Acta Anaesthesiol Scand 2005; 49: 702—706 Copyright # Acta Anaesthesiol Scand 2005

Printed in UK. All rights reserved

ACTA ANAESTHESIOLOGICA SCANDINAVICA
doi: 10.1111/j.1399-6576.2005.00679.x

Observations and warning signs prior to cardiac arrest.

Should a medical emergency team intervene earlier?

J. NURMI1, V. P. HARJOLA2, J. NOLAN3 and M. CASTRÉN1

1
Department of Anesthesiology and Intensive Care Medicine, Uusimaa Emergency Medical Services, Helsinki University Hospital, Finland,
2
Department of Medicine, Division of Emergency Medicine, Helsinki University Central Hospital, Finland, and 3Department of Anesthesia,
Royal United Hospital, Bath, UK

Background: The Medical Emergency Team (MET) has
evolved in some hospitals as a means of delivering effective
treatment early enough to prevent cardiac arrests. Our aim was
to analyze the effectiveness of observation practice to detect
abnormalities in vital signs prior to cardiac arrest and to deter-
mine the need for a MET system in Finnish hospitals.
Methods: The charts of patients who suffered cardiac arrest
during 18 months in four hospitals were reviewed. The vital
signs, symptoms and interventions during 8 h prior to arrest
were recorded and analyzed against trigger criteria of the MET.
Results: During the study period, 110 patients suffered car-
diac arrest in hospitals, and 56 (51%) of the arrests occurred on
the wards. Of those patients, 30 (54%) had an abnormal vital
sign fulfilling the MET criteria, documented on average 3.8 h
prior to the arrest. During this period, 13 patients did not
receive any intervention (e.g. supplemental oxygen or medica-
tion), eight received intervention within 1 h and nine received
intervention after more than 1 h. Response to the first interven-
tion was not attained in any patient; nevertheless re-interven-
tions took place in one patient only.
Conclusion: Significant physiological deterioration seems to
be common in the hours before a cardiac arrest on the wards of
Finnish hospitals, suggesting that implementation of a MET-
system may be worthwhile. However, the practice of vital sign
observation by the nursing staff should be improved before
maximal benefit of a MET can be achieved.
Accepted for publication 20 December 2004
Key words: Cardiopulmonary resuscitation; critical care;
nursing staff, hospital; respiratory insufficiency.
# Acta Anaesthesiologica Scandinavica 49 (2005)

P REVIOUS studies have shown that survival to hos-

pital discharge after in-hospital cardiac arrest is
approximately 17—24% (1, 2). Survival rates after car-
diac arrest on hospital general wards are significantly
lower than in those occurring in monitored areas
(1—3). Up to 80% of these patients will show signs
of significant physiological deterioration in the hours
before their cardiac arrest (4—7). Despite documenta-
tion of these warning signs, interventions are fre-
quently inappropriate or not undertaken at all (8, 9).
The Medical Emergency Team (MET) concept has
evolved as a means of delivering effective treatment
early and, hopefully, preventing deterioration to
cardiac arrest (10—13). A few studies have con-
cluded that implementation of a MET reduces the
incidence of and mortality from unexpected cardiac
arrest, but all these studies have involved historical
control groups (13, 14). The previous studies of
antecedents of cardiac arrest and interventions
have been undertaken in British, Australian and
North American hospitals, where the hospital cul-
ture and practices on wards may differ from many
other countries. The current study was undertaken in
several Finnish hospitals to determine the prevalence
of abnormal vital signs prior to cardiac arrest and to
estimate the sufficiency of current practice in patient
observation on the wards to screen those patients at
risk of cardiac arrest.
Methods
We reviewed the charts of patients who suffered a
cardiac arrest during 18 months (December 2001 to
May 2003) in four Finnish hospitals (i.e. a tertiary
teaching hospital, a tertiary trauma center and two
secondary general hospitals). In the secondary hos-
pitals and the tertiary trauma center, the resuscitation
attempts in the hospital are managed by an estab-
lished resuscitation team, and a resuscitation officer

