The anaesthetized patient is at risk of complications resulting from the actions, or inaction, of the anaesthetist, from the actions of the surgeon, and from failure or malfunction of anaesthetic equipment. The state of anaesthesia may be considered to be intrinsically unsafe. Patients are subjected to administration of drugs which have side-effects, particularly on the cardiovascular and respiratory systems. Unconsciousness carries with it risks of airway obstruction, soiling of the lungs, and inability to detect peripheral injury. Pharmacological muscle paralysis necessitates the use of artificial ventilation, making the patient dependent on the anaesthetist and his equipment for the fundamental functions of oxygenation and excretion of carbon dioxide. The anaesthetist may deliberately alter physiological functions, for example by inducing hypotension or ventilating only one lung.

Estimates of mortality

Mortality is a vital estimate of risk associated with anaesthesia, the most important reason being that the definition is clear, in contrast to the more debatable definitions of morbidity. Mortality76 is also a somewhat crude estimate of risk, because of the relative rarity of this complication.

During the three decades up to 1980, a number of investigators in various countries had attempted to estimate the frequency with which death was associated with anaesthesia (Table 1).710213637444748546569707887 There was a general trend towards reduced mortality attributable primarily to anaesthesia during this period (from about 1:2500 to about 1:5000), but the same principal causes of death continued to be identified: inadequate supervision of trainees, drug overdose, drug mistakes, airway obstruction, aspiration of gastric contents, insufficient monitoring, and lack of postoperative care.

Table 1

Estimates of the incidence of mortality due to anaesthesia between 1954 and 1980

Authors
 
Year of publication
 
Number of anaesthetics
 
Primary cause
 
Primary and associated cause
 
Beecher and Todd7 1954 599 548 1:2680 1:1560 
Dornette and Orth36 1956 63 105 1:2427 1:1343 
Schapira et al.87 1960 22 177 1:1232 1:821 
Phillips et al.78 1960 – 1:7692 1:2500 
Dripps et al.37 1961 33 224 1:852 1:415 
Clifton and Hotten21 1963 295 640 1:6048 1:3955 
Memery69 1965 114 866 1:3145 1:1082 
Gebbie44 1966 129 336 – 1:6158 
Minuck70 1967 121 786 1:6766 1:3291 
Harrison47 1968 177 928 – 1:3068 
Marx et al.65 1973 34 145 – 1:1265 
Bodlander10 1975 211 130 1:14 075 1:1703 
Harrison48 1978 240 483 – 1:4537 
Hovi-Viander54 1980 338 934 1:5059 1:1412 
Authors
 
Year of publication
 
Number of anaesthetics
 
Primary cause
 
Primary and associated cause
 
Beecher and Todd7 1954 599 548 1:2680 1:1560 
Dornette and Orth36 1956 63 105 1:2427 1:1343 
Schapira et al.87 1960 22 177 1:1232 1:821 
Phillips et al.78 1960 – 1:7692 1:2500 
Dripps et al.37 1961 33 224 1:852 1:415 
Clifton and Hotten21 1963 295 640 1:6048 1:3955 
Memery69 1965 114 866 1:3145 1:1082 
Gebbie44 1966 129 336 – 1:6158 
Minuck70 1967 121 786 1:6766 1:3291 
Harrison47 1968 177 928 – 1:3068 
Marx et al.65 1973 34 145 – 1:1265 
Bodlander10 1975 211 130 1:14 075 1:1703 
Harrison48 1978 240 483 – 1:4537 
Hovi-Viander54 1980 338 934 1:5059 1:1412 

One of the problems which renders comparison between these studies difficult is that different criteria were used to define anaesthetic death. A spectrum of time limits has been used, starting with all deaths occurring before the time of transfer of the patient from the operating theatre or from the recovery room. A limit of deaths occurring within 24–48 h after anaesthesia has reflected coronial requirements in many jurisdictions, with a period of 7–10 days used in other studies. However, some patients who suffer anaesthetic-related complications may not die for weeks, months or even years after the anaesthetic. These deaths would not be captured in studies using such limits. In addition, some studies of anaesthetic mortality have included patients who suffered hypoxic cerebral damage, with resultant persistent coma.

More recent studies of mortality are shown in Table 2. In 1982, the Association of Anaesthetists of Great Britain and Ireland (AAGBI) published the results of a major study of mortality in five regions in the UK. An anonymous and confidential system was established to report deaths which occurred within 6 days of surgery. During the study period, an estimated 1 147 362 operations took place.63 The overall perioperative mortality was 0.53%. Anaesthesia was considered to have been totally responsible for death in less than 1:10 000 operations, but may have contributed to death in 1:1700 operations.

Table 2

Estimates of the incidence of mortality due to anaesthesia between 1982 and 2003

Authors
 
Year of publication
 
Number of anaesthetics
 
Primary cause
 
Primary and associated cause
 
Lunn and Mushin63 1982 1 147 362 1:10 000 1:700 
Tiret et al.98 1986 198 103 1:13 207 – 
Buck et al.13 1987 555 258 1:185 086 1:1354 
Holland50 1987 – 1:26 000 – 
Chopra et al.19 1990 113 074 1:16 250 – 
Pedersen77 1994 200 000 1:2500 – 
Tikkanen and Hovi-Viander96 1995 325 585 1:66 667 – 
Warden and Horan100 1996 – – 1:20 000 
Arbous et al.3 2001 869 483 1:124 212 1:7143 
Willis et al.103 2002 10 336 000 1:220 000 1:79 509 
Kawashima et al.55 2003 2 363 038 1:47 619 – 
Authors
 
Year of publication
 
Number of anaesthetics
 
Primary cause
 
Primary and associated cause
 
Lunn and Mushin63 1982 1 147 362 1:10 000 1:700 
Tiret et al.98 1986 198 103 1:13 207 – 
Buck et al.13 1987 555 258 1:185 086 1:1354 
Holland50 1987 – 1:26 000 – 
Chopra et al.19 1990 113 074 1:16 250 – 
Pedersen77 1994 200 000 1:2500 – 
Tikkanen and Hovi-Viander96 1995 325 585 1:66 667 – 
Warden and Horan100 1996 – – 1:20 000 
Arbous et al.3 2001 869 483 1:124 212 1:7143 
Willis et al.103 2002 10 336 000 1:220 000 1:79 509 
Kawashima et al.55 2003 2 363 038 1:47 619 – 

Because of the importance of the findings, and because of the difficulty in separating anaesthetic and surgical factors when reports came only from anaesthetists, the AAGBI initiated the first Confidential Enquiry into Perioperative Deaths (CEPOD) in conjunction with the Association of Surgeons of Great Britain and Ireland. Three regions in the UK were studied over a 12-month period. The overall perioperative mortality was 0.7%.13 There were 410 deaths associated with anaesthesia out of a total of 2928 deaths after 555 258 anaesthetics. Factors which were believed to have contributed to death are shown in Table 3. However, expert assessors considered that only three deaths resulted solely from anaesthesia, an incidence of 1 in 185 086 anaesthetics.

Table 3

Factors involved in deaths attributable in part to anaesthesia, in decreasing order of frequency.13

Including failure of a trainee to consult a more senior anaesthetist, grossly inadequate monitoring, inappropriate drug doses or other clear indication of a poor standard of practice

Failure to apply knowledge 
Lack of care 
Failure of organization 
Lack of experience 
Lack of knowledge 
Drug effect 
Failure of equipment 
Fatigue 
Failure to apply knowledge 
Lack of care 
Failure of organization 
Lack of experience 
Lack of knowledge 
Drug effect 
Failure of equipment 
Fatigue 

Studies from other countries have suggested higher rates of death related to anaesthesia than that reported in the CEPOD study. From 1978 to 1982, the French Health Ministry conducted a prospective nationwide survey of major complications during anaesthesia. A representative sample of 198 103 anaesthetics was analysed from 460 institutions selected at random; this represented approximately 8% of the total estimated number of anaesthetics undertaken in France.98 During anaesthesia, or within 24 h, 268 major anaesthesia-related complications occurred (one in every 739 anaesthetics). There were 67 deaths within 24 h, and 16 patients suffered coma which persisted after 24 h. The incidence of death and persistent coma attributable totally to anaesthesia was 1:7924; death due solely to anaesthesia occurred with an incidence of 1:13 207. In half of all the patients who died or suffered coma, postoperative respiratory depression was responsible.

In New South Wales, Australia, a system was put in place in 1960 to undertake a confidential investigation of deaths related to anaesthesia. Deaths were categorized as anaesthetic, surgical, inevitable, fortuitous or unassessable. Between 1960 and 1985, the incidence of death attributable to anaesthesia decreased by a factor of five, from 1:5500 to 1:26 000.50 However, the pattern of errors by anaesthetists remained largely unchanged during the 25-yr period; inadequate preparation of patients, wrong choice of agent or technique, inadvertent overdose, inadequate crisis management and inadequate resuscitation remained the commonest errors. The proportion of deaths attributable to anaesthesia in which no error could be found in management increased from 2.8% in the period 1960–69 to 10% in the period 1983–85. Over the same period, the proportion of specialist anaesthetists involved in deaths attributable to anaesthesia increased from 27 to 62%.

In the Netherlands, a retrospective study of faults, accidents, near-accidents and complications associated with anaesthesia in one institution was conducted between 1978 and 1987.20 During that period, 97 496 anaesthetics were administered for non-cardiac procedures. Cardiac arrest occurred with an incidence of 1:3362 anaesthetics and mortality from cardiac arrest in these patients had an incidence of 1:5417 anaesthetics. Anaesthesia was considered to have contributed to cardiac arrest in 1:7500 anaesthetics, with a fatal outcome in 1:16 250 anaesthetics. Failure to check, lack of vigilance and carelessness were the most frequently associated human factors.

In a prospective study conducted in Denmark,77 mortality attributable to anaesthesia occurred with a frequency of 1:2500 (0.04%). The overall perioperative mortality rate was 1.2%, and 0.05% of patients died during anaesthesia. Mortality in patients who developed postoperative cardiovascular complications was 20%.

