessary


What is the optimum amount of time initial CPR
should be performed
At precisely what time does the optimum
strategy become one of performing CPR first
Prehospital Triage
Is it accurate?
Is it safe?
Background

Overcrowding of ED’s is a problem

Reasons are complex and multifactorial

Prehospital discharge and/or diversion of
patients to alternate sites of care has been
suggested as a way of alleviating the problem
Schmidt T, et al. Evaluation of Protocols
Allowing Emergency Medical Technicians to
Determine Need for Treatment and Transport.
Acad Emerg Med 2000;7:663-669

Study Objective
“…whether it is feasible and safe to
develop a set of protocols allowing
EMTs to offer alternatives to an ED
for patients who access the 911
system.”
Study Design


Prospective observational study of scene
categorization by EMTs using a triage protocol
Associated retrospective chart review by an
expert panel.
Triage Protocols


Developed by multidisciplinary expert panel
Protocol assumptions
1.
Triage guidelines are conservative
•
2.
3.
overtriage better than undertriage
Triage guidelines must be easy to use
Triage guidelines must offer appropriate
alternatives to ambulance transport and not refuse
anyone proper care.
Traige Protocols

Categorize patients by presentation to
1.
2.
3.
4.
Ambulance (ALS) transport to an ED
Transport to an ED by alternate means (BLS, etc.)
Referral to a primary care provider within 24 hours
Treatment at the scene only
Cardinal Conditions
protocol always requires ALS transport (category 1)




Unconscious or ALOC
Respiratory distress
Sustained abnormal VS
– sBP<90
– HR<60 or >110
– RR<12 or >24
Compromised airway




Uncontrolled bleeding
Suspected abuse of any
type
Suspected c-spine injury
Infants less than 3
months of age
Triage Protocol Example
MSK Injury

Category 1:
–
–
–
–
–
–
–

Amputations
Femur fracture
Multiple trauma
NV compromise
Severe pain
Open fractures
Hip#/disclocation
Category 2:
–
–
Avulsion of digit tips
Isolated extremity injury

Category Three
–
–

Isolated injury with no loss
of function, point
tenderness, or swelling
Need for tetanus
prophylaxis
Category Four
–
–
Contusions
Abrasions
Training

Video describing study and protocols

Workbook and 50 question exam asking EMTs
to categorize patients using protocols

Passing score: 90%
Outcome Measure
Critical Events in Ambulance
used to identify need for ambulance transport








CPR
Intubation
Assisted ventilation
Extensive bleeding controlled
in ambulance
Cardioversion or pacing
SaO2<90% or transported on
O2
sBP < 90 (or age appropriate)
Dysrhythmia treated








Traction placed for femur
fracture
Needle thoracotomy
Cricothyrotomy
GCS<12
Spinal precautions
Received medication during
transport
Criteria for trauma activation
>1 litre fluid infused
Results

1300 patients transported and categorized
–
–
–
–
1023 patients category 1 (needed ambulance)
277 patients category 2
63 patients category 3
14 patients category 4
Comparison of triage category and
critical events



277 patients categorized as not needing
ambulance
30/277 (10.8%) – had critical event
7/277 (2.5%) – had critical event warranting
ambulance transport




Acministration of NTG for chest pain
GCS 10, 12
SaO2 83%, 86%
sBP <90mmHg



6/7 critical events requiring ALS ambulance
were misclassified (provider error)
1/7 critical events involved protocol error
Conclusion
–

3-11% of patients categorized as not needing an
ambulance actually did
Limitations
–
–
BLS considered acceptable option for category 2
patients
Applicability to ALS only system unclear
Pointer J et al. Can paramedics using
guidelines accurately triage patients?
Ann Emerg Med 2001;38:268-277.

Study Hypothesis
“…paramedics can use written
guidelines to accurately and safely
triage patients in the field.”
Protocol

Paramedics assign triage category based on
chief complaint
–

Used same four triage categories as Schmidt
Added two cardinal conditions
–
–
Trauma activation (was a critical event in Schmidt’s
study)
Suspected stroke <6 hours ago

Training
–
–
–
30 minute overview of project
Study workbook and multiple choice exam
Passing grade = 80%
Data Analysis

Gold Standard
–
–
Panel of three emergency physicians
Assignment to same categories using retrospective
review of:



ED treatment / monitoring
Diagnosis
Need for admission
Results
Paramedic Score
MD
Score
1
2
3
4
1
127
592
83
2
2
0
140
52
4
3
1
81
65
1
4
0
17
14
1
Results

Need to come to ED (categories 1,2) vs no
need to come to ED (categories 3,4),
kappa=0.28

113/1180 (9.6%) undertriaged by paramedics
–
–
Represent 43 different diagnoses
19.5% admitted

99/1180 (8.4%) triaged away from ED
inappropriately
–
–

paramedics category 3-4, MD category 1-2
Represents 55% of all patients rated 3-4 by
paramedics
Incidence of incorrect assignment of category 4
(no MD follow-up necessary ) was 2.6%
–
–
Three patients required admission
CHF, ARF, psychosis
Why were patients undertriaged?