702

continuously collects records of resuscitations. In these
hospitals, the patients included in the study were
identified from the resuscitation sheets. In the tertiary
teaching hospital, no centralized collection of resusci-
tation sheets exists and the patients for the study were
identified from the cardiopulmonary resuscitation
related diagnoses entered in the hospital’s database.
We analyzed retrospectively the documented symp-
toms and vital signs, and any interventions under-
taken during the 8 h before the cardiac arrest. We
also recorded the number of observations during the
24 h before the cardiac arrest. The symptoms and vital
signs were analyzed against the MET trigger criteria
used by Buist et al. (Table 1; 13). Ninety-five percent
confidence intervals (95% CI) were calculated for pro-
portions using a modified Wald method.
Results
The total number of patients was 110 (64 males and 46
females). The mean age was 68 years (standard devi-
ation [SD] 16, range 8—97), and length of stay before
the cardiac arrest was 4.4 days (SD 4.3, range 1—21).
The reasons for hospitalization of the patients who
suffered a cardiac arrest on the wards are shown in
Table 2. Thirty-three (30%, 95% CI 22—39) patients
presented with a shockable rhythm and 39 (35%,
95% CI 27—45) achieved spontaneous circulation. The
number of resuscitation attempts in the tertiary teach-
ing hospital, tertiary trauma center and two secondary
general hospitals was 15, 17, 20 and 58, respectively.
Table 1
Medical Emergency Team calling criteria,
defined by Buist et al. (13).
Airway
* Respiratory distress
* Threatened airway
Breathing

1
* Respiratory rate >30 min
* Respiratory rate <6 min

1
* SaO2<90% on oxygen
* Difficulty speaking
Circulation
* Blood pressure <90 mmHg despite treatment
* Pulse rate >130 min

1
Neurology
* Any unexplained decrease in consciousness
* Agitation or delirium
* Repeated or prolonged seizures
Other
* Concern about patient
* Uncontrolled pain
* Failure to respond to treatment
* Unable to obtain prompt assistance
Pre-arrest observations and signs
Fifty-four (49%, 95% CI 40—58) of the cardiac arrests
occurred in the intensive and coronary care units, the
operating room or in the emergency department. Of
those cases, 12 (22%, 95% CI 13—35) fulfilled at least
one of the MET criteria. All those 12 patients received
intensive treatment immediately after recognition of
an abnormal vital sign and re-interventions took place
without delay. The remainder of the cardiac arrests
[56 (51%, 95% CI 42—60)] occurred on a ward and of
those patients, 30 (54%, 95% CI 41—66) fulfilled at
least one of the MET criteria. The mean delay from
the first documented abnormal vital sign to the car-
diac arrest was 3.8 h (SD 2.8, range 0.5—8.0). The most
common MET criteria displayed were respiratory dis-
tress (17 cases), SpO2 < 90% on oxygen (10 cases) and
a decreased level of consciousness (10 cases). Inter-
ventions made for the patients with abnormal vital
signs on the wards and the frequency of observations
of the ward patients who did not fulfill the MET
criteria are shown in a flow chart (Fig. 1).
The frequency of documented observations in the
24 h before the cardiac arrest varied considerably
among the patients on the wards (Fig. 2). Only pulse
rate and blood pressure were recorded more than
once during these 24 h (mean 1.6 and 1.4 times per
24 h, respectively).
Discussion
Our study has shown that the cardiac arrest patients on
the wards of the four Finnish hospitals of different types
showed evidence of deterioration long before their car-
diac arrest. Fifty-four percent of the ward patients suf-
fering a cardiac arrest had abnormal vital signs in the
preceding 24 h. We have also shown that interventions
during this period are usually insufficient and are per-
formed too late. Our results are consistent with those
reported by authors from other countries (4, 6, 8, 9).
In our study, the most common warning signs
before cardiac arrest on the wards were respiratory
distress, decreased oxygen saturation and a decreased
level of consciousness. The respiratory rate was docu-
mented in one patient only. Tachypnoea is one of the
most important predictors of cardiopulmonary arrest
and yet in our study hospitals respiratory rate was
virtually never recorded (7). In one study from the
UK, respiratory rate was recorded in the 24 h before
the cardiac arrest in only 27% of the patients (8). This
is not surprising given the fact that only 20% of text-
books dealing with clinical examinations mention
the important role of respiratory rate in critical illness
(15). Nurses and medical staff must be taught the
703
J. Nurmi et al.