Mortality associated with anaesthesia and surgery was studied in Finland in 1986 (although not published until 1995).96 Death was caused primarily by anaesthesia, with an incidence of approximately 1:67 000. This was a substantial improvement on the mortality rate of 1:5000 reported after an earlier study in 1975.54

Between 1984 and 1990, there were 1503 deaths within 24 h of, or as a result of, anaesthesia in New South Wales. The Special Committee Investigating Deaths Under Anaesthesia in New South Wales attributed 172 deaths (11.4%) to factors under anaesthetists' control. About 10% of the deaths occurred in patients undergoing urgent non-emergency operations, of which 45 (31.3%) were attributed to anaesthetic factors.52,100 General anaesthesia was used in 31 cases and major regional block (10 spinals and four epidurals) in the others. Anaesthetic factors most often identified as contributing to death were inadequate preparation for anaesthesia and surgery, inappropriate choice or application of technique, inadequate postoperative care, and overdose. Patients most likely to suffer an anaesthetic-related death were those who were elderly (between 70 and 79 yr) and male (almost twice as often as females). The calculated rate of death attributable wholly or in part to anaesthesia was 1 in 20 000 operations.

In Canada, the risk factors associated with death within 7 days of anaesthesia were analysed in a study involving 100000 surgical procedures.25 There were 71 deaths per 10000 patients, and differences in anaesthetic practice were of much less importance in contributing to death than were age, physical status and the type of surgery.

Another study from the Netherlands3 found that the incidence of mortality in which anaesthesia contributed to death was 1:7143 procedures, but that it was solely responsible for death in only 1:124 000 operations. A number of inadequacies in management of the patients were identified, including poor preoperative assessment, inappropriate anaesthetic technique, inadequate cardiovascular management, poor management of ventilation and inadequate monitoring. Inadequate communication and lack of supervision were also noted. The highest frequency of anaesthesia-related deaths (45.4%) occurred among ASA III patients, whereas mortality rate associated with all operations increased with increasing ASA grade; 39.5% of all deaths occurred in ASA V patients, in comparison with only 10.1% of all anaesthesia-related deaths.

All states in Australia have government-supported special committees to collect data about anaesthetic-related deaths. Four successive triennial reports have been collated from all State committees. The report for 1994–199632 concluded that anaesthetic-related deaths occurred with a frequency of no more than 1:63 000 operative or diagnostic procedures; the death rate attributable to anaesthesia alone was considered to be about 1 in over 150 000 procedures. The latest report available at the time of writing103 found that the incidence of death had decreased during the triennium 1997–1999; anaesthesia caused or contributed to death in 1:79 500 procedures and was the definite cause in approximately 1:220 000. In most deaths there was more than one causal or contributory factor (mean 2.5). Correctable anaesthetic factors identified most frequently were related to anaesthetic technique, inadequate preoperative assessment and use of anaesthetic drugs. The authors noted that 73% of deaths involved specialist anaesthetists, but 27% involved non-specialist anaesthetists or trainees working unsupervised. The proportion of ASA I or II patients among those who died decreased from 42% in 1985–1987 to 10% in 1994–1996, although it increased to 15% in 1997–1999.

A study in Japan55 investigated mortality between 1994 and 1998. The findings may not be as reliable as those from other studies because data were gathered retrospectively from questionnaires, only 40% of which were returned. Nevertheless, data were gathered relating to 2.3 million operations. The mortality rate attributable to anaesthesia was approximately 1:48 000.

In the USA, an incidence of 1.7 cardiac arrests per 10 000 anaesthetics was reported in 1985,56 although not all were fatal. The study involved 163 240 anaesthetics administered over a 15-yr period. Of 449 cardiac arrests, 27 were judged to be attributable solely to anaesthesia, and mortality was 0.9 per 10 000 anaesthetics. Three-quarters of these cardiac arrests were considered to have been preventable. In 1991, the same authors published results relating to 241934 anaesthetics over a 20-yr period from 1969 to 1988.57 During the second decade, pulse oximetry and capnography were introduced. Cardiac arrest related to anaesthetic causes decreased from 2.1 per 10 000 anaesthetics in the first decade to 1.0 per 10 000 in the second decade. Most of this difference was due to a decrease in cardiac arrests from preventable respiratory causes.

Kawashima et al.55 also investigated the incidence of cardiac arrest during anaesthesia. They reported that cardiac arrest attributable to anaesthesia (from 1994–1998) occurred with an incidence of 1.0 per 10 000, a figure identical to that reported by Keenan and Boyan57 for the period from 1979 to 1988. However, mortality following cardiac arrest was 0.13 per 10 000 anaesthetics, much lower than that reported from the USA in the previous decade. The 10 most common causes of cardiac arrest attributable to anaesthesia alone are shown in Table 4.

Table 4

The 10 most common causes of intra-operative cardiac arrest attributable to anaesthesia55

Cause
 
Frequency (%)
 
Drug overdose or selection error 15.3 
Serious arrhythmia 13.9 
Myocardial infarction, ischaemia 8.8 
Inadequate airway management 7.9 
High spinal 7.4 
Inadequate vigilance 6.9 
Massive haemorrhage managed badly 5.1 
Overdose of inhaled anaesthetic 2.8 
Suffocation, aspiration 2.8 
Dis/misconnection 2.3 
Cause
 
Frequency (%)
 
Drug overdose or selection error 15.3 
Serious arrhythmia 13.9 
Myocardial infarction, ischaemia 8.8 
Inadequate airway management 7.9 
High spinal 7.4 
Inadequate vigilance 6.9 
Massive haemorrhage managed badly 5.1 
Overdose of inhaled anaesthetic 2.8 
Suffocation, aspiration 2.8 
Dis/misconnection 2.3 

Despite the difficulty in comparing the results of these studies directly, there does appear to have been a considerable reduction in mortality caused primarily by anaesthesia within the last 25 yr in countries with high standards of training and well-equipped hospitals. The figures in Table 2 should be compared with the mortality rate reported by McKenzie.67 Death or persistent coma attributable to anaesthesia occurred with a frequency of 1:388 in Zimbabwean teaching hospitals in 1992, and there were avoidable factors in 51% of deaths.

Estimates of morbidity

Reference has been made above to some studies which have estimated the frequency of serious morbidity as well as mortality. Pedersen77 reported a very high incidence (9%) of intraoperative cardiopulmonary complications associated with anaesthesia or surgery, and requiring intervention during the procedure. There was a trend suggesting that regional anaesthesia was safer in elderly patients with chronic obstructive airways disease who underwent major orthopaedic surgery. One-third of all complications were considered to be preventable.

Cooper et al.27 studied patients admitted to an intensive care unit (ICU) as a result of serious complications of anaesthesia. Two per cent of ICU admissions were related to complications of anaesthesia (1 in 1543 anaesthetics). Most complications (62%) occurred in the recovery period. Most complications involved the respiratory or cardiovascular systems. One quarter of the complications were judged to have been avoidable, and 17% of the patients died.

The Multicenter Study in the USA4243 was conducted in an attempt to analyse predictors of severe perioperative adverse outcome related to general anaesthesia using four specific anaesthetic agents. A total of 17 201 patients were followed up for 7 days for the occurrence of 14 specified outcomes: hypotension, hypertension, tachycardia, ventricular arrhythmia, bradycardia, atrial arrhythmia, cardiac failure, myocardial ischaemia, myocardial infarction, bronchospasm, respiratory failure, secretions, and ‘any severe myocardial outcome’ or ‘any severe respiratory outcome’. The major risk factors for severe outcome were: cardiovascular surgery, thoracic surgery or abdominal surgery; history of cardiac failure, myocardial infarction, myocardial ischaemia or hypertension; age over 50 yr; ASA grade III or IV; and type of anaesthetic.

In Canada, data relating to complications from anaesthesia were collected over two decades.24 Data from follow-up of 112 961 patients showed that nearly 10% of patients were either inconvenienced or suffered some morbidity as a result of the anaesthetic, while 0.45% suffered significant morbidity. The most common complications were nausea, vomiting and sore throat.

The Canadian Four-Centre Study reported on the follow-up of 27 184 patients who underwent anaesthesia and surgery between 1987 and 1989.2326 A panel of experts defined 115 major events and classified them into anaesthesia-related, surgery-related or disease-related categories. There was an anaesthetic involvement in 10.3% of major events. There were no anaesthetic deaths.

More recently, Myles and colleagues72 developed a 40-item questionnaire, the QoR-40, which covered various aspects of recovery: physical comfort, pain, physical independence, emotional state and psychological support. In an accompanying paper,73 the authors compared the occurrence of specific outcomes, such as nausea, vomiting, pain and other complications, with patient satisfaction. The level of satisfaction was high (96.8%) and the level of dissatisfaction low (0.9%). There was a very strong correlation between dissatisfaction and the occurrence of intraoperative awareness (patients who were aware were 55 times more likely to be dissatisfied with their anaesthetic care). Patients who suffered severe nausea and vomiting, or moderate or severe postoperative pain were about four times more likely to be dissatisfied.

Other sources of information

Because of the high rate of litigation against anaesthetists in some countries, analysis of claims for compensation has been used to examine the pattern of injury which patients may suffer, or believe that they have suffered, as a result of the actions of anaesthetists. There is a risk of bias in analysing claims for compensation, in that the complaints relate predominantly to events which the patient does not expect. For example, totally inept treatment of postoperative pain is unlikely (at present) to result in a claim for compensation, because the patient expects to experience postoperative pain. In addition, the pattern of claims is influenced by the personality of the patient, and, in some countries, by the availability of free legal advice. While patients are prepared to accept that surgery may not be entirely successful, or may be associated with a small incidence of complications, they are often unwilling to acknowledge that any consequence which they attribute to anaesthesia, or the anaesthetist, is acceptable. In one extreme example, a patient attempted to sue her anaesthetist for failing to diagnose breast cancer at the preoperative visit when she underwent cystoscopy; the tumour was diagnosed 6 months later.

