Suspected anaphylactic reactions associated with anaesthesia.

The AAGBI has published guidance on management of anaphylaxis during anaesthesia in 1990, 1995 and 2003. This 2008 update was necessary to disseminate new information.

Death or permanent disability from anaphylaxis in anaesthesia may be avoidable if the reaction is recognised early and managed optimally.

Recognition of anaphylaxis during anaesthesia is usually delayed because key features such as hypotension and bronchospasm more commonly have a different cause.

Initial management of anaphylaxis should follow the ABC approach. Adrenaline (epinephrine) is the most effective drug in anaphylaxis and should be given as early as possible.

If anaphylaxis is suspected during anaesthesia, it is the anaesthetist’s responsibility to ensure the patient is referred for investigation.

Serum mast cell tryptase levels may help the retrospective diagnosis of anaphylaxis: appropriate blood samples should be sent for analysis.

Specialist (allergist) knowledge is needed to interpret investigations for anaesthetic anaphylaxis, including sensitivity and specificity of each test used. Specialist (anaesthetist) knowledge is needed to recognise possible non-allergic causes for the ‘reaction’. Optimal investigation of suspected reactions is therefore more likely with the collaboration of both specialties.

Details of specialist centres for the investigation of suspected anaphylaxis during anaesthesia may be found on the AAGBI website http://www.aagbi.org.

Cases of anaphylaxis occurring during anaesthesia should be reported to the Medicines Control Agency and the AAGBI National Anaesthetic Anaphylaxis Database. Reports are more valuable if the diagnosis is recorded following specialist investigation of the reaction.

This guidance recommends that all Departments of Anaesthesia should identify a Consultant Anaesthetist who is Clinical Lead for anaesthetic anaphylaxis.

Introduction

The AAGBI published its first guidelines on suspected anaphylactic reactions in 1990. Subsequent revisions were published in 1995 and 2003. The current guidelines incorporate advice from a large number of clinical immunologists, allergists and anaesthetists throughout the UK. In common with the 1995 and the 2003 editions, this report is published jointly with the British Society for Allergy and Clinical Immunology (BSACI).

This document is intended to be concordant with, and complementary to, the 2007 Scandinavian Clinical Practice Guidelines, the 2008 Resuscitation Council UK guidelines and the BSACI guidelines: Investigation of suspected anaphylaxis during anaesthesia.

These guidelines apply only to suspected anaphylactic reactions associated with anaesthesia and it is presupposed that the patient is in the care of a trained anaesthetist.

To clarify the definitions used in allergy and anaphylaxis.

To review the epidemiology of anaesthesia-related anaphylaxis.

To provide advice on the recognition of anaesthetic anaphylaxis.

To make recommendations on the immediate management and initial investigation of suspected anaesthetic anaphylaxis.

To make recommendations concerning the further investigation of suspected anaesthetic anaphylaxis.

To assist anaesthetists in obtaining access to a specialist centre for comprehensive investigation of suspected anaesthetic anaphylaxis.

To assist anaesthetists to provide appropriate information to the specialist centre.

To make recommendations about the reporting and collection of data on anaesthetic anaphylaxis in Great Britain and Ireland.

Definitions

The term ‘anaphylaxis’ has been used for all types of acute life-threatening illness triggered by abnormal sensitivity (hypersensitivity) to a trigger agent, and for apparently spontaneous attacks with similar features (idiopathic anaphylaxis). This has made it difficult to define. The EAACI Nomenclature Committee proposed the following broad definition [1]:

Anaphylaxis is a severe, life-threatening, generalized or systemic hypersensitivity reaction.

Minor, localised or non-systemic reactions are outside the definition of anaphylaxis. Anaphylaxis may be divided into ‘allergic anaphylaxis’ and ‘non-allergic anaphylaxis’. The clinical features of allergic anaphylaxis and non-allergic anaphylaxis may be identical. The EAACI committee proposed the term ‘allergic anaphylaxis’ should be used only when the reaction is mediated by an immunological mechanism (such as IgE, IgG, or complement activation by immune complexes). An anaphylactic reaction mediated by IgE antibodies, such as to amoxicillin, is referred to as ‘IgE-mediated allergic anaphylaxis’.

The term ‘anaphylactoid’ reaction had been introduced for non-IgE-mediated anaphylactic reactions but the EAACI committee has recommended this term should no longer be used. This proposal has not been universally accepted. An authoritative recent American practice parameter [2] states: ‘Anaphylaxis is defined … as a condition caused by an IgE-mediated reaction [2]. Anaphylactoid reactions are defined as those reactions that produce the same clinical picture as anaphylaxis but are not IgE mediated.’ In these guidelines we will follow the European (EAACI) nomenclature.

