Editorial II
15 Maze M, Tranquilli W. Alpha-2 adrenergic agonists: de®ning
the role in clinical anesthesia. Anesthesiology 1991; 74:
581±605
16 Angel A, Majeed AB. Alterations of `sleeping time' in the rat
induced by drugs which modulate central monoaminergic
systems. Br J Anaesth 1990; 64: 594±600
17 Mason ST, Angel A. Brain noradrenaline and anaesthesia: further
characterization of the beta-receptor. Neuropharmacology 1983;
22: 1065±9
18 Smart D. Orexins: a new family of neuropeptides. Br J
Anaesth1999; 83: 695±7
19 Hagan JJ, Leslie RA, Patel S, et al. Orexin A activates locus
coeruleus cell ®ring and increases arousal in the rat. Proc Natl
Acad Sci USA 1999; 96: 10911±6
20 Lin L, Faraco J, Li R, et al. The sleep disorder canine narcolepsy is
caused by a mutation in the hypocretin (orexin) receptor 2 gene.
Cell 1999; 98: 365±76
21 Kubota K, Hirota K, Yoshida H, et al. Effects of sedatives on
Editorial II
Comparative mortality in anaesthesia
Ever since the advent of general anaesthesia in 1846, the
subject of anaesthetic mortality has been a source of
considerable debate and discussion. In the late 1880s, the
Hyderabad Commissions concluded that chloroform was a
completely safe anaesthetic agent in the human species.1
The Lancet commissioned its own study and this was
probably the ®rst epidemiological investigation of the
effects of anaesthetic agents in any species.2 On the basis
of animal experiments, the Hyderabad Commissions had
concluded that chloroform was a safe agent in man but the
Lancet Commission challenged these ®ndings. They con-
cluded that death under chloroform anaesthesia was 8.7
times more likely than death under ether anaesthesia,
whereas chloroform was only administered 6.1 times more
often than ether. Whilst the members of the commission
failed to establish the relative safety of ether or chloroform,
they focused attention on the role of human error in the
aetiology of anaesthetic deaths. It took the greater part of
50 yr before the real problems of anaesthetic deaths and
their prevention became the subject of intense study by a
number of anaesthetists throughout the world. It was only
during that time that the real value of properly maintained
anaesthetic records was emphasized by Waters.3 However,
it was not until 1954 that a well-designed and executed
survey was reported by Beecher and Todd.4 They collected
data on 600 000 patients from 10 university hospitals over a
5-yr period. The report caused considerable controversy, in
that it reported a considerably higher mortality in patients
that had received the newly introduced muscle relaxant
drugs compared with those that did not.
One of the major problems in surveys of anaesthetic
mortality is its actual de®nition and how this relates to
Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2001
noradrenaline release from the medial prefrontal cortex in rats.
GABAA receptor activator vs NMDA receptor inhibitor.
Psychopharmacology 1999; 146: 335±8
22 Kubota T, Anzawa N, Hirota K, Yoshida H, Matsuki A.
Effects of ketamine and pentobarbital on norepinephrine
release from the median frontal cortex in rats. Can J Anaesth
1999; 46: 388±92
23 Yoshida H, Kushikata T, Kubota T, Hirota K, Ishihara H, Matsuki
A. Increased norepinephrine release from anterior and posterior
hypothalamus following xenon inhalation. Can J Anesth 2001; 48:
651±5
24 Alam MN, McGinity D, Szymusiak R. Neuronal discharge of
preoptic/anterior hypothalamus thermosensitive neurons:
relation to NREM sleep. Am J Physiol 1995; 269: R1240±9
25 Kubota K, Hirota K, Anzawa N, Yoshida H, Takahashi S, Matsuki
A. Physostigmine antagonizes ketamine-induced noradrenaline
release from the medial prefrontal cortex in rats. Brain Res 1999;
840: 175±8
surgical and overall mortality; these problems have been
discussed by Duberman and Bendixen.5 With the re®nement
of both anaesthetic and surgical techniques, patients with
extensive pathology are being anaesthetized who in a
previous era would not have undergone surgery. Hence, it is
extremely dif®cult to compare anaesthetic mortality on a
historical basis.
