Resuscitation (2006) 71, 47—55
Clinical consequences of the introduction of
mechanical chest compression in the EMS system
for treatment of out-of-hospital cardiac arrest—–
A pilot study?
Christer Axelssona,∗, Johan Nestinb, Leif Svenssonc,
˚ Asa B. Axelssond, Johan Herlitze
aGothenburg EMS-system, Box 5204, SE-402 24 G¨ oteborg, Sweden
bSamariten EMS-system, Huddinge Fire station, SE-141 41 Huddinge, Sweden
cDivision of Cardiology, South Hospital, SE-118 83 Stockholm, Sweden
dSahlgrenska Academy, Institute of Health and Care Sciences, Box 457, SE-405 30 G¨ oteborg, Sweden
eDepartment of Metabolism and Cardiovascular Research, Sahlgrenska University Hospital,
SE-413 45 G¨ oteborg, Sweden
Received 13 January 2006; accepted 16 February 2006
Aim: To evaluate the outcome among patients suffering from out-of-hospital car-
diac arrest (OHCA) after the introduction of mechanical chest compression (MCC)
compared with standard cardiopulmonary resuscitation (SCPR) in two emergency
medical service (EMS) systems.
Methods: The inclusion criterion was witnessed OHCA. The exclusion criteria were
age <18 years, the following judged etiologies behind OHCA: trauma, pregnancy,
hypothermia, intoxication, hanging and drowning or return of spontaneous circu-
lation (ROSC) prior to the arrival of the advanced life support (ALS) unit. Two MCC
devices were allocated during six-month periods between four ALS units for a period
of two years (cluster randomisation).
Results: In all, 328 patients fulfilled the criteria for participation and 159 were
allocated to the MCC tier (the device was used in 66% of cases) and 169 to the SCPR
In the MCC tier, 51% had ROSC (primary end-point) versus 51% in the SCPR tier.
The corresponding values for hospital admission alive (secondary end-point) were
38% and 37% (NS).
?A Spanish translated version of the summary of this article appears as Appendix in the online version at
∗Corresponding author. Tel.: +46 705 400 304.
E-mail address: firstname.lastname@example.org (C. Axelsson).
0300-9572/$ — see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.
48C. Axelsson et al.
In the subset of patients in whom the device was used, the percentage who had
ROSC was 49% versus 50% in a control group matched for age, initial rhythm, aetiol-
ogy, bystander-/crew-witnessed status and delay to CPR. The percentage of patients
discharged alive from hospital after OHCA was 8% versus 10% (NS) for all patients and
2% versus 4%, respectively (NS) for the patients in the subset (where the device was
used and the matched control population).
Conclusion: In this pilot study, the results did not support the hypothesis that the
introduction of mechanical chest compression in OHCA improves outcome. However,
there is room for further improvement in the use of the device. The hypothesis that
this will improve outcome needs to be tested in further prospective trials.
© 2006 Elsevier Ireland Ltd. All rights reserved.
Cardiopulmonary resuscitation (CPR) skills vary
among health care professionals.1It is easy to
understand how difficult it can be to perform CPR
during an ambulance transport, both on a stretcher
and in a moving ambulance.2,3Fatigue is another
important factor that occurs within 2—3min after
the start of CPR. Fatigue reduces the effect of
the chest compressions for some minutes before
the rescuer feels subjectively exhausted.4Wik et
al. concluded that, during out-of-hospital cardiac
arrest (OHCA), chest compressions were delivered
for only half the time and most were too shallow.
They also discovered that ECG analysis and defib-
rillation accounted for only 20% of intervals with-
out chest compression.5Sanders and Ewy consider
that CPR guidelines are too complex and result in
patients not receiving known benefits such as chest
compressions.6The Swedish national registry states
that the mortality rate in out-of-hospital cardiac
arrest is more than 95%.7
In patients with OHCA, survival with a good neu-
rological outcome is dependent upon the genera-
tion of blood flow to the heart and brain during
resuscitation.8A coronary perfusion pressure (CPP)
of 15mm Hg also appears to be necessary for the
return of spontaneous circulation (ROSC).9Blood
flow and CPP during cardiac arrest are related to
the quality and continuity of chest compressions
during CPR.10It has recently been suggested that
CPR prior to defibrillation may be of benefit when
there have been more than 5min of delay before
shock to patients suffering from out-of-hospital
ventricular fibrillation (VF).11,12Steen et al. found
in animal studies that the most optimal outcome
for ROSC (after cardiac arrest for 6.5min) occurred
3min before and during defibrillation. Steen et al.