Despite attempts during the early 1990s to persuade the NHS Litigation Authority to analyse claims made against hospitals so that the pattern of injury could be understood, the causes of injury analysed and lessons from mistakes made widely known in an attempt to reduce the risk of recurrence, virtually no information has been made available about the injuries which lead to litigation in NHS hospitals. The Medical Defence Union has, for many years, allowed its database to be analysed in an attempt to educate doctors about the causes of injuries which lead to litigation, with a view to reducing their incidence through education. Since the early 1990s, the Medical Defence Union, in common with other medical defence organizations, has indemnified doctors in private practice in the UK, and also indemnifies doctors in a number of other countries. The information which the Medical Defence Union gathers is not necessarily typical of injuries in NHS hospitals, but the similarities are probably very considerable.

Data generated by the Medical Defence Union (personal communication) indicate that 63% of claims against anaesthetists arise from damage to teeth; in the vast majority of these claims, the injury is very minor and the cost of repair is small. After damage to teeth, the commonest complaints against anaesthetists were that a medication error had occurred (7% of all claims), or that death was alleged to have resulted from anaesthesia (6%). Neurological problems (other than brain injury) and awareness during anaesthesia each accounted for 5% of claims. Brain damage was alleged in 4% of claims.

Critical incident reporting is a technique which was developed by psychologists to evaluate aspects of human behaviour and to study the causes of good and bad performance. An incident is any observable human activity that is sufficiently complete in itself to permit inferences and predictions to be made about the person performing the act. A critical incident is an incident whose purpose or intent is clear to the observer and the consequences of which are sufficiently definite to leave little doubt about its effect. The technique was described first in 1954,40 but had been used extensively by the US Air Force during the Second World War. Trained observers collected large numbers of factual observations which made an important positive or negative contribution to the activity being studied (e.g. the reasons for failure of bombing missions). When these critical incidents were studied, they could be categorized and steps could be taken to improve success (for example, in guiding recruitment policy so that pilots with good performance could be predicted) and to reduce failure (for example, by improving cockpit design).

In medicine, the critical incident technique has been applied to identifying the causes of errors in administration of drugs by nurses,84 the quality of care by paediatricians, obstetricians, physicians and surgeons,85 and the analysis of professional behaviour in doctors involved in child health.101 The first report of its use in anaesthesia was in 1978. Cooper et al.30 examined errors and equipment failures in anaesthetic practice. Their definition of a critical incident was:

an occurrence that could have led (if not corrected and discovered in time) or did lead to an undesirable outcome ranging from increased length of hospital stay to death. It must also involve error by a member of the anaesthetic team or a failure of the anaesthetists' equipment to function; occur while the patient is under anaesthetic care; be described in clear detail by an observer or member of the anaesthetic team; and be clearly preventable.

The large majority of critical incidents do not lead to an adverse outcome. The technique is therefore valuable because much larger quantities of data can be collected than by identification of adverse outcomes and analysis of the causes. In their first study,30 Cooper et al. employed an interview technique: anaesthetists were asked to describe preventable happenings which they had observed involving either equipment failure or human error. The interviews lasted between 60 and 90 min, and a total of 72 interviews were conducted. Staff anaesthetists reported an average of seven incidents, and trainees reported an average of eight; more than half of the incidents had occurred within the previous 6 months, but one-fifth had occurred more than 3 yr earlier. The most frequent events involved breathing system disconnections, inadvertent changes in gas flows, errors with syringes, problems with gas supply and disconnection of intravenous lines. Inadequate experience, unfamiliarity with equipment, poor communication, haste, inattention, fatigue and failure to perform equipment checks were cited as associated factors.

In subsequent studies,282975 the same group collected data prospectively, using anonymized reporting, produced a more specific analysis of mishaps with substantive negative outcomes, and evaluated the effectiveness of some monitoring devices. The early studies can be criticized on the basis that retrospective reporting is prone to error because memory decays rapidly, but their early work was nevertheless invaluable. It was instrumental in the routine adoption of the use of low pressure alarms in ventilator circuits.

Critical incident reporting has been adopted widely in anaesthesia over the last two decades. Many anaesthetic departments collect data internally, using the data to identify and correct faults with specific items of equipment, to modify protocols, guidelines and training, and to provide feedback at departmental meetings. One of the largest studies is the Australian Incident Monitoring Study,51 which, in 1993, reported in detail the findings from analysis of the first 2000 incident reports. The commonest incidents reported are shown in Table 5, and the most commonly quoted associated factors in Table 6. There is a close similarity with the factors which contributed to death in the CEPOD study13 (Table 3). Cooper et al.29 reported that 70 patients suffered a substantive negative outcome (death, cardiac arrest, cancelled operation, extended stay in the recovery room, ICU or hospital) from 1089 critical incidents. There was an average of 2.5 critical incidents in patients with a negative outcome and 3.4 associated factors. There was also a higher frequency of moderately or severely ill patients among those who had a negative outcome (71%) in comparison with the overall frequency of similar patients in whom critical incidents were reported (41%). In the Australian Incident Monitoring Study, minor physiological changes occurred in association with 30% of incidents, major physiological changes but no injury followed 18%, physical morbidity occurred in 6%, awareness in 1%, and death was associated with 1.5%.102

Table 5

Examples of the most commonly quoted critical incidents in the Australian Incident Monitoring Study51

Problems with breathing system 
    Disconnections 
    Misconnections 
    Leaks 
Problems in administration of drugs 
    Overdosage 
    Underdosage 
    Wrong drug 
Problems with intubation and control of airway 
    Failed intubation 
    Oesophageal intubation 
    Endobronchial intubation 
    Accidental or premature extubation 
    Aspiration 
Failure of equipment 
    Laryngoscopes 
    Intravenous infusion devices 
    Breathing system valves 
    Monitoring devices 
Problems with breathing system 
    Disconnections 
    Misconnections 
    Leaks 
Problems in administration of drugs 
    Overdosage 
    Underdosage 
    Wrong drug 
Problems with intubation and control of airway 
    Failed intubation 
    Oesophageal intubation 
    Endobronchial intubation 
    Accidental or premature extubation 
    Aspiration 
Failure of equipment 
    Laryngoscopes 
    Intravenous infusion devices 
    Breathing system valves 
    Monitoring devices 
Table 6

Examples of the commonest factors associated with critical incidents in the Australian Incident Monitoring Study51

Inattention/carelessness 
Inexperience 
Haste 
Failure to check equipment 
Unfamiliarity with equipment 
Poor communication 
Restricted visual field or access 
Failure of planning 
Distraction 
Lack of skilled assistance 
Lack of supervision 
Fatigue and decreased vigilance 
Inattention/carelessness 
Inexperience 
Haste 
Failure to check equipment 
Unfamiliarity with equipment 
Poor communication 
Restricted visual field or access 
Failure of planning 
Distraction 
Lack of skilled assistance 
Lack of supervision 
Fatigue and decreased vigilance 

Critical incident techniques have an important place in the identification of risk and improvement of safety by drawing the attention of anaesthetists to potential errors and by identifying deficiencies in equipment design and function. However, there are potential disadvantages associated with continuous critical incident reporting. An individual anaesthetist who is conscientious in reporting critical incidents in a departmental system, or an individual department in a national scheme, may be regarded as a poor performer if there is any breach of anonymity, and this may encourage under-reporting. Similarly, an apathetic individual or department not only causes under-reporting, but may appear (justifiably or not) to be a good performer. Continuous reporting may result in loss of enthusiasm with time, causing progressively increasing under-reporting; limiting reporting to specific areas of anaesthetic practice in rotation might help to maintain enthusiasm. In addition, it is necessary to define the term ‘critical incident’ carefully; some authors have used Cooper's original definition, but others have used the term to describe events which may or may not be preventable, which may or may not include anaesthetic error or equipment failure, and which may include only incidents which result in an adverse outcome. The term ‘real or potential adverse event’ has been proposed34 as a self-explanatory and unambiguous alternative.

Severe brain injury

The incidence of severe brain injury caused by anaesthesia is difficult to ascertain. Severe brain injury is usually associated with either severe hypoxaemia or cerebral ischaemia. Severe hypoxaemia in the perioperative period may occur as a result of inadequate ventilation (caused usually by disconnection of the breathing system, oesophageal intubation, ventilatory depression, airway obstruction or failure or misuse of mechanical ventilation), severe lung disease, delivery of an inappropriately low inspired oxygen concentration, bronchospasm (most commonly caused by asthma, anaphylaxis or aspiration), endobronchial intubation or haemo- or pneumothorax. Cerebral ischaemia may result from profound hypotension caused by vasodilatation due to adverse effects of drugs or anaphylaxis, severe hypovolaemia, or a greatly reduced cardiac output related to cardiac failure, cardiac arrest, severe bradycardia or a malignant arrhythmia. Kawashima et al.55 found that the incidence of a persistent vegetative state attributable to anaesthesia in surgical patients was approximately 1:170 000, more than three times less frequent than death attributable to anaesthesia, but this figure may be misleading because it is possible that some brain-damaged patients were allowed to die and were therefore classified as deaths related to anaesthesia.

In the mid-1980s, the Committee of Professional Liability of the American Society of Anesthesiologists (ASA) began a structured evaluation of adverse anaesthetic outcomes with the purpose of improving safety by devising strategies to prevent anaesthetic mishaps. Data were extracted from ‘closed claims’ files of 17 insurance organizations which indemnify doctors (ASA Closed Claims Analysis). Because of the long delays incurred in the legal process, this has generated information about the pattern of injury over three decades, in addition to data derived from analysing the causes of injury. In the 1970s, 64% of claims involved permanent and disabling injuries, or death. In the 1980s, approximately half of all claims fell into this category, and in the 1990s, this figure had decreased to 41%.79 There was a clear reduction in the frequency of complications caused by respiratory events over the three decades (Fig. 1). The most common complications associated with claims in the 1990s are shown in Table 7.