Anaphylaxis is not a homogeneous process: the pathways, mediators, time course and response to treatment depend on the trigger agent, its route and rate of administration, the nature of the patient’s hypersensitivity and the state of health of the patient, including incidental pathology such as respiratory or cardiovascular disease and the effects of concomitant medication such as β-blockers and ACE inhibitors. Although anaphylaxis commonly involves respiratory, cutaneous and circulatory changes, variations such as shock with gastrointestinal disturbance or shock alone are possible. Alternatively, reactions may be fatal without significant shock except as the terminal event following respiratory arrest [3]. Angioedema and urticaria may be features of anaphylaxis but commonly result from mechanisms other than anaphylaxis.

Intravascular volume redistribution is an important component of anaphylactic shock. Cardiac output may be decreased as a result of reduced coronary artery perfusion pressure as well as impaired venous return. Local release of mediators may cause coronary artery spasm and there may be features of acute left or right ventricular failure. Myocardial ischaemia with ECG changes is expected within minutes of anaphylactic shock becoming severe.

Asphyxia may be due to upper airway occlusion caused by angioedema, or bronchospasm with mucus plugging of the lower airways; the latter most commonly occurs in patients taking daily treatment for asthma. Both these processes may occur simultaneously in patients reacting to foods, latex, β-lactam antibiotics or aspirin.

Anaphylaxis usually resolves in 2–8 h but secondary pathology arising from the reaction or its treatment may prolong this. Resolution is complete except when cerebral hypoxia at the peak of the reaction has caused significant brain damage, or when disordered clotting leads to bleeding.

Acknowledgements

The involvement of the British Society for Allergy and Clinical Immunology and the Clinical Immunology and Allergy section of the British Society for Immunology is gratefully acknowledged. In addition, the working party wishes to acknowledge the assistance of the following:

Sr Alex Farragher Specialist Immunology Nurse

Mr Chris Hirst AAGBI Anaphylaxis website designer

Dr David Noble

Dr Martin Shields

Epidemiology

Geographical variation

Most reports on anaesthesia-related anaphylaxis originate from France, Australia, New Zealand and the United Kingdom. Other case series have been described from Scandinavia and the USA. The true incidence and their associated morbidity and mortality remain poorly defined. Both the accuracy and completeness of reporting is not optimal.

10% of anaesthesia-related reactions reported to the UK Medicines Control Agency (MCA) were fatal. These data should be interpreted with caution because it is likely that many less-severe reactions are not reported [1]. Reactions that are accepted as side-effects of certain drugs, for instance, histaminergic reactions due to atracurium and mivacurium, may be under-reported. During a time period of over 6 years only 361 reactions were reported to the MCA. In a 2-year period 789 reactions were reported in France in a comparable population where there is a well-established culture of reporting anaesthesia-related reactions [2].

Based on studies in Australia and France, the incidence of anaphylaxis during anaesthesia has been estimated at between 1 in 10 000 and 20 000 [3, 4] The true incidence of anaphylaxis during anaesthesia in the UK is not known. By extrapolating the French and Australian data to the UK, it is estimated that there are approximately 500 severe reactions in the UK each year. Anaphylactic reactions are more common when drugs are given intravenously.

In one large study an immune basis was demonstrated in two-thirds of patients investigated for anaphylaxis [2]: the remainder comprised non-allergic anaphylaxis and mechanisms other than anaphylaxis. The tests currently available for the diagnosis of anaesthetic anaphylaxis are imperfect. Skin tests and blood tests have limited sensitivity and specificity and their positive and negative predictive value varies between drugs.

Patient characteristics (age, sex, ethnicity, smoking etc)

Neuromuscular blocking drugs and latex appear to cause anaphylaxis more commonly in female patients. There appears to be a connection between smoking and antibiotic anaphylaxis, possibly because smokers may become sensitised by exposure to repeated courses of antibiotics for respiratory tract infections. Individuals with a history of atopy, asthma or allergy to some foods appear to be at increased risk of latex allergy but not anaphylaxis to neuromuscular blocking drugs or antibiotics [2, 5]. Anaphylaxis associated with radiographic contrast media appears to be associated with atopy [6]. Patients with asthma or taking β-blocking drugs may suffer a more severe reaction. Some of these reactions may be refractory to conventional therapy.