Table 2
Causes of hospitalization of patients who suffered a cardiac arrest on the wards.
Patients with documented abnormal vital Patients without documented abnormal
signs prior to cardiac arrest (n ¼ 30) vital signs prior to cardiac arrest (n ¼ 26)
Cardiovascular diseases
Ischemic cardiac disease 3 5
Arrhythmias 2 1
Pulmonary embolism 1 1
Congestive heart disease 0 2
Abdominal aortic aneurysm (postoperative) 1 1
Mitral valve replacement (postoperative) 1 0
Critical peripheral ischemia 1 1
Cardiomyopathy 0 1
Pulmonary diseases
COPD 1 0
Pleural effusion 1 0
Orthopaedic problems
Hip fracture (postoperative) 9 2
Other fractures 1 1
Neurological diseases 1 3
Infection diseases
Pneumonia 2 0
Erysipelas 1 1
Gluteal abscess 0 1
Other diseases
Uremia 2 0
Anuria 1 0
Anemia 0 2
Icterus 0 1
Ileus 1 0
Arthrosis 0 1
Lymphoma 0 1
Myalgia 1 0
Nasal bleeding 0 1

importance of the respiratory rate and to record this
observation. This is emphasised in the recently
described ALERTTM(ALERTTM, Portsmouth, UK)
course, which is targeted at nurses and junior doctors
in the UK (16). In our view, there is a role for a similar
course in Finland as well.
Of the patients whose cardiac arrest occurred in the
intensive and coronary care units, the operating room
or in the emergency department, only 12 of 54 patients
had documented abnormal vital signs prior to their
cardiac arrest, even though they were monitored more
intensively. Moreover, all the patients with abnormal
vital signs in those areas received interventions with-
out unnecessary delays. Thus, the major benefit of
MET could be observed within patients deteriorating
in the unmonitored wards.
Of the patients who suffered a cardiac arrest on the
wards, 46% did not have any documented abnormal
vital sign. This proportion includes the patients whose
cardiac arrest occurred suddenly and those whose vital
signs were not measured or documented. The preva-
lence of abnormal vital signs prior to cardiac arrest in
our study is comparable to previous ones (4—7).
The introduction of a MET system should enable
patients at risk of cardiac arrest to be treated
promptly and appropriately and might reduce the
incidence of cardiac arrest in hospital wards (12). In
the current study, we did not assess the potential
benefit that could be provided to the patients after
earlier recognition of abnormal vital signs, and thus
we do not know whether some of the cardiac arrests
could have been avoided. However, the causes of
hospitalization of the patients who suffered their
cardiac arrest in the ward were mainly diseases not
expected to be necessarily fatally. Previous studies
(13, 14) have demonstrated a decreased rate of
cardiac arrests and mortality after implementation
of MET. However, no randomized, controlled trials
testing the benefit achieved by early interventions
provided by the MET have been published.
Our study has several weaknesses. Due to the retro-
spective nature of the study, we may have omitted
some patients who sustained a cardiac arrest. In three
out of four hospitals in the study, resuscitation sheets
were routinely collected, and virtually all patients
resuscitated were included. In contrast, in the tertiary

704
 Pre-arrest observations and signs

Patients suffering cardiac representative of all Finnish hospitals and collectively

arrest on the wards
56 they yielded just 110 cardiac arrests during 18 months.
In conclusion, 54% of the ward patients who suf-
fered a cardiac arrest had had abnormal vital signs
At least one MET trigger None of MET trigger criteria
documented in the preceding 24 h. During this period,
criteria fulfilled
30
fulfilled
26
the interventions were either insufficient or per-
formed too late. These deficiencies should be
addressed by appropriate education in identifying
No observations in 8 h At least 1 observation in 8 h
7 19 patients at risk of cardiac arrest and in effective inter-
blood pressure (12)
heart rate (10) ventions for patients with abnormal vital signs. Intro-
pain (5)
oxygen saturation (3) duction of a MET could well support the achievement
counsciousness (3)
work of breathing (2) of this goal. However, the current practices of observ-
respiration rate (1)
ing vital signs on the wards need to be improved to
achieve the maximal benefit of MET.
Intervention in > 1h Intervention in < 1h No intervention prior to
9 8 cardiac arrest
13

Acknowledgements
No response to first No response to first No response to first
We wish to thank Dr Mauri Pehko and Dr Tero Varpula for their
intervention, no new intervention, no new intervention, new contribution to the study. This study was supported by The
intervention intervention intervention in > 1h
9 7 1 Laerdal Foundation for Acute Medicine.