used a mechanical device to perform chest com-
pressions during defibrillation.10
LUCAS (Lund University Cardiopulmonary Assist
System, Jolife AB, Lund, Sweden) is a gas-driven
CPR device providing mechanical active compres-
sion/decompression (ACD) CPR.13Active decom-
pression increases the naturally occurring nega-
tive intrathoracic pressure by physically expand-
ing the chest wall. During manual CPR, incomplete
chest wall recoil is common, resulting in signifi-
cantly less blood flow back to the heart.14In ran-
carotid/artery blood flow were found with mechan-
ical ACD CPR compared with manual CPR.13,15At
present, no randomised trial shows improved sur-
vival with mechanical chest compressions, but suc-
cessful clinical cases when prolonged CPR was indi-
cated have been published.16—18
The purpose of the present study was to describe
the outcome during a limited period in witnessed,
non-traumatic, out-of-hospital cardiac arrest (>18
years) treated with standard CPR or standard CPR
followed by mechanical ACD CPR in the Municipali-
ties of G¨ oteborg and S¨ odert¨ alje, Sweden.
Patients and methods
This descriptive, non-randomised, controlled trial
in G¨ oteborg and S¨ odert¨ alje, Sweden, was approved
by the Ethics Committee at G¨ oteborg/M¨ olndal
and Stockholm Universities, Sweden. During a
period of 2 years (22 May 2003—25 May 2005)
in G¨ oteborg/M¨ olndal and 1 year (1 September
2003—31 August 2004) in S¨ odert¨ alje, LUCAS was
exchanged between the four Advanced Life Support
(ALS) units for approximate 6-month periods. This
trial therefore used a cluster method to create a
Patients with a witnessed OHCA of presumed car-
diac aetiology were enrolled. Exclusion criteria
were age <18 years, with trauma, pregnancy,
Clinical consequences of the introduction of mechanical chest compression in the EMS system 49
hypothermia, intoxication, hanging and drowning,
as the judged aetiologies of OHCA, Return Of Spon-
taneous Circulation (ROSC) before the arrival of the
second tier and other reasons, such as terminal ill-
The emergency medical services (EMS) system in
G¨ oteborg/M¨ olndal serves about 542,000 inhabitants
in an area of 595km2. The S¨ odert¨ alje EMS sys-
tem serves about 344,000 inhabitants in an area
of 580km2. In both cities, ambulances were dis-
patched according to a two-tier system—–i.e., for
of health, an ALS unit, if available, and the near-
est Basic Life Support (BLS) unit were dispatched
simultaneously. The BLS units in both cities were
staffed by at least one nurse and the ALS units by a
paramedic and a well-trained anaesthesia nurse. In
the G¨ oteborg/M¨ olndal EMS system, three ALS units
were available for 24h every day. In S¨ odert¨ alje,
one ALS unit was available every day between 7
a.m. and 7 p.m. All OHCA were treated according
to American Heart Association and European Resus-
citation Council guidelines.
Before starting the study, 50 EMS personnel
(paramedics and anaesthesia nurses) were trained
to perform mechanical ACD CPR and re-trained in
manual chest compressions. The instructor was an
anaesthesia nurse educated as a LUCAS instructor
by Jolife AB. Each training session lasted 3h and
ended with a practical and a theoretical test. The
practical test was a scenario at which the EMS per-
sonnel worked in pairs. On arriving, the first EMS
staff member immediately started manual chest
compressions on a manikin, while the second pre-
pared LUCAS for attachment to the manikin. To pass
the test, they had to minimise the hands-off time
between manual and mechanical chest compres-
sions to <20s. During training, they were informed
about the importance of minimising hands-off sit-
uations and preparing for fatigue by rotating the
rescuers during manual CPR. In the trial group, they
were told to attach LUCAS to the patient as soon as
possible after arriving.
Before every period in which the device was
used, the EMS personnel took part in a 2h re-
training session. During the trial period, personnel
were contacted regularly by the project manager
and had the opportunity to call a special phone
number to report aspects such as ethical and safety
questions to the project manager.