Fig 1

Most common ‘damaging events’ in the ASA Closed Claims Study; the ‘damaging event’ is the mechanism which allegedly caused the injury. Data from the ASA Closed Claims Study.79 (Figure reproduced with the permission of the Anesthesia Patient Safety Foundation.)

Fig 1

Most common ‘damaging events’ in the ASA Closed Claims Study; the ‘damaging event’ is the mechanism which allegedly caused the injury. Data from the ASA Closed Claims Study.79 (Figure reproduced with the permission of the Anesthesia Patient Safety Foundation.)

Table 7

Most common complications found on analysis of 1784 American closed claims related to anaesthetic care in the 1990s79

Injury
 
Frequency (%)
 
Death 23 
Nerve injury 21 
Brain damage 
Burns/skin inflammation 
Awareness 
Eye injury 
Backache 
Headache 
Pneumothorax 
Aspiration pneumonitis 
Injury to newborn 1.5 
Injury
 
Frequency (%)
 
Death 23 
Nerve injury 21 
Brain damage 
Burns/skin inflammation 
Awareness 
Eye injury 
Backache 
Headache 
Pneumothorax 
Aspiration pneumonitis 
Injury to newborn 1.5 

A number of factors are known to influence the risk of serious complications associated with anaesthesia and surgery. Morita et al.71 reported that the incidences of critical events, death and cardiac arrest were greatest at the extremes of age (Fig. 2), and lowest between the ages of 1 and 65 yr. In the Australian Incident Monitoring Study,102 the incidence of ‘harm’ which might have resulted from critical incidents was related to the ASA grading of the patient. ASA V patients were much more likely to be harmed by an incident than patients in categories I–IV. Chopra20 found that all ASA IV or V patients who suffered perioperative cardiac arrest died, whereas the majority of ASA I–III patients survived.

Fig 2

Relationship between age and complications associated with anaesthesia.71

Fig 2

Relationship between age and complications associated with anaesthesia.71

In a study from Denmark, Pedersen77 investigated the relationship between complications attributable to anaesthesia and both ASA grade and whether the procedure was an emergency or elective procedure. He also studied the influence of these factors on outcome after a complication, and the frequency of preventable errors. He found that complications attributable to anaesthesia were slightly less common during emergency procedures than elective operations, and that the number of preventable errors leading to complications was lower in emergency procedures. However, the probability of a negative outcome from an anaesthetic complication was higher in emergency patients. The number of complications attributable to anaesthesia was almost eight times higher in patients in ASA grades III–V than in those with ASA grade I or II, and the incidences of a negative outcome and preventable errors were also much higher in patients with a high ASA grade.

Oesophageal intubation is a well-known complication of anaesthesia, and can have devastating consequences if not detected rapidly. The incidence of undetected oesophageal intubation has decreased since the introduction of routine capnography, but it has not been eliminated. The Australian Incident Monitoring Study49 reported that 35 of its first 2000 critical incidents involved oesophageal intubation. In 15 of these incidents there had been no difficulty reported during intubation. In 18 cases, a trained specialist anaesthetist intubated the trachea; in 15 cases a trainee was responsible, and in two incidents the identity of the individual involved was not known. In the ASA Closed Claims Study,15 analysis of legal claims for damages related to oesophageal intubation revealed that, in more than 80% of claims, the standard of care provided by the anaesthesiologist was considered, on retrospective review, to be substandard (Fig. 3). A later study18 showed that in more than 80% of claims related to oesophageal intubation the patient had died and the remainder had suffered brain damage (Fig. 4). This information is, of course, rather misleading; provided oesophageal intubation is detected promptly, no injury is caused and a patient has no grounds for legal action.

Fig 3

Proportion of respiratory and non-respiratory events judged to have been associated with acceptable or substandard care in the ASA closed claims analysis.15 (Figure reproduced with permission.)

Fig 3

Proportion of respiratory and non-respiratory events judged to have been associated with acceptable or substandard care in the ASA closed claims analysis.15 (Figure reproduced with permission.)

Fig 4

Frequencies of death and brain damage associated with respiratory and non-respiratory complications in the ASA closed claims analysis.1518

Fig 4

Frequencies of death and brain damage associated with respiratory and non-respiratory complications in the ASA closed claims analysis.1518

The ASA Closed Claims Study also found1518 that the standard of care in claims relating to injury caused by inadequate ventilation (again almost always resulting in death or brain injury, but for the same reason) was judged to be inadequate in the large majority of cases (Fig. 3). In almost 60% of cases of injury related to difficult intubation, the actions of the anaesthesiologist were judged to have been substandard. In contrast, the standard of care was judged to have been inadequate in approximately 30% of claims which did not involve the respiratory system. As noted above, the incidence of claims involving the respiratory system has diminished since the introduction of pulse oximetry and capnography, but hypoxic brain damage and death resulting from failure to ventilate the lungs do still occur and, arguably, are even more culpable if appropriate monitors have not been used, or if the information which they provide has been ignored.

Inability to ventilate the lungs, again with the potential to cause hypoxic brain injury and death, may occur as a result of an obstruction in a tracheal tube, laryngeal mask airway or the anaesthetic breathing system. The Australian Incident Monitoring Study93 received reports of obstruction of the tracheal tube caused by kinking, biting, secretions or blood, a surgical gag and foreign bodies, and there has been a recent incident in the UK in which a child died as a result of obstruction of an angle piece by the cap of an intravenous infusion set.53 Obstruction of the breathing system was one of a number of ‘lesser’ adverse respiratory events considered by the ASA Closed Claims Study.18 Other injuries in this category were airway trauma, pneumothorax, bronchospasm and aspiration. The commonest site of airway trauma was the larynx, but there were also injuries to the pharynx and oesophagus (usually associated with difficult intubation), the nasopharynx, the temporomandibular joint and the trachea.

In the Australian Anaesthetic Incident Monitoring Study,58 244 of 5000 critical incidents were related to vomiting or regurgitation associated with anaesthesia, and there were 133 incidents which involved aspiration. Passive regurgitation occurred three times more commonly than active vomiting. Both regurgitation/vomiting and aspiration occurred approximately twice as commonly in patients undergoing elective surgery in comparison with patients who required emergency treatment. Most incidents occurred during induction of anaesthesia, some took place during maintenance, emergence and later recovery, and very few occurred in association with the process of extubation. The immediate effect of aspiration was judged to be ‘major’ in more than half of the incidents, but the final outcome was serious in only about one-third of patients. In the ASA Closed Claims Study,18 approximately half of the patients who sued because of injury associated with aspiration died or suffered brain damage, but this figure is distorted by outcome bias.

Drug administration errors

Drug administration errors are a common source of injury in all areas of hospitals, including the operating theatre.11 Drugs may be omitted, repeated, substituted, given in error, given in an incorrect dose or given by an incorrect route. In anaesthetic claims, Bowdie11 reported that drug administration errors involved, in decreasing order of frequency: succinylcholine; inhaled anaesthetic agents; opioids; local anaesthetics; epinephrine; cardiovascular agents; antibiotics; and non-depolarizing neuromuscular blocking drugs. Most drug administration errors are harmless (e.g. administration of cefuroxime in an attempt to induce anaesthesia, believing the syringe to contain thiopental), but some are not (e.g. administration of succinylcholine before induction of anaesthesia, believing the syringe to contain fentanyl).

Awareness during general anaesthesia

Awareness during anaesthesia was found to occur in 1–2% of general anaesthetics in the 1970s, but the reported frequency in the last 10–15 yr has been of the order of 1.5–2:1000. Many patients identified in carefully controlled studies as having been aware at some stage during anaesthesia have not suffered pain or distress, and many had not appreciated that they had been awake at a time when consciousness was not intended. The incidence of awareness associated with distress is unknown, but is probably only 10–20% of the reported overall incidence of awareness (i.e. 1.5–4.0:10 000). In my experience, the incidence of awareness leading to litigation in the UK has remained relatively constant over the last 25 yr. Awareness claims accounted for 1.9% of claims in the ASA Closed Claims Project,35 but in my experience the proportion of claims relating to awareness forms a substantially higher proportion of claims against anaesthetists in the UK.

Awareness is usually the result of provision of inadequate concentrations of anaesthetic drugs in the paralysed patient. The commonest causes of awareness are the use of an anaesthetic technique which is predictably inadequate to ensure lack of spontaneous recall in all patients and failure to check or monitor the use of gas delivery equipment. The commonest time for awareness to occur is at, or shortly after, the time of skin incision. This is often because, either as a result of failure of understanding of the pharmacokinetics of anaesthetic agents, or due to fears of producing cardiovascular depression before surgery starts, insufficient concentrations of inhaled agents have been administered to provide adequate anaesthesia for surgery as the effects of the intravenous induction agent decline. Awareness may occur also if air is entrained into mechanical ventilators, if vaporizers become empty during anaesthesia or during difficult intubation if the initial induction dose of anaesthetic is not supplemented.

The commonest cause of awareness during total intravenous anaesthesia is an interruption of the supply of drug, caused either by disconnection of, or obstruction to, the infusion.86 However, a number of cases have arisen as a result of confusion on the part of the anaesthetist concerning infusion regimens. The commonest mistake is the failure to appreciate that most published infusion regimens were effective because nitrous oxide and/or large doses or infusions of an opioid were also given; if these regimens are used without nitrous oxide or an appropriate dose of opioid, they are associated with a significant risk of awareness. Another cause of awareness is the failure to appreciate that, in order to achieve a stable blood concentration of intravenous agent, the infusion must be started immediately after the induction dose; delay until after the patient has been transferred into the operating theatre is likely to result in periods in which the brain concentration of anaesthetic agent is less than that which produces effective anaesthesia.