In 1973, following a computer analysis of deaths
associated with anaesthesia, Marx and his colleagues
reported that the mortality within the ®rst 7 days of surgery
was 1.9%.6 They considered that two main factors deter-
mined the mortality rateÐthe physical status of the patient
and the skill and judgement of the physicians. Two-thirds of
the deaths were considered to be preventable. In a further
survey from Australia in 1975, Bodlander showed that the
incidence of deaths associated with anaesthesia was one in
502 (0.2%).7 The number of deaths attributable to anaes-
thesia showed a marked fall from 20% to 3.7% of the total
mortality over the 10-yr period of the survey. The de®nitive
work on anaesthetic mortality in the UK has been published
by Lunn and Mushin.8 They reported that one in 166 (0.6%)
of patients died within 6 days of surgery but only one in 10
000 deaths (0.01%) is directly attributable to anaesthesia.
They went on to report that, in a much larger number of
1800 deaths, anaesthesia may play some part in one in 1700
(0.06%), and they suggested that this could, in large
measure, be avoided. The events that caused these deaths
do not appear to have changed over the past 30 yr.
In a study of 7306 patients in Denmark by Pedersen and
colleagues in 1990, it was reported that one in 1800
(approximately 0.05%) died during anaesthesia, one in 730
(>0.1%) during the recovery period, and the overall hospital
 Editorial II
mortality rate was one in 81 (1.2%).9 In an assessment of
anaesthetic mortality, Keats suggested that `anesthesia
mortality has not decreased because we create new
mechanisms of mortality at the same rate as we solve
them'.10
In veterinary anaesthetic practice, it is extremely dif®cult
to collect meaningful statistics on the subject of anaesthetic
mortality, due to the diverse nature of veterinary practice
and the absence of a Coroner's Court system. Hence, there is
no real need or incentive to record or report anaesthetic
deaths. However, on the basis of scienti®c curiosity, a
number of surveys have been reported and, overall, there is a
greater interest in the subject over the past 10 yr.
Horses
Anaesthetic mortality in horses has always been considered
to be relatively high compared with most other species and
has been attributed to the size and bulk of the horse, and to its
unique cardiopulmonary physiology as a highly evolved
athletic animal. However, in a series of some 600 horses
which had received chloral hydrate by the i.v. route, Wright
and Hall reported only two deaths (0.3%).11 With the
development of extensive gastrointestinal surgery in the
horse, the mortality rate has risen and a rate of 1.18% has
been recorded more recently. 12 It is important to separate
the animals undergoing elective surgery from those
undergoing emergency abdominal surgery, which have
extensive abdominal pathology, and are extremely hypovo-
laemic and in endotoxic shock. Resuscitation of these
animals presents a massive challenge. In a series of horses
reported in 1983 from Norway, a mortality of 33 in 1216
(2.7%) was reported but anaesthesia was responsible for
only 0.8%.13 In a further study of 1314 equine general
anaesthetics reported from a single centre in 1993 by Young
and Taylor, the mortality was 0.68%.14 The main causes of
death were ischaemic myopathy (which is similar to
compartment syndrome), fracture of the cervical or long
bones, and cardiac arrest. It was not until 1995 that accepted
epidemiological principles were applied to anaesthetic
mortality in horses. A con®dential enquiry by Johnston
and colleagues surveyed a total of 6255 equine general
anaesthetics and reported an overall mortality of 1.6%.15
When abdominal surgery and foals were excluded, the rate
fell to 0.9%. They identi®ed a number of factors, which led
to a higher mortality, including anaesthesia in the third
trimester of pregnancy, emergency abdominal surgery,
orthopaedic procedures involving internal ®xation, and the
administration of xylazine. Xylazine is a relatively non-
speci®c alpha2-adrenoreceptor agonist, which was the ®rst
one of that group of drugs to be used in domestic animals. It
has similar cardiovascular effects to the other drugs in that
group, but the actual cause of death after xylazine is
uncertain. The risk of death also increased with the duration
of anaesthesia, when it was carried out outside normal