LUCAS weighs about 6.5kg and is kept in a back-
pack. To run LUCAS, we used compressed air in
double cylinders (Spirolight 3, 4 from Interspiro)
made of composite material. One double cylinder
weighs about 10kg and runs LUCAS for approxi-
Data relating to the witnessed cardiac arrest cohort
were obtained from the G¨ oteborg and S¨ odert¨ alje
EMS. Data were also collected from the dispatch
centre and national registry for out-of-hospital car-
diac arrest in Sweden. Further medical data relat-
ing to patients admitted alive to hospital were
obtained from hospital records. The primary study
end-point was ROSC at any time during the treat-
ment and the secondary end-point was survival
at hospital admission in witnessed cardiac arrest.
Additional major clinical study end-points, anal-
ysed for all enrolled patients, were survival to
hospital discharge and neurological recovery. Data
were collected according to the Utstein criteria and
the Glasgow-Pittsburgh CPC (Cerebral Performance
The distribution of variables is given as means,
medians and percentages.
For comparison between groups of continuous vari-
ables, Fisher’s non-parametric permutation test
was used. For comparison of dichotomous variables
between groups, Fisher’s exact test was used. A
p-value of less than <0.05 was regarded as signif-
icant. Two-tailed tests were applied. Only p-values
of <0.20 are given in the tables.
About 536 patients who suffered from an out-of-
were available for inclusion in the trial. About 379
patients fulfilled the inclusion criteria and 51 of
these met various exclusion criteria (Table 1). The
rhythm on of the arrival of the ALS unit.
50 C. Axelsson et al.
and exclusion criteria
Number of patients who fulfilled inclusion
aPatients could have more than one exclusion criterion.
As a result, 328 patients were evaluated (159 in
the mechanical chest compression group and 169 in
the control group). From now on, these two groups
will be compared.
Characteristics (Table 2)
The patients were relatively old (mean age 71
years in both groups) and more than one-third
were women. About half had a history of coronary
artery disease and nearly one-third had a history of
heart failure. No significant differences were found
between the two groups (Table 2).
Place of cardiac arrest (Table 3)
The majority of cardiac arrests (about two-thirds)
took place in the patient’s home. There was
no significant difference between the two groups
Aetiology (Table 3)
ology was more frequent in the control group.
Status on the arrival of the BLS team
A high percentage of patients received bystander
CPR prior to the arrival of the rescue team. Less
than one-third had ventricular fibrillation on the
arrival of the BLS team. The largest percentage had
asystole. With the exception of pulseless electrical
activity, which was more frequent in the mechani-
cal chest compression group, the two groups were
similar in all comparisons (Table 4).
Status on the arrival of the ALS team
The results appeared to be fairly similar to those
relating to the arrival of the BLS team.
Delay (Table 5)
The median delay from cardiac arrest to the arrival
of the ALS unit was 12min. Among the patients
in the mechanical chest compression group, the
median delay between cardiac arrest until the start
of mechanical chest compression was 18min (i.e. a
delay between the arrival of ALS and the start of
mechanical chest compression of 6min). The ALS
unit arrived on the scene a median of 2min after
the BLS unit. There were no significant differences
between the two groups (Table 5).
Age, sex and previous history
Mechanical chest compression,
Control, n=169 (%)
Age (years; mean±S.D.)
Previous history (%)
Coronary artery disease (34/35)a
Congestive heart failure (37, 40)
Peripheral vascular disease (40, 40)
Cerebrovascular disease (37, 39)
Pulmonary disease (37, 35)
Diabetes (39, 36)
Other (35, 37)
aNumber of patients with missing information.
**p-Values denoted if <0.2.
Clinical consequences of the introduction of mechanical chest compression in the EMS system 51
Place and aetiology of cardiac arrest
compression, n=159 (%)
Control, n=169 (%)
Place of cardiac arrest (1.0)
On the street
At a nursing home
In the ambulance
Some other place
aA few patients were judged to have more than one aetiology behind their cardiac arrest.