In obstetric anaesthesia, concerns regarding the effects of anaesthetic agents on the newborn baby and on the risk of haemorrhage from the uterus resulted in the use of anaesthetic techniques which would not be regarded as adequate in other areas of surgery. The perceived risks to the baby and the uterus have almost certainly been overemphasized, and a substantial reduction in the incidence of awareness has been achieved since the introduction of improved techniques.64 However, the incidence of awareness remains higher in obstetric anaesthesia than in other areas of anaesthetic practice. Most episodes of awareness now occur between skin incision and the delivery of the baby, and result from the short period between induction of anaesthesia and the start of surgery, and reluctance to use ‘overpressure’ to achieve a rapid increase in the concentration of volatile agent in the blood and brain.66 A small proportion of patients who complain of intraoperative awareness have experienced events in the immediate postoperative period, but interpreted the events as having occurred during operation.94 There is also a small proportion of patients who have suffered intraoperative awareness, but in whom an episode of awareness can be justified, most commonly because an emergency has arisen which has necessitated a reduction in the concentration of anaesthetic delivered to the patient, or temporary cessation of administration of anaesthetic agents.

Patients who have been aware during anaesthesia, and who have experienced pain and distress, may develop the post-traumatic stress disorder, which may last for many years. Advice to anaesthetists confronted with a patient who claims to have been aware during anaesthesia has been published.1

Obstetric anaesthesia

Most claims relating to obstetric anaesthesia in the ASA Closed Claims Study involved regional anaesthesia, in comparison with only about one-fifth of non-obstetric claims.16 Claims related to maternal death accounted for 42% of claims in which general anaesthesia had been used, but only 12% of claims when regional anaesthesia had been employed. The proportions of claims related to maternal brain damage, death of a newborn baby and brain damage to the baby were higher when general anaesthesia had been used, but claims provoked by nerve damage to the mother, pain during anaesthesia, emotional distress and back pain were commoner when the mother had received regional anaesthesia.

Peripheral nerve damage

Peripheral nerve damage is usually the result of compression or stretching of the nerve. In a few cases, nerves are damaged by direct trauma from needles or extravasated drugs. Most lesions involve demyelination at a point of localized compression. In most cases, remyelination results in recovery in 6–8 weeks, but in some patients recovery is more prolonged, and in a few the injury is permanent. In the ASA closed claims analysis,1759 15% of claims against anaesthetists related to peripheral nerve injury. Injury to the ulnar nerve represented about one-third of these claims, brachial plexus injury 23% and damage to the lumbosacral roots 16%. Frequently, a cause could not be identified clearly.

The ulnar and common peroneal nerves are particularly susceptible to damage because they lie, at some point, in a very superficial position close to bony promontories. However, there is evidence that many patients have pre-existing subclinical lesions of these nerves related to normal posture, often occupational, such as sitting with the legs crossed or leaning habitually on an elbow. In these patients, a brief period of normally innocuous stretching or compression may result in a clinical lesion. Often, electrophysiological studies in patients who develop a lesion in the postoperative period reveal that the nerve on the contralateral side is also abnormal.2 Patients with a pre-existing general neuropathy are also at increased risk. The presence of a neuropathy and any history of regular paraesthesiae on waking from normal sleep should be elicited at the preoperative visit, and special care should be taken during the perioperative period.

The brachial plexus may suffer injury if the arm is placed in any extreme position during anaesthesia, but particularly if it is abducted and externally rotated. In the supine position, the risk is highest if both arms are abducted, if an arm is abducted beyond 90°, if the arm is abducted and the head is extended and rotated to the opposite side, and if the arm lies below the horizontal plane of the body. In the prone position, there is a risk of damage to the brachial plexus if the arms are abducted excessively; this is most easily prevented by keeping the arms at the side of the body. There is also a risk of damage to the brachial plexus if shoulder braces are used to prevent a patient slipping from a steep head-down position on the operating table. Brachial plexus injury can also complicate cardiac surgery, as a result of stretching during median sternotomy.

Damage to the radial nerve in the upper arm occurs usually as a result of compression, for example under a pneumatic tourniquet.41 The sciatic nerve is at risk in thin patients placed on a hard operating table for a long operation, but may also be damaged by stretching in the lithotomy position, or by compression by a hard pad placed under the pelvis when patients are placed in the Lloyd-Davies position. Lumbosacral nerve root injuries are associated most commonly with pain or paraesthesiae accompanying a regional anaesthetic technique. Nerves injured less commonly include the tibial, femoral, obturator, saphenous, pudendal, supraorbital and facial.

Anaphylaxis

An anaphylactic reaction is an exaggerated response of an organism to a foreign protein (or other substance to which it has become sensitized) associated with the liberation of histamine, serotonin and other vasoactive substances. The clinical manifestations are shown in Tables 8 and 9.39 An anaphylactoid reaction is a term that encompasses all reactions which are clinically indistinguishable from anaphylactic reactions but for which other mechanisms are involved. Examples include histamine-mediated transient hypotension after high-dose opioid administration and most cases of pruritus or urticaria after small doses of opioids. The incidence of anaphylactic reactions associated with anaesthesia has been estimated to be approximately 1:6000 in France61 and 1:10 000 to 1:20 000 in Australia.39 Reactions are more common in female patients. Anaphylactic reactions during anaesthesia may occur to any drug given intravenously, or to latex. A history of previous exposure to the drug is not always necessary, particularly in relation to neuromuscular blocking drugs, where the absence of a history of previous exposure may be as high as 80%. This may be due to sensitization by some foods and cosmetics, and is also related to the fact that patients may be sensitized to a number of muscle relaxants by previous administration of one drug of this class.

Table 8

Clinical features reported in 206 patients with severe anaphylactic reactions during anaesthesia. Numbers represent percentages of patients in whom clinical feature was noted39

Clinical features
 
Induction agents
 
Muscle relaxants
 
Other drugs
 
Number of patients 44 115 47 
Cutaneous features    
    Rash 14 13 
    Urticaria 28 
    Flush 50 41 53 
Bronchospasm    
    Transient 16 23 17 
    Severe 15 11 
Cardiovascular features    
    Tachycardia 93 90 83 
    Bradycardia 15 
    Other arrhythmia 11 19 
    Hypotension 93 92 72 
    Vasodilatation 86 86 79 
    Vasoconstriction   
Oedema    
    Pulmonary  16 11 
    Angio-oedema 16 33 17 
    Generalized 
Gastrointestinal 20 10 
Clinical features
 
Induction agents
 
Muscle relaxants
 
Other drugs
 
Number of patients 44 115 47 
Cutaneous features    
    Rash 14 13 
    Urticaria 28 
    Flush 50 41 53 
Bronchospasm    
    Transient 16 23 17 
    Severe 15 11 
Cardiovascular features    
    Tachycardia 93 90 83 
    Bradycardia 15 
    Other arrhythmia 11 19 
    Hypotension 93 92 72 
    Vasodilatation 86 86 79 
    Vasoconstriction   
Oedema    
    Pulmonary  16 11 
    Angio-oedema 16 33 17 
    Generalized 
Gastrointestinal 20 10 
Table 9

The first clinical feature of an anaphylactic reaction in 206 patients39

Clinical feature
 
Percentage of patients
 
No pulse detected 27.6 
Flush 26.7 
Difficulty in inflation 23.8 
Cough 5.3 
Rash 3.9 
Urticaria 3.4 
Cyanosis 3.4 
Subjective feeling 2.9 
Oedema 2.4 
Hypotension 0.4 
No bleeding 0.4 
Clinical feature
 
Percentage of patients
 
No pulse detected 27.6 
Flush 26.7 
Difficulty in inflation 23.8 
Cough 5.3 
Rash 3.9 
Urticaria 3.4 
Cyanosis 3.4 
Subjective feeling 2.9 
Oedema 2.4 
Hypotension 0.4 
No bleeding 0.4 

There are three important factors which influence the risk of damage associated with a severe anaphylactic reaction.

(i) History. It is essential at the preoperative visit to elicit any history of an adverse event associated with previous anaesthesia which might indicate an anaphylactic reaction. Patients should be asked specifically if they have ever been told that they are allergic to any anaesthetic drug. If a positive history is obtained, the previous records should be scrutinized if they are available. If they are not available, then unless the operation is a dire emergency, skin testing should be undertaken for the drugs which the anaesthetist proposes to use.92

(ii) Prompt recognition and appropriate treatment. The diagnosis of anaphylaxis is clinical. A serious anaphylactic reaction starts rapidly after the causative drug is injected, and can be detected almost immediately if appropriate monitoring is employed. Various treatment regimens have been recommended.92 Prompt recognition and appropriate treatment are associated with a successful response and outcome in most patients, even after a severe anaphylactic reaction. Because anaphylactic reactions are uncommon, it has been recommended that anaesthetists and operating room personnel should rehearse a simulated anaphylaxis drill at regular intervals.92

(iii) Investigation. It is essential to attempt to identify the cause of an anaphylactic reaction and any patient who has a suspected anaphylactic reaction should be investigated fully. Current recommendations in the UK92 are that serum tryptase concentration should be measured in samples of blood taken immediately after the reaction has been treated, 1 h after the reaction and 6–24 h after the reaction; an elevated concentration is a sensitive indicator of the existence of an anaphylactic reaction. If the serum tryptase concentration is elevated, or if the reaction was severe, then skin prick tests should be performed 4–6 weeks later. All drugs administered intravenously during anaesthesia, and all neuromuscular blocking drugs, should be tested; it is preferable also to test for other anaesthetic agents and for allergy to latex. Positive and negative reactions should be documented clearly, and the patient should be given a letter, and if possible a warning bracelet, indicating the drugs which showed a positive reaction.

No drug which is proved, or strongly suspected, to have caused an anaphylactic reaction should be administered again.

Damage to teeth

As noted above, dental damage is the most frequent cause of claims against anaesthetists, although it is often overlooked because the claims are not usually expensive. Lockhart et al.62 surveyed 133 anaesthesia training programmes and estimated the incidence of dental damage to be 1 per 1000 anaesthetics which involved tracheal intubation. Difficult or emergency intubation accounted for about half of the cases of dental damage, but 25% of injuries occurred during extubation or in the recovery room. The upper left incisor tooth was most frequently involved. In 62% of cases in one study,14 dental damage occurred in teeth which had been restored previously or damaged by periodontal disease. A plastic dental shield is used routinely by some anaesthetists, and is claimed to minimize the risk of injury. However, Burton and Baker14 found that ‘the vast majority’ of anaesthetists did not use a protective guard routinely and that 45% of anaesthetists never used one.