working hours, and in foals under 1 yr old. In a further single
814
centre study, Mee and colleagues reported on a series of
2276 horses undergoing elective surgery under general
anaesthesia.16 A death rate of 0.63% was attributable to
surgery/anaesthesia but the rate for anaesthesia was only
0.08%. The cause of death in the one horse was cardiac
arrest 15 min after induction of anaesthesia in which
resuscitation was unsuccessful. Post-mortem examination
failed to establish a cause of death. In contrast, in a further
series of 995 horses undergoing emergency procedures, the
same authors reported a mortality of 31.4%.17 In abdominal
surgery, the gross mortality was 35.5% and in the non-
abdominal surgery group it was 15.3%. The surgical/
anaesthesia death rate was 4.3% in the abdominal surgical
group and 2% in the other group. The main causes of death
were cardiac arrest, uncontrollable massive haemorrhage,
and irreversible endotoxic shock. In a later report in 2000,
Johnston and his colleagues reported on 41 787 equine
anaesthetics.18 The overall death rate was 1.9%. When
abdominal surgery was excluded, the rate was 0.9% but, for
abdominal surgery, it was 7.9%. The most common causes
of death were cardiac arrest (33%), fractures (23%), and
ischaemic myopathy (7%).
Small animals
In cats and dogs, there is a little more information available
from surveys conducted over the past 50 yr, although the
main information has been published in the past decade. A
survey of 2912 anaesthetics administered in one institution,
between 1955 and 1957, recorded a mortality of 1.1% in
dogs and 1.8% in cats.12 A follow-up survey in 1979±1981
showed a reduction to 0.43% in dogs and 0.25% in cats. The
main cause of death was attributed to human error leading to
anaesthetic overdosage and to hypoxia. Equipment prob-
lems, hypothermia, and cardiovascular collapse were other
causes of death. A survey of 10 000 feline anaesthetics in
Scotland reported a death rate of 0.3%.19 The main causes of
death appeared to be related to the failure to obtain accurate
body weights of the animals, to use adequate and effective
pre-medication, and to intubate the trachea.
In 1990, a survey of 20 814 dogs and 20 103 cats
undergoing general anaesthesia in the UK reported a
mortality of 0.23% in dogs and 0.29% in cats.20 However,
when they were divided into two groups, on the basis of the
absence or presence of pathology, a different situation was
observed. In dogs, the ®gure was 3.12% with pathology and
0.11% without. In cats, a similar pattern was observed with
a mortality of 3.33% with pathology and 0.29% without.
There were few factors that could be highlighted as the main
cause of death. However, the administration of xylazine was
again associated with an exceptionally high mortality rate.
Complications following tracheal intubation in the cats
appeared to be associated with several deaths.
In a survey involving some 30 000 anaesthetics reported
in 1992, a mortality rate of 0.11% was reported in dogs and
0.06% in cats.21 The most common cause of death was
 Editorial III
cardiopulmonary arrest. A further single institution survey
published in 1994 recorded a death rate of 0.43% in dogs
and cats.22 Most of the deaths were associated with cardiac
arrest and, due to the nature of the surgical condition,
resuscitation was not attempted in some of these animals. A
more recent survey was published in 1998 on anaesthesia in
the dog and cat.23 Death occurred in 0.11% of 8087 dogs
and in 0.1% of 8702 cats. Signi®cant odds ratios were
calculated for a number of factors including the association
of cardiac arrest with xylazine administration in dogs. A
modi®ed ASA classi®cation of patient status was adopted,
and a classi®cation of 3, 4, or 5 was also associated with an
increased incidence of cardiac arrest in both species.
It would appear that the incidence of death associated
with anaesthesia over a number of years has shown a
reduction in most species. Whilst this may well have been
associated with the development of `safer anaesthetic
techniques', it is most likely also to have been associated
with attempts to reduce human error by improved training
and wider use of sophisticated monitoring methods.
However, with the developments in surgery, which have
presented greater challenges to the anaesthetist, it remains to
be seen whether further major improvements in mortality
®gures will occur in the foreseeable future.