**p-Value denoted if <0.2.
ROSC and survival (Table 6)
For patients allocated to the mechanical chest com-
pression group (n=159), the device was actually
given below. When all the patients (including those
in whom the device was not used) were included
in the analyses, there was no significant differ-
ence between the groups with regard to ROSC, sur-
vival to hospital admission or to hospital discharge
When patients in whom the device was used
were compared with a matched control population
according to age, initial rhythm, bystander-/crew-
witnessed status, aetiology and delay to start of
CPR, no difference was found between the two
groups in any of the variables evaluated. The results
did not change when relating outcome to the inter-
life support unit
Rhythm and treatment prior to and on arrival of the basic life support unit and on arrival of the advanced
Mechanical chest compression, n=159 (%)
Control, n=169 (%)
Health care professional*
Rhythm on arrival of BLS
Pulseless electrical activity
Treatment by BLS
Rhythm on arrival of ALS
Pulseless electrical activity
*Percentage of patients that received bystander CPR.
52 C. Axelsson et al.
Control, n=169 (%)
Time from cardiac arrest to (median; min)
Call for ambulance (n=126, 125)
Start of CPR (n=154, 160)
First ECG recording (n=142, 147)
Arrival of BLS (n=125, 123)
Arrival of ALS (n=125, 126)
First defibrillation (n=58, 65)
Start of mechanical chest compression (n=77)
End of mechanical chest compression (n=64)
Return of spontaneous circulation (n=71, 76)
aCrew-witnessed not included.
bWhen VF/VT was first rhythm, crew-witnessed included.
*p-Value denoted if <0.2.
val between collapse and the arrival of the ALS
CPC score (Figure 1)
As shown in the figure, the majority of survivors had
a CPC score of 1 or 2 in both groups with no signifi-
cant difference between the groups (Figure 1).
Three technical problems were reported during
this trial. Two were solved using a longer air hose
between the device and the double cylinders con-
taining compressed air and one by replacing the
gasket in the regulator with a low-temperature-
resistant gasket. Both groups reported rib frac-
tures. In the group receiving both manual and
mechanical CPR, one case of suspected flail
chest and several cases of skin injury from
the suction cup were reported. The ALS per-
sonnel also noted that, during a long sequence
of mechanical chest compressions, the device
slid in an abdominal direction, especially when
the patient was lying on a stretcher during the
ROSC and survival
Mechanical chest compression (%)
Patients in whom the device was used vs. a matched control group
Delay to ALS arrival ≤11mina
Delay to ALS arrival >11min
aDelay between cardiac arrest and arrival of ALS was missing in 4 and 2 patients, respectively.
*p-Value denoted if <0.20.
Clinical consequences of the introduction of mechanical chest compression in the EMS system53
Figure 1 CPC-score at hospital discharge among sur-
vivors. In total n=29; MCC: mechanical chest compres-
sion; SCPR: standard cardiopulmonary resuscitation.
Reasons for not using mechanical chest com-
Reasons according to field
Patients too small
Patients too large
Lack of experience, forgot
to bring the device
Cardiac arrest too close to
Chest pain, cardiovascular
disease, angina pectoris
Diabetes, back and
Reasons for not using mechanical chest
compression (Table 7)
There were a variety of reasons for not using the
device in 34% of the cases (Table 7).
The two most common reasons were:
(a) The dispatchers had interpreted the case as
something other than a cardiac arrest.
(b) The patients had such a short delay from the
arrival of the ALS unit until ROSC that there
was no time to use the device.
Our hypothesis that mechanical chest compres-
sion improves outcome after out-of-hospital car-
diac arrest was not confirmed. However, there are
a number of ways in which the use of mechanical
chest compression could be improved.19
(1) A critical issue is the way in which patients are
randomised. Scientifically, the most appropri-
ate way to randomise patients is by envelope
randomisation or its equivalent. However, this
procedure increases the delay until the start
of the intervention. Further, extra equipment
could compromise the control group, since it
has to be carried back to the ambulance while
the personnel perform manual chest compres-
sions. However, as is clearly shown in this pilot
study with cluster randomisation, there is a risk
that the intervention is not always used in the
intervention group. This will make it more dif-
ficult to interpret the results. However, with
this type of randomisation, knowledge about
the feasibility of the intervention will increase.