Fear of damaging teeth may lead some anaesthetists to take unreasonable risks of other complications.22 Dental damage was avoided in one vulnerable, obese patient who required Caesarean section; a face mask was used, but she died after massive aspiration of gastric fluid.38

Eye injuries

Injuries to the eye during anaesthesia can result in corneal abrasions or blindness.59195 The ASA Closed Claim Study found45 that eye injuries accounted for 3% of total claims against anaesthetists. Corneal abrasion occurred in 35% of these cases, and was most probably the result of the eye opening and the cornea drying during anaesthesia, or trauma from the anaesthetic face mask, laryngoscope or surgical drapes. In most cases (84%), injury was temporary. The second commonest source of injury (30%) was related to movement during ophthalmic anaesthesia, usually due to coughing or bucking; in 81% of these cases the reviewers considered that the standard of care was inadequate. Another potential source of blindness is pressure applied to the eye as a result of inappropriate positioning in the prone position, particularly during neurosurgical procedures.

Embolism

Embolism of air or thrombus may result in cerebral ischaemia. Massive venous embolism causes a great reduction in, or cessation of, cardiac output. Arterial emboli may cause focal or generalized cerebral lesions. Massive pulmonary embolism is unusual during anaesthesia, but may occur in patients who have been immobile, or occasionally after release of a pneumatic tourniquet on the leg; postoperative embolism is much more common. Arterial thrombi may originate in the heart or great vessels, and are usually neither predictable nor preventable by the anaesthetist. Air embolism may occur during neurosurgery or operations in the pelvis if the patient is positioned in such a way that the operation site is above the heart. Oxygen emboli have caused cardiac arrest after instillation of hydrogen peroxide into surgical wounds.97 Fatal carbon dioxide emboli have occurred during laparoscopy.6 Emboli of air or clot, or reduced cerebral perfusion pressure, may cause focal or global ischaemia after cardiopulmonary bypass;90 microemboli of platelets, air bubbles or other substances result in a high incidence of minor neurological and psychological sequelae, most of which are of little clinical significance and some of which are temporary.

Regional anaesthesia

Regional anaesthesia is associated with both morbidity and mortality. The ASA Closed Claims Project8 found that, during the 1990s, in 10% of claims related to the use of regional anaesthesia the patient had died; however, this was an improvement on the previous two decades. The primary reason for death was cardiac arrest associated with neuraxial blockade. Permanent disabling injuries were responsible for 23% of claims relating to regional anaesthesia, the commonest cause being nerve blocks of the eye. Second in frequency were pain management-related claims. Nerve injuries associated with neuraxial and peripheral nerve blocks accounted for 20% of claims in this category, and epidural haematoma for 13%.

Spinal cord/nerve root damage

In the UK there is, in my experience, a much greater risk of spinal cord damage associated with the use of spinal or combined spinal and epidural block than is apparent from the literature. There is a wide range of estimates of the risk of serious neurological damage associated with regional anaesthesia. A study based on data gathered from the Finnish no-fault compensation system4 reported that the risk of serious complications from spinal block was 0.46 per 10 000 procedures, with an incidence of neurological damage of 0.34 per 10 000 procedures. For epidural block, there was a risk of serious complications of 0.52 per 10 000 procedures, and an incidence of neurological damage of 0.24 per 10000 procedures. However, that study may have underestimated the risk because many patients who suffered complications may not have applied for compensation. A study of more than 17 000 patients in Sweden31 found that the risk of permanent neurological damage associated with spinal block was 1:2800 and that the risk of permanent neurological damage associated with epidural block was 1:923. Reynolds82 reported seven cases of spinal cord injury associated with attempted spinal block in the UK. The denominator is not known, but it seems likely that the incidence of neurological damage associated with spinal and epidural block is not insignificant. Broadbent et al.12 found that anaesthetists identified the correct lumbar interspace in only 29% of cases; in most instances the actual space selected was higher than the level at which the anaesthetist believed it to be.

There have also been reports of damage to the cervical spinal cord associated with interscalene block performed under general anaesthesia.9 Magnetic resonance imaging demonstrated a syrinx at the cervical level resulting from injection of local anaesthetic into the spinal cord.

The early signs of contact between a needle tip and nerve root or spinal cord are masked if the block is performed after induction of general anaesthesia, and thus the likelihood of more severe and permanent damage is probably increased; in my opinion, it is prudent to perform spinal or epidural block before inducing general anaesthesia if a combination of regional and general anaesthesia is to be employed.

There have been instances in which an inappropriate substance has been injected inadvertently into the epidural or subarachnoid space, causing extensive neurological damage. A number of substances are known to cause neurotoxicity, including potassium chloride and calcium chloride.89 Most local anaesthetic drugs used in the UK are not associated with neurotoxicity, although there are concerns about their intrathecal use with epinephrine and their administration through a subarachnoid catheter.83 There have also been reports of neurotoxicity associated with the use of chloroprocaine, related probably to toxicity of a preservative.80 However, a substantial proportion of injuries to nerve roots have no obvious cause. There is increasing suspicion that inadvertent subdural injection of local anaesthetic may be responsible for a proportion of unexplained neurological lesions in patients who have undergone epidural anaesthesia;81 injection of radio-opaque dye through a catheter placed inadvertently in the subdural space has demonstrated that fluid may produce compression of nerve roots and the spinal cord.68 In the ASA Closed Claims study,59 2.5% of claims against anaesthetists related to alleged damage to the lumbosacral nerve roots.

Damage to the spinal cord may occur also as a result of ischaemic lesions, because of either systemic hypotension or local occlusion of the blood supply (the latter usually in association with coeliac plexus block). Injection of steroids into the epidural space (without a product licence) has been claimed to cause damage to nerve roots in a number of cases;74 this is a controversial issue.99

Pain during surgery

Up to 25% of patients experience pain or discomfort during operations performed under regional anaesthesia, and particularly during Caesarean section. A proportion of these patients suffer severe pain. It is essential that the degree and extent of block are tested adequately before surgery starts, and that the anaesthetist is prepared to take appropriate steps, including induction of general anaesthesia, if the block becomes inadequate during the operation. Consequently, preparation of patients for major regional techniques must be the same as that for patients scheduled to undergo general anaesthesia, and the anaesthetist must remain available to monitor the effects of the block, and intervene if necessary, during the procedure.

There have been a number of recent cases in which patients have claimed to have been assaulted because regional or local techniques were employed without specific consent, and performed under general anaesthesia. In one claim,33 a patient who suffered localized damage to lumbar and sacral nerve roots unilaterally claimed that the caudal anaesthetic to which she attributed the injury should not have been performed during general anaesthesia without her specific consent. In finding against the patient, the judge indicated that, in his opinion, it was inappropriate to expect that patients should be told every detail of the proposed procedure, and that there was no realistic distinction between omitting to tell a patient that she would receive a caudal block for provision of intraoperative analgesia and postoperative pain relief, and omitting to tell her that a tube would be placed in her trachea or that an intramuscular injection of morphine would be given during emergence from the anaesthetic. However, it cannot necessarily be assumed that a similar judgement would be given in respect of a potentially more hazardous block, such as subarachnoid, or lumbar or thoracic epidural.

Miscellaneous injuries

Pain related to administration of succinylcholine is a very common complaint, and patients should be warned of the possibility at the preoperative visit. Fractured ribs and other injuries have resulted from patients being dropped from the operating table on to the floor. Tissued infusions may result in damage to skin or nerves, and, in rare circumstances, ischaemic damage to an entire limb because of compartment syndrome. Skin necrosis may occur if irritant drugs intended for deep intramuscular injection are delivered instead into the subcutaneous tissues. Skin damage may occur also as a result of burns, pooling of disinfectant, and pressure.

Laryngeal damage may be minor (e.g. granulomata) or major (e.g. dislocation of cartilages, or damage to nerves); prolonged and strenuous attempts to intubate the trachea may result rarely in perforation of the pharynx or oesophagus, resulting in retropharyngeal abscess formation or potentially fatal mediastinitis.18

Development of pneumothorax in the perioperative period is associated most commonly with attempts to insert central venous or pulmonary artery catheters. While this is a recognized risk in a small proportion of patients, the risk is increased if multiple attempts have been made, or if an inexperienced anaesthetist has made the attempts with inadequate supervision. Bilateral attempts at cannulation of the internal jugular or subclavian veins should be avoided unless absolutely necessary, because of the potentially catastrophic consequences of bilateral tension pneumothoraces. The anaesthetist must always have a high index of suspicion for the development of pneumothorax in the perioperative period, particularly if a central venous catheter has been inserted, or if supraclavicular brachial plexus block, intercostal block or intrapleural injection of local anaesthetic has been performed.

Anaesthetized patients may become hypothermic if steps are not taken to maintain normothermia by minimizing heat loss and by using warming devices. In patients undergoing total hip arthroplasty, intraoperative blood loss was significantly higher and blood transfusion needs were significantly greater in patients who were allowed to become mildly hypothermic (mean final core temperature 35.0°C) than in patients in whom normothermia was maintained (mean final core temperature 36.6°C).88 In a study of patients undergoing colorectal surgery, hypothermia (mean final core temperature 34.7°C) was associated with an incidence of wound infection of 19%, in comparison with an incidence of 6% in patients in whom normothermia was maintained (mean final core temperature 36.6°C). The duration of hospitalization was increased by a mean of 2.6 days in the patients who had been hypothermic.60 The authors suggested that this finding could be explained by tissue hypoxia related to peripheral vasoconstriction during and shortly after surgery. The same research group has demonstrated that administration of oxygen 80% during colorectal surgery and for 2 h postoperatively was associated with a lower incidence of postoperative wound infection (5.2%) in comparison with the incidence in patients who received an inspired oxygen concentration of 30% during the same period (11.2%).46

Anaesthetized patients are also susceptible to burns. In the ASA Closed Claims Analysis there were 54 cases of burns (in a total of 3000 claims), mostly from devices intended to prevent hypothermia.19 Injury related to the use of electrical equipment is also an important issue; the literature on electrical safety is vast, and detailed discussion is beyond the scope of this article.