R. S. Jones
University Department of Anaesthesia
University Clinical Department
The Duncan Building
Daulby Street
Liverpool L69 3GA
UK
References
1 Bunker JP. Historical aspects. In: Lunn JN, ed. Epidemiology in
Anaesthesia. London: Edward Arnold, 1986; 1±7
2 Commission on Anaesthetics. Report of the Lancet Commission
appointed to investigate the subject of the administration of
chloroform and other anaesthetics from a clinical standpoint.
Lancet 1893; 1: 629±38, 693±708, 761±76, 889±914, 971±8,
1111±8, 1236±40, 1479±98
3 Waters RM. The evolution of anesthesia I and II. Proceedings of
the Staff Meetings of the Mayo Clinic 1942; 17: 428±40
Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2001
4 Beecher HK, Todd DP. A study of the deaths associated with
anesthesia and surgery based on a study of 599,348 anesthesias
in 10 institutions 1948±1952 inclusive. Ann Surg 1954; 140: 2±35
5 Duberman SM, Bendixen H. Mortality, morbidity and risk in
anaesthesia. In: Lunn JN, ed. Epidemiology in Anaesthesia. London:
Edward Arnold, 1986; 37±73
6 Marx GF, Mateu CV, Orkin LR. Computer analysis of
postanesthetic deaths. Anesthesiology 1973; 39: 54±9
7 Bodlander FMS. Deaths associated with anaesthesia. Br J
Anaesth1975; 47: 36±40
8 Lunn JN, Mushin WW. Mortality associated with anaesthesia.
Anaesthesia1982; 37: 856
9 Pedersen T, Eliasen K, Henriksen E. A prospective study of
mortality associated with anaesthesia and surgery: risk indicators
of mortality in hospital. Acta Anaesthesiol Scand 1990; 34: 176±82
10 Keats AS. Anesthesia mortality in perspective anesthesia and
analgesia. Anesth Analg 1990; 71: 113±9
11 Wright JG, Hall LW. Veterinary Anaesthesia and Analgesia, 5th
Edn. London: BaillieÁre Tindall and Cox, 1961; 161
12 Lumb WV, Jones EW. Veterinary Anaesthesia, 2nd Edn.
Philadelphia: Lea and Febiger, 1973; 611±29
13 Tevik A. The role of anaesthesia in surgical mortality in horses.
Nord Vet Med 1983; 35: 175±9
14 Young SS, Taylor PM. Factors in¯uencing the outcome of equine
anaesthesia: a review of 1,314 cases. Equine Vet J 1993; 25: 147±51
15 Johnston GM, Taylor PM, Holmes MA, Wood JLN. Con®dential
enquiry of perioperative equine fatalities (CPEF-1): preliminary
results. Equine Vet J 1995; 27: 193±200
16 Mee AM, Cripps PJ, Jones RS. A restrospective study of
mortality associated with general anaesthesia in horses: elective
procedures. Vet Rec 1998; 142: 275±6
17 Mee AM, Cripps PJ, Jones RS. A restrospective study of
mortality associated with general anaesthesia in horses:
emergency procedures. Vet Rec 1998; 142: 307±9
18 Johnston GM. Equine anaesthesia ± a chance to cut is a chance
to kill. Proc Assoc Vet Anaesth 2000; 1±2
19 Dodman NH. Feline anaesthetic survey. J Small Anim Pract 1977;
10: 653±8
20 Clarke KC, Hall LW. A survey of anaesthesia in small animal
practice. AVA/BSAVA report. J Ass Vet Anaesth 1990; 17: 4±10
21 Dodman NH, Lamb LA. Survey of small animal anesthetic
practice in Vermont. J Anim Hosp Ass 1992; 28: 439±45
22 Gaynor JS, Dunlop CI, Wagner AE, Wertz EM, Golden AE,
Demme WC. Complications and mortality associated with
anesthesia in dogs and cats. J Anim Hosp Ass 1994; 35: 13±7
23 Dyson DH, Maxi MG. Morbidity and mortality associated with
anesthetic management in small animal veterinary practice in
Ontario. J Anim Hosp Ass 1998; 35: 325±35