(2) Another critical issue is the delay between col-
lapse and the start of the intervention. In this
trial, the median delay was 18min, which is
very long. One of the major components was
the 6min delay between the arrival of the ALS
unit and the start of mechanical chest compres-
sion. One interesting observation was, however,
that this delay decreased from 6min in the first
indicates that, with prolonged experience, the
ambulance crew can learn more quickly to use
this, the delay between collapse and the start
of mechanical chest compression is unaccept-
It is possible that a shorter delay to start of treat-
ment might improve outcome. It is important to
stress that, during the interval between the arrival
of BLS and the start of mechanical chest compres-
sion, manual chest compressions were performed—
–hopefully in the optimal way.
One way to further reduce the delay would be to
introduce the device in the first tier, which reached
the patient 2min before the ALS unit. However,
there will always be a delay from the arrival of the
rescue team to setting up the equipment and organ-
ising things according to field conditions. So, even
if the device is placed on the first tier, the per-
sonnel will always start the resuscitation attempt
with manual chest compressions, because this is the
quickest way to start, there is no need for extra
equipment and manual chest compression is always
It has been suggested that, if there is a pro-
longed delay until the arrival of the rescue team
and the patient is found in ventricular fibrillation,
54C. Axelsson et al.
chest compressions should be performed first.11,12
This strategy was not used in the present study. Fur-
thermore, mechanical chest compressions were not
performed during nor directly after defibrillation.
So there are various ways in which the use of the
device could be optimised.
The device was not used in 54 (34%) of the cases.
Sixteen of these (three technical problems, seven
forgotten, two of three cases too close to hospital
and four unconscious patients) were attributed to
lack of routine, broad outlines and start-up prob-
lems. This means that the remaining 38 cases (24%)
of patients with a cardiac arrest would not be given
mechanical chest compressions in an equivalent
EMS system with an experienced crew and with the
device placed on the second tier. It is important to
investigate these figures in more detail whether the
device not used in the remaining 24% of the cases in
the intervention group? One major reason was that
the cardiac arrest was frequently not interpreted
as such by the dispatchers and the ALS unit crew
therefore failed to take the device to the patient.
It is important to stress that the majority of the
crew-witnessed cases are concealed in this group.
The ALS units in G¨ oteborg undertake about 11,000
assignments each year, 200—300 of which are car-
associated with introducing bulky extra equipment.
Most cardiac arrests occur at home and bringing
extra equipment from the ambulance (in addition
to standard equipment) requires better information
or broader outlines.
Other reasons included the patient having a
pulse-generating rhythm during the first few min-
utes after the ALS unit arrived and before LUCAS
was prepared for use. Consequently, among the 54
patients in whom the device was not used, 22% had
a pulse-generating rhythm during the first few min-
utes after the arrival of the ALS unit. This explains
why these 54 patients had a relatively high rate of
survival to hospital discharge (19% versus 2% among
patients in whom the device was actually used).
The device proved to be impact resistant and
dependable. According to the ALS personnel, resus-
citation efforts were facilitated by freeing the
hands of the rescuer from chest compression. For
the same reasons, safety increased during trans-
port in a moving ambulance. To further increase
the level of safety, we used compressed air instead
of oxygen during this trial. Air consumption was
higher than that mentioned in the manual, thereby
reducing the uptime. The loud noise is also a fac-
tor that we hope the manufacturer will address
in later versions. A more serious problem is the
fact that the device slid from its original posi-
tion, mostly in an abdominal direction. This prob-
lem occurs frequently when LUCAS cannot be kept
exactly vertical, during transport, for example. The
problem was resolved locally in an appropriate way
using a specially designed strap between the device
and the patient. A permanent solution is vital, as
injuries have been reported that might emanate
from the device sliding from its original position.20
With regard to the patient characteristics,
i.e. age, previous history, witnessed status and
bystander CPR, there were no unexpected findings.
The percentage of patients found in ventricular fib-
rillation was low, in agreement with many recent
Conclusions and implications
In this pilot study, our hypothesis that mechani-
cal chest compression improves outcome after out-
of-hospital cardiac arrest was not confirmed. It is
important to perform further randomised trials to
investigate how to use mechanical chest compres-
sions in accordance with pre-hospital standards and
to determine whether mechanical chest compres-
sion in other conditions would improve outcome
in OHCA. Mechanical chest compressions might be
started at an earlier stage if the device is placed in
the first responding tier. If so, it is extremely impor-
tant to create an algorithm for mechanical chest
1. Thoren AB, Axelsson A, Holmberg S, Herlitz J. Measurement
of skills in cardiopulmonary resuscitation—do professionals
follow given guidelines? Eur J Emerg Med 2001;8(3):169—76.