Conclusions

Although many claims have been made that the risk associated with anaesthesia has decreased, there is little hard supportive evidence except in relation to serious respiratory complications, where improved monitoring appears to have reduced the incidence substantially over the last three decades. Many assumptions have been based on retrospective analysis of events which ‘could have been prevented’, but numerous studies have demonstrated that the same pattern of errors, incidents and accidents continues to occur. Although a small number of studies have suggested that there has been a huge reduction in the incidence of mortality attributable to anaesthesia, other studies suggest that the incidence is not changing significantly. The safety of patients does not depend solely on the application of standards of practice, the purchase of new equipment and the institution of new monitoring techniques. Safety can be increased only by combining the use of modern technology with improvements in education, training, supervision, attitudes, standards of clinical practice, audit and vigilance.

References

1
Aitkenhead AR. Awareness during anaesthesia: what should the patient be told?
Anaesthesia
 
1990
;
45
:
351
–2
2
Alvine FG, Schurrer ME. Postoperative ulnar-nerve palsy. Are there predisposing factors?
J Bone Joint Surg Am
 
1987
;
69
:
255
–9
3
Arbous MS, Grobbee DE, van Kleef JW, et al. Mortality associated with anaesthesia: a qualitative analysis to identify risk factors.
Anaesthesia
 
2001
;
56
:
1141
–53
4
Aromaa U, Lahdensuu M, Cozanitis DA. Severe complications associated with epidural and spinal anaesthesias in Finland 1987–1993. A study based on patient insurance claims.
Acta Anaesthesiol Scand
 
1997
;
41
:
439
–44
5
Batra YK, Bali IM. Corneal abrasions during general anesthesia.
Anesth Analg
 
1977
;
56
:
863
–6
6
Beck DH, McQuillan PJ. Fatal carbon dioxide embolism and severe haemorrhage during laparoscopic salpingectomy.
Br J Anaesth
 
1994
;
72
:
243
–5
7
Beecher HK, Todd DP. A Study of the Deaths Associated with Anesthesia and Surgery. Springfield, Illinois: Charles C Thomas,
1954
8
Ben-David B. Complications of regional anesthesia: an overview.
Anesthesiol Clin North Am
 
2002
;
20
:
665
–7
9
Benumof JL. Permanent loss of cervical spinal cord function associated with interscalene block performed under general anesthesia.
Anesthesiology
 
2000
;
93
:
1541
–4
10
Bodlander FMS. Deaths associated with anaesthesia.
Br J Anaesth
 
1975
;
47
:
36
–40
11
Bowdie TA. Drug administration errors from the ASA Closed Claims Project.
ASA Newsletter
 
2003
;
67
:
11
–3
12
Broadbent CR, Maxwell WB, Ferrie R, Wilson DJ, Gawne-Cain M, Russell R. Ability of anaesthetists to identify a marked lumbar interspace.
Anaesthesia
 
2000
;
55
:
1122
–6
13
Buck N, Devlin HB, Lunn JN. Report on the confidential enquiry into perioperative deaths. London: Nuffield Provincial Hospitals Trust, The Kings Fund Publishing House,
1987
14
Burton JF, Baker AB. Dental damage during anaesthesia and surgery.
Anaesth Intensive Care
 
1987
;
15
:
262
–8
15
Caplan RA, Posner KL, Ward RJ, Cheney FW. Adverse respiratory events in anesthesia: a closed claims analysis.
Anesthesiology
 
1990
;
72
:
828
–33
16
Chadwick HS, Posner K, Caplan RA, Ward RJ, Cheney FW. A comparison of obstetric and nonobstetric anesthesia malpractice claims.
Anesthesiology
 
1991
;
74
:
242
–9
17
Cheney FW, Domino KB, Caplan RA, Posner KL. Nerve injury associated with anesthesia: a closed claims analysis.
Anesthesiology
 
1999
;
91
:
1062
–9
18
Cheney FW, Posner KL, Caplan RA. Adverse respiratory events infrequently leading to malpractice suits: a closed claims analysis.
Anesthesiology
 
1991
;
75
:
932
–9
19
Cheney FW, Posner KL, Caplan RA, Gild WM. Burns from warming devices in anesthesia: a closed claims analysis.
Anesthesiology
 
1994
;
80
:
806
–10
20
Chopra V, Bovill JG, Spierdijk J. Accidents, near accidents and complications during anaesthesia: a retrospective analysis of a 10-year period in a teaching hospital.
Anaesthesia
 
1990
;
45
:
3
–6
21
Clifton BS, Hotten WIT. Deaths associated with anaesthesia.
Br J Anaesth
 
1963
;
35
:
250
–9
22
Clokie C, Metcalf I, Holland A. Dental trauma in anaesthesia.
Can J Anaesth
 
1989
;
36
:
675
–80
23
Cohen MM, Duncan PG, Pope WDB, et al. The Canadian four-centre study of anaesthetic outcomes. II. Can outcomes be used to assess the quality of anaesthesia care?
Can J Anaesth
 
1992
;
39
:
430
–9
24
Cohen MM, Duncan PG, Pope WDP, Wolkenstein C. A survey of 112,000 anaesthetics at one teaching hospital (1975–83).
Can Anaesth Soc J
 
1986
;
33
:
22
–31
25
Cohen MM, Duncan PG, Tate RB. Does anesthesia contribute to operative mortality?
JAMA
 
1988
;
260
:
2859
–63
26
Cohen MM, Duncan PG, Tweed AW, et al. The Canadian four-centre study of anaesthetic outcomes. I. Description of methods and populations.
Can J Anaesth
 
1992
;
39
:
420
–9
27
Cooper AL, Leigh JM, Tring IC. Admissions to the intensive care unit after complications of anaesthetic techniques over 10 years.
Anaesthesia
 
1989
;
44
:
953
–8
28
Cooper JB, Long CD, Newbower RS, Philip JH. Critical incidents associated with intraoperative exchanges of anesthesia personnel.
Anesthesiology
 
1982
;
56
:
456
–61
29
Cooper JB, Newbower RS, Kitz RJ. An analysis of major errors and equipment failures in anesthesia management: considerations for prevention and detection.
Anesthesiology
 
1984
;
60
:
34
–42
30
Cooper JB, Newbower RS, Long CD, McPeek B. Preventable anesthesia mishaps: a study of human factors.
Anesthesiology
 
1978
;
49
:
399
–406
31
Dahlgren N, Tornebrandt K. Neurological complications after anaesthesia. A follow-up of 18,000 spinal and epidural anaesthetics performed over three years.
Acta Anaesthesiol Scand
 
1995
;
39
:
872
–80
32
Davis NJ (ed). Anaesthesia Related Mortality in Australia. 1994–1996. Melbourne: Australian and New Zealand College of Anaesthetists,
1999
33
Davis v Barking, Havering and Brentwood Health Authority.
Med LR
 
1993
;
4
:
85
–91
34
Derrington MC. Critical incidents in anaesthesia. In: Walker JS, ed. Quality and safety in anaesthesia. London: BMJ Publishing Group
1994
; 105–28
35
Domino KB, Posner KL, Caplan RA, Cheney FW. Awareness during anesthesia: a closed claim analysis.
Anesthesiology
 
1999
90
:
1053
–61
36
Dornette WHL, Orth OS. Death in the operating room.
Anesth Analg
 
1956
;
35
:
545
–51
37
Dripps RD, Lamont A, Eckenhoff JE. The role of anesthesia in surgical mortality.
JAMA
 
1961
;
178
:
261
–6
38
Endler GC, Marion FG, Sokol RJ, Stevenson LB. Anesthesia-related maternal mortality in Michigan, 1972–1984.
Am J Obstet Gynecol
 
1988
;
159
:
187
–93
39
Fisher MM, Baldo BA. Anaphylactoid reactions during anaesthesia.
Clin Anaesthesiol
 
1984
;
2
:
677
–92
40
Flanagan JC. The critical incident technique.
Psychol Bull
 
1954
;
51
:
327
–58
41
Fletcher IR, Healy TEJ. The arterial tourniquet.
J R Coll Surg Engl
 
1983
;
65
:
409
–17
42
Forrest JB, Cahalan MK, Rehder K, et al. Multicenter study of general anesthesia. II. Results.
Anesthesiology
 
1990
;
72
:
262
–8
43
Forrest JB, Rehder K, Cahalan MK, Goldsmith CH. Multicenter study of general anesthesia. III. Predictors of severe perioperative adverse outcomes.
Anesthesiology
 
1992
;
76
:
3
–15
44
Gebbie D. Anaesthesia and death.
Can Anaesth Soc J
 
1966
;
13
:
390
–6
45
Gild WM, Posner KL, Caplan RA, Cheney FW. Eye injuries associated with anesthesia.
Anesthesiology
 
1992
;
76
:
204
–8
46
Greif R, Akea O, Horn EP, Kurz A, Sessler DI. Supplemental perioperative oxygen to reduce the incidence of surgical wound infection. Outcomes Research Group.
N Engl J Med
 
2000
;
342
:
161
–7
47
Harrison GG. Anaesthetic contributory death—its incidence and causes. I. Incidence.
S Afr Med J
 
1968
;
42
:
514
–8
48
Harrison GG. Death attributable to anesthesia: a 10 year survey (1967–1976).
Br J Anaesth
 
1978
;
50
:
1041
–6
49
Holland R, Webb RK, Runciman WB. The Australian Incident Monitoring Study. Oesophageal intubation: an analysis of
2000
incident reports.
Anaesth Intensive Care
 
1993
;
21
:
608
–10
50
Holland R. Anaesthetic mortality in New South Wales.
Br J Anaesth
 
1987
;
59
:
834
–41
51
Holland R. Symposium—The Australian Incident Monitoring Study.
Anaesth Intensive Care
 