Guidelines not applied correctly in 55 cases

No appropriate guideline in 32 cases

9 patients with “occult conditions” not
detectable in field even with improved
guidelines and training
Conclusions

Even with use of guidelines, the accuracy of
paramedic triage was unacceptably low

Given the potentially serious consequences of
undertriage, 9.6% is an unacceptably high rate
Silvestri S et al. Can paramedics accurately
identify patients who do not require
emergency department care. Prehospital
Emergency Care 2002;6:387.

Objective
“To determine whether paramedics can
identify patients contacting 9-1-1 who do
not require emergency department care.”
Methodology

Subjective (no guidelines) classification of
patients by paramedics as either
1.
2.
3.
4.
Needs ED care
Needs medical care within 24 hours
Needs medical care in greater than 24 hours
Medical care is not required
Gold Standard

Patients deemed to have required ED care if
they:
1.
2.
3.
Were admitted
Required surgical, surgical subspecialty, obstetric,
or gynecologic consultation
Required advanced radiological procedures
(excluding plain films)
Results

313 patients enrolled

Paramedic assessment:
–
–
–
–
81% sensitivity (CI 72-88%)
34% specific (CI 28-41%)
50% PPV (CI 44-56%)
68% NPV (CI 56-79%)


Paramedics triaged 85 patients as not needing
ED transport
Of these, 27 (32%) met criteria for ED
treatment
–
–
–
–
15 (18%) admitted
5 (6%) admitted to ICU
19 (22%) required advanced radiologic imaging
1 (1%) required surgical consultation

Of 15 admitted patients felt not to need ED,
paramedics stated
–
–
–
2 could be seen at health care clinic within 24 hours
1 could be seen >24 hours later
12 did not require any medical attention
Final Diagnoses in these patients







Femur fracture
Aspiration pneumonia
MRSA pneumonia
Osteomyelitis
Pancreatitis
Tibia fracture
Septic arthritis







CHF
CVA
Cellulitis
Pericardial effusion
Syncope
Hepatitis
Cocaine toxicity
Conclusions

Unacceptable rate of undertriage in seriously ill
patients (6 ICU admissions!)

“…use of paramedic discretion to determine
whether patients calling 9-1-1 required ED care
was inaccurate.”
Hauswald M. Can paramedics safely decide
which patients do not need ambulance
transport of emergency department care?
Prehospital Emergency Care 2002;6:383-386.

Objectives
“To determine whether paramedics
can safely decide which patients do
not require ambulance transport or
emergency department care.”
Methods


Prospective survey and linked chart review
Paramedics were asked:
–
–

“Could this patient have been safely transported by
a non-medical transport service?”
“Could this patient have been safely transported to a
clinic or urgent care center?”
Responses were yes/no, and were purely
subjective (ie: no guidelines used)
Gold Standard Criteria

ED MD reviewed medical records

Ambulance transport was “needed” if the
charted DDx included Dx that could
necessitate treatment in an ambulance
–
–
eg. ASA for possible MI
eg. rapid transport for head injury

ED care was “needed” if patient required
diagnostic or therapeutic procedures not
available outside of hospital.
–
–
Eg. CT scans
Eg. fracture reduction
Results

183 patients enrolled

Paramedics felt that
–
97 patients (53%) did not need ambulance

–
23 (24%) of these needed ambulance by gold standard
71 (38%) were appropriate for triage to alternative
care site

32 (45%) of these needed ED care by gold standard
ED Diagnoses of patients for whom
paramedics recommended alternative
transportation








Alcohol intoxication
Assault
Coma x 4 days
Chest pain
Dislocated hip
Hepatic failure
Hypoxia
Laceration (severe
bleeding)






MVC / C-spine
precautions
Narcotic overdose
Pyelonephritis – 6
months pregnant
Adult first seizure
Sepsis
syncope
ED Diagnoses of patients for whom
paramedics recommended alternative
care








Assault (multiple injuries)
Child abuse
Chest pain (neg workup)
Alcohol abuse
Multiple medical
problems (died in
hospital)
Adult first seizure
Narcotic OD
Active labor






Laceration (required
extensive repair)
MVC – C-spine
precautions
Multiple drug overdose
Post traumatic headache
Hypoxia
Hepatic failure
Conclusions



“…prehospital triage will result in an
unacceptably high number of inappropriate
non-transports.”
“Allowing paramedics to transport to non-ED
care settings will result in many patients
requiring a second transport from clinic to ED.”
“Paramedics as currently trained are not
capable of making these decisions accurately
or safely.”
Conclusions



Paramedics are taught to, and are excellent at
providing initial treatment, and transport of
patients with life-threatening complaints
They are not taught to diagnose and manage
less acute complaints
Many important diagnoses and disposition
decisions cannot be made without tests
available in the ED
Implications for Calgary



Some of the literature not directly applicable to
Calgary as we have an ALS only system
Calgary EMS has treat and release protocols
for hypoglycemia and narrow complex
tachycardia
Ongoing QA is required to ensure that the
protocols are adequate and that they are being
followed

It is possible that Urgent Care Centres (eg. 8th
and 8th) will begin to accept ambulances

Protocols must be established and ongoing QA
is necessary to ensure that patients are
appropriately triaged to these sites