2. Stapleton ER. Comparing CPR during ambulance transport.
Manual vs. mechanical methods. Jems 1991;16(9):63—72,
66, 68 passim.
3. Sunde K, Wik L, Steen PA. Quality of mechanical, manual
standard and active compression-decompression CPR on the
arrest site and during transport in a manikin model. Resus-
4. Hightower D, Thomas SH, Stone CK, Dunn K, March JA. Decay
5. Wik L, Kramer-Johansen J, Myklebust H, et al. Quality of
arrest. JAMA 2005;293(3):299—304.
6. Sanders AB, Ewy GA. Cardiopulmonary resuscitation in the
real world: when will the guidelines get the message? JAMA
7. Holmberg M, Holmberg S, Herlitz J, G˚ ardel¨ ov B. Survival
after cardiac arrest outside hospital in Sweden. Resuscita-
8. Berg RA, Sanders AB, Kern KB, et al. Adverse hemodynamic
effects of interrupting chest compressions for rescue breath-
ing during cardiopulmonary resuscitation for ventricular fib-
rillation cardiac arrest. Circulation 2001;104(20):2465—70.
Clinical consequences of the introduction of mechanical chest compression in the EMS system55 Download full-text
9. Paradis NA, Martin GB, Rivers EP, et al. Coronary perfusion
pressure and the return of spontaneous circulation in human
cardiopulmonary resuscitation. JAMA 1990;263(8):1106—13.
10. Steen S, Liao Q, Pierre L, Paskevicius A, Sjoberg T. The
critical importance of minimal delay between chest com-
pressions and subsequent defibrillation: a haemodynamic
explanation. Resuscitation 2003;58(3):249—58.
11. Cobb LA, Fahrenbruch CE, Walsh TR, et al. Influence
of cardiopulmonary resuscitation prior to defibrillation in
patients with out-of-hospital ventricular fibrillation. JAMA
12. Wik L, Hansen TB, Fylling F, et al. Delaying defibrillation to
give basic cardiopulmonary resuscitation to patients with
out-of-hospital ventricular fibrillation: a randomized trial.
13. Steen S, Liao Q, Pierre L, Paskevicius A, Sjoberg T. Evaluation
of LUCAS, a new device for automatic mechanical compres-
sion and active decompression resuscitation. Resuscitation
14. Frascone RJ, Bitz D, Lurie K. Combination of active com-
pression decompression cardiopulmonary resuscitation and
the inspiratory impedance threshold device: state of the art.
Curr Opin Crit Care 2004;10(3):193—201.
15. Rubertsson S, Karlsten R. Increased cortical cerebral blood
flow with LUCAS; a new device for mechanical chest com-
pressions compared to standard external compressions dur-
ing experimental cardiopulmonary resuscitation. Resuscita-
16. Holmstrom P, Boyd J, Sorsa M, Kuisma M. A case of hypother-
mic cardiac arrest treated with an external chest com-
pression device (LUCAS) during transport to re-warming.
17. Vatsgar TT, Ingebrigtsen O, Fjose LO, Wikstrom B, Nilsen JE,
Wik L. Cardiac arrest and resuscitation with an automatic
mechanical chest compression device (LUCAS) due to ana-
phylaxis of a woman receiving caesarean section because of
pre-eclampsia. Resuscitation, 2005.
18. Wik L, Kiil S. Use of an automatic mechanical chest com-
19. Jerry N, Peter B, editors. European resuscitation coun-
cil guidelines for resuscitation 2005 (Resuscitation 2005;
20. Englund E, Kongstad PC. Active compression-decompression
CPR necessitates follow-up post mortem. Resuscitation
21. Herlitz J, Engdahl J, Svensson L, Young M,¨Angquist K-A,
Holmberg S. Decrease in the occurrence of ventricular fib-
rillation as the initially observed arrhythmia after out-of-
hospital cardiac arrest during 11 years in Sweden. Resusci-
22. Kuisma M, Repo J, Alaspaa A. The incidence of out-of-
hospital ventricular fibrillation in Helsinki, Finland, from
1994 to 1999. Lancet 2001;358(9280):473—4.