Fig. 1. Interventions for the ward patients with an abnormal vital

sign, and the number of recorded vital signs in those patients who
did not fulfill any medical emergency team criteria. Data for 8 h References
before the cardiac arrest.
1. Gwinnutt CL, Columb M, Harris R. Outcome after cardiac
arrest in adults in UK hospitals: effect of the 1997 guidelines.
teaching hospital, in which the patients were Resuscitation 2000; 47: 125—35.
collected by diagnosis codes in the hospital’s data- 2. Skogvoll E, Isern E, Sangolt GK, Gisvold SE. In-hospital
base, only 15 patients were found. Obviously, this cardiopulmonary resuscitation. 5 years’ incidence and sur-
represents only a minority of the resuscitation vival according to the Utstein template. Acta Anaesthesiol
Scand 1999; 43: 177—84.
attempts. The four hospitals in our study may not be 3. Herlitz J, Bang A, Aune S et al. Characteristics and outcome
among patients suffering in-hospital cardiac arrest in
monitored and non-monitored areas. Resuscitation 2001;
10 48: 125—35.
4. Schein RM, Hazday N, Pena M, Ruben BH, Sprung CL.
Number of observations

Clinical antecedents to in-hospital cardiopulmonary arrest.

8 Chest 1990; 98: 1388—92.
5. Smith AF, Wood J. Can some in-hospital cardio-respiratory
6 arrests be prevented? A prospective survey. Resuscitation
1998; 37: 133—7.
6. Hodgetts T, Kenward G, Vlachonikolis I, Payne S, Castle N.
4 The identification of risk factors for cardiac arrest and for-
mulation of activation criteria to alert a medical emergency
2 team. Resuscitation 2002; 54: 125.
7. Fieselmann JF, Hendryx MS, Helms CM, Wakefield DS.
Respiratory rate predicts cardiopulmonary arrest for
0 internal medicine inpatients. J Gen Intern Med 1993; 8:
354—60.
8. Hodgetts T, Kenward G, Vlackonikolis I et al. Incidence,
te

te

re

ss
2

location and reasons for avoidable in-hospital cardiac

O
ra

ra

su

ne
Sp

arrest in a district general hospital. Resuscitation 2002;

es
ry

rt

us
ea
to

pr

io

54: 115.
ra

sc
d
pi

9. Franklin C, Mathew J. Developing strategies to prevent

oo

on
es

Bl

inhospital cardiac arrest: analyzing responses of physicians

R

and nurses in the hours before the event. Crit Care Med 1994;
Fig. 2. Number of documented observations of vital signs in the 22: 244—7.
wards during the 24 h before the cardiac arrest. Each circle 10. Lee A, Bishop G, Hillman KM, Daffurn K. The Medical
represents a single patient and the lines represent means. Emergency Team. Anaesth Intensive Care 1995; 23: 183—6.

705
J. Nurmi et al.
11. Parr MJ, Hadfield JH, Flabouris A, Bishop G, Hillman K. The
Medical Emergency Team. 12 month analysis of reasons for
activation, immediate outcome and not-for-resuscitation
orders. Resuscitation 2001; 50: 39—44.
12. Cretikos M, Hillman K. The medical emergency team: does it
really make a difference? Intern Med J 2003; 33: 511—4.
13. Buist MD, Moore GE, Bernard SA et al. Effects of a medical
emergency team on reduction of incidence of and mortality
from unexpected cardiac arrests in hospital: preliminary
study. BMJ 2002; 324: 387—90.
14. Bellomo R, Goldsmith D, Uchino S et al. A prospective
before-and-after trial of a medical emergency team. Med J
Aust 2003; 179: 283—7.
706
15. Cook CJ, Smith GB. Do textbooks of clinical examination
contain information regarding the assessment of critically
ill patients? Resuscitation 2004; 60: 129—36.
16. Smith GB, Osgood VM, Crane S. ALERT — a multiprofes-
sional training course in the care of the acutely ill adult
patient. Resuscitation 2002; 52: 281—6.
Address:
J. Nurmi
Eteläinen Hesperiankatu
22 A 26
FI-00100 Helsinki
Finland
e-mail: address: jouni.nurmi@helsinki.fi