1993
;
21
:
501
52
Horan BF, Warden JC, Dwyer B. Urgent non-emergency surgery and death attributable to anaesthetic factors.
Anaesth Intensive Care
 
1996
;
24
:
694
–8
53
Hospital ‘neglect’ led to boy's death in routine surgery. The Times, 20 May
2003
, p. 9
54
Hovi-Viander M. Death associated with anaesthesia in Finland.
Br J Anaesth
 
1980
;
52
:
483
–9
55
Kawashima Y, Takahashi S, Suzuki M et al. Anaesthesia-related mortality and morbidity over a 5-year period in 2,363,038 patients in Japan.
Acta Anaesthesiol Scand
 
2003
;
47
:
809
–17
56
Keenan RL, Boyan CP. Cardiac arrest due to anesthesia: a study of incidence and causes.
JAMA
 
1985
;
253
:
2373
–7
57
Keenan RL, Boyan CP. Decreasing frequency of anesthetic cardiac arrests.
J Clin Anesth
 
1991
;
3
:
354
–7
58
Kluger MT, Short TG. Aspiration during anaesthesia: a review of 133 cases from the Australian Anaesthetic Incident Monitoring Study (AIMS).
Anaesthesia
 
1999
;
54
:
19
–26
59
Kroll DA, Caplan RA, Posner K, Ward RJ, Cheney FW. Nerve injury associated with anesthesia.
Anesthesiology
 
1990
;
73
:
202
–7
60
Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of surgical wound infection and shorten hospitalization. Study of Wound Infection and Temperature Group.
N Engl J Med
 
1996
;
334
:
1209
–15
61
Laxenaire MC. Drugs and other agents involved in anaphylactic shock occurring during anaesthesia. A French multicenter epidemiological inquiry.
Ann Fr Anesth Réanim
 
1993
;
12
:
91
–6
62
Lockhart PB, Feldbau EV, Gabel RS, Connolly SF, Silversin JB. Dental complications during and after tracheal intubation.
J Am Dent Assoc
 
1986
;
112
:
480
–3
63
Lunn JN, Mushin WW. Mortality Associated with Anaesthesia. London: Nuffield Provincial Hospitals Trust,
1982
64
Lyons G, Macdonald R. Awareness during Caesarean section.
Anaesthesia
 
1991
;
46
:
62
–4
65
Marx GF, Matteo CV, Otkin LR. Computer analysis of post anesthetic deaths.
Anesthesiology
 
1973
;
39
:
54
–8
66
McCrirrick A, Evans GH, Thomas TA. Overpressure isoflurane at Caesarean section: a study of arterial isoflurane concentrations.
Br J Anaesth
 
1994
;
72
:
122
–4
67
McKenzie AG. Mortality associated with anaesthesia at Zimbabwean teaching hospitals.
S Afr Med J
 
1996
86
:
338
–42
68
McMenemin IM, Sissons GRJ, Brownridge P. Accidental subdural catheterization: radiological evidence of a possible mechanism for spinal cord damage.
Br J Anaesth
 
1992
;
69
:
417
–9
69
Memery HN. Anesthesia mortality in private practice.
JAMA
 
1965
;
194
:
1185
–8
70
Minuck M. Death in the operating room.
Can Anaesth Soc J
 
1967
;
14
:
197
–204
71
Morita K, Kawashima Y, Irita K, Iwao Y, Seo N, Tsuzaki K. Perioperative mortality and morbidity in the year 2000 in 520 certified training hospitals of Japanese Society of Anesthesiologists: with a special reference to age—report of Japanese Society of Anesthesiologists Committee on Operating Room Safety.
Masui
 
2002
;
51
:
1285
–96
72
Myles PS, Weitkamp B, Jones K, Melick J, Hensen S. Validity and reliability of a postoperative quality of recovery score: the QoR-40.
Br J Anaesth
 
2000
;
84
:
11
–5
73
Myles PS, Williams DL, Hendrata M, Anderson H, Weeks AM. Patient satisfaction after anaesthesia and surgery: results of a prospective survey of 10,811 patients.
Br J Anaesth
 
2000
;
84
:
6
–10
74
Nelson DA. Intraspinal therapy using methylprednisolone acetate. Twenty-three years of clinical controversy.
Spine
 
1993
;
18
:
278
–86
75
Newbower RS, Cooper JB, Long CD. Failure analysis—the human element. In: Gravenstein JS, Newbower RS, Ream AK, Smith NT, eds. Essential Non-invasive Monitoring in the Operating Room. New York: Grune and Stratton,
1980
; 269–81
76
Owens WD, Spitznagel EL. Anesthetic side effects and complications: an overview.
Int Anesthesiol Clin
 
1980
;
18
:
1
–9
77
Pedersen T. Complications and death following anaesthesia. A prospective study with special reference to the influence of patient-, anaesthesia-, and surgery-related risk factors.
Dan Med Bull
 
1994
;
41
:
319
–31
78
Phillips OC, Frazier TM, Graff TD, De Kornfeld TJ. The Baltimore Anesthesia Study Committee. A review of 1,024 postoperative deaths.
JAMA
 
1960
;
174
:
2015
–20
79
Posner KL. Closed claims project shows safety evolution. Anesthesia Patient Safety Foundation Newsletter
2001
80
Ravindran RS, Bond VK, Tasch MD, Gupta CD, Luerssen TG. Prolonged neural blockade following regional anesthesia with 2-chloroprocaine.
Anesth Analg
 
1980
;
59
:
447
–51
81
Reynolds F, Speedy HM. The subdural space: the third place to go astray.
Anaesthesia
 
1990
;
45
:
120
–3
82
Reynolds F. Damage to the conus medullaris following spinal anaesthesia.
Anaesthesia
 
2001
;
56
:
238
–47
83
Rigler ML, Drasner K, Krejcie TC, et al. Cauda equina syndrome after continuous spinal anesthesia.
Anesth Analg
 
1991
;
72
:
275
–81
84
Safren MA, Chapanis A. A critical incident study of hospital medication errors.
Hospitals
 
1960
;
34
:
32
–66
85
Sanazaro PJ, Williamson JW. Physician performance and its effect on patients: a classification based on reports by internists, surgeons, pediatricians and obstetricians.
Med Care
 
1970
;
8
:
299
–308
86
Sandin R, Nordström O. Awareness during total i.v. anaesthesia.
Br J Anaesth
 
1993
;
71
:
782
–7
87
Schapira M, Kepes ER, Hurwitt ES. An analysis of deaths in the operating room and within 24 hours of surgery.
Anesth Analg
 
1960
;
39
:
149
–52
88
Schmied H, Kurz A, Sessler DI, Kozek S, Reiter A. Mild hypothermia increases blood loss and transfusion requirements during total hip arthroplasty.
Lancet
 
1996
;
347
:
289
–92
89
Shanker KB, Palkar NV, Nishkala R. Paraplegia following epidural potassium chloride.
Anaesthesia
 
1985
;
40
:
45
–7
90
Smith PLC, Treasure T, Newman SP, et al. Cerebral consequences of cardiopulmonary bypass.
Lancet
 
1986
;
1
:
823
–5
91
Snow JC, Kripke BJ, Norton ML, Chandra P, Woodcome HA. Corneal injuries during general anesthesia.
Anesth Analg
 
1975
;
54
:
465
–8
92
Suspected Anaphylactic Reactions Associated with Anaesthesia. Report of the third working party. London: Association of Anaesthetists of Great Britain and Ireland,
2003
93
Szekely SM, Webb RK, Williamson JA, Russell WJ. The Australian Incident Monitoring Study. Problems related to the endotracheal tube: an analysis of 2000 incident reports.
Anaesth Intensive Care
 
1993
;
21
:
611
–6
94
Taylor v Worcester and District Health Authority.
Med LR
 
1991
;
2
:
215
–31
95
Terry HR, Kearns TP, Love JG, Orwoll G. Untoward ophthalmic and neurologic events of anesthesia.
Surg Clin North Am
 
1965
;
45
:
927
–9
96
Tikkanen J, Hovi-Viander M. Death associated with anaesthesia and surgery in Finland in 1986 compared to 1975.
Acta Anaesthesiol Scand
 
1995
;
39
:
262
–7
97
Timperley AJ, Bracey DJ. Cardiac arrest following the use of hydrogen peroxide during arthroplasty.
J Arthroplasty
 
1989
;
4
:
369
–70
98
Tiret L, Desmonts JM, Hatton F, Vourc'h G. Complications associated with anaesthesia—a prospective survey in France.
Can Anaesth Soc J
 
1986
;
33
:
336
–44
99
Walsh EM. Steroid epidurals for low back pain and sciatica: still safe and effective after all these years?
J Pain Soc
 
1992
;
10
:
21
–4
100
Warden JC, Horan BF. Deaths attributed to anaesthesia in New South Wales, 1984–1990.
Anaesth Intensive Care
 
1996
;
24
:
66
–73
101
Waterson T. A critical incident study in child health.
Med Educ
 
1988
;
22
:
27
–31
102
Webb RK, Currie M, Morgan CA, et al. The Australian Incident Monitoring Study: an analysis of 2000 incident reports.
Anaesth Intensive Care
 
1993
;
21
:
520
–8
103
Willis RJ, ed. A Review of Anaesthesia Related Mortality 1997–1999. Melbourne: Australian and New Zealand College of Anaesthetists, 2002 http://www.anzca.edu.au/publications/reports/mortality/

Comments

1 Comment
Anaesthetic Errors, Australia, 2005
10 December 2008
John S Wright

In 2005, according to The Sydney Morning Herald health writer and a leading medical indemnifier, the following was the situation:

1. In New South Wales, for deaths to be attributed to anaesthesia, they must occur within 24 hours of its commencement. This definition has been repeatedly questioned for being "protective" of anaesthetists at the expense of others.

2. One in three anaesthetists in Australia has a "significant mishap" every year. Of those, one in five leads to litigation.

3. Such mishaps affect anaesthetists three times as often as any other doctor.

Conflict of Interest:

None declared

Submitted on 10/12/2008 7:00 PM GMT