Possible environmental sensitising agents

The prevalence of anaphylactic reactions to neuromuscular blocking agents (NMBAs) is reported to be at least six times more frequent in some countries [7, 8]: it is considerably more common in Norway than in Sweden. Quaternary ammonium ions (QAI) are proposed to be the allergenic epitopes in NMBAs. Common environmental chemicals such as toothpastes, washing detergents, shampoos, and cough medicines share these allergenic epitopes with the NMBAs [8]. Numerous possibilities exist for a predisposed individual to become sensitised to QAIs and thus be at risk of developing anaphylaxis to NMBAs during anaesthesia. In a recent Scandinavian survey, use of certain cough medicines was found to be the only significant difference in environmental chemical exposure [8]. In Norway, but not Sweden, cough syrups containing pholcodine are available without prescription. IgE-antibodies to pholcodine were seen in 6% of a general population (blood donors) in Norway but not in Sweden (0%).

Muscle relaxants

Approximately 60% of cases of anaesthesia-related anaphylaxis are thought on the basis of skin tests to be due to neuromuscular blocking agents. Mivacurium and atracurium are associated with non-allergic anaphylaxis in which release of mediators (notably histamine) from mast cells exactly imitates allergic anaphylaxis. Cisatracurium, although sharing a benzylisoquinolinium structure, is not associated with non-allergic anaphylaxis, although several cases of allergic anaphylaxis have been reported. It is generally accepted that succinylcholine is the NMBA most likely to be associated with allergic anaphylaxis, although rocuronium has been implicated in a similar number of cases in France.

The prevalence of sensitisation to NMBAs in the community is higher than the incidence of reactions would suggest. In one study, approaching 10% of the general population exhibited skin reactivity to NMBAs, an incidence that far exceeds the incidence of anaphylaxis on administration of these drugs during anaesthesia [9]. A previous history of specific drug exposure is not necessary, particularly for neuromuscular blocking drugs. A history of previous exposure is found in fewer than 50% of patients who are allergic to neuromuscular blocking drugs. Conversely, an uneventful exposure may sensitise an individual to subsequent administration of the drug.

Cross-sensitivity between different NMBAs is relatively common, probably because they share a quaternary ammonium epitope. If anaphylaxis to an NMBA is suspected, the patient should undergo skin prick testing with all the NMBAs in current use. If a patient demonstrates a positive skin prick test (SPT) to an NMBA, the patient should be warned against future exposure to all NMBAs if possible. If it is mandatory to use an NMBA during anaesthesia in the future, it would seem appropriate to permit the use of an NMBA which has a negative skin test, accepting that a negative skin test does not guarantee that anaphylaxis will not occur.

The apparent excess of cases of anaphylaxis to rocuronium in some countries should be interpreted with caution until much more data have been collected. Large sample sizes are needed to estimate the true incidence: if the true incidence is 1 in 5000, a sample size of 7 million would be needed to have a 95% chance of being within 5% of the true value [10]. There is a need for further epidemiological studies.

Latex

Latex hypersensitivity is the second most common cause of anaesthesia-related anaphylaxis in many studies (up to 20% of cases). However, the incidence of anaphylaxis to latex is considerably less than would be suggested by the prevalence of positive skin tests or positive IgE tests. It appears that the incidence may be waning, at least in some countries; possibly as a result of a change or decline in the use of latex gloves. Certain patient groups are more susceptible to latex anaphylaxis (see Appendix II).

Antibiotics

Approximately 15% of anaesthesia-related anaphylactic episodes are due to antibiotics. This proportion has increased in recent years and may merely mirror increased exposure to antibiotics in the community [2, 11]. The pre-operative history is important. Although only a minority of patients who report allergy to antibiotics have a true allergy, the consequence of anaphylaxis to intravenous antibiotics may be catastrophic, and self-reporting should be taken seriously. Skin testing is only approximately 60% predictive of clinical hypersensitivity. Penicillins and cephalosporins which share the β-lactam ring are responsible for approximately 70% of antibiotic-induced anaphylaxis. Many antibiotics possess a β-lactam ring:

Benzylpenicillin, phenoxymethylpenicillin

Flucloxacillin, temocillin

Amoxicillin, ampicillin, co-amoxiclav

Co-fluampicil

Piperacillin, ticarcillin,

Pivmecillinam HCl

Cephalosporins

Aztreonam

Ertapenem, imipenem with cilastatin, meropenem

The structure of the side chains attached to the β-lactam ring is also important in determining the response of the immune system. First generation cephalosporins and cefamandole share a similar side chain with penicillin and amoxicillin. A recent meta-analysis suggested that patients who are allergic to penicillin or amoxicillin have a higher incidence of allergic reactions to first generation cephalosporins and cefamandole, but not other cephalosporins [12]. The issue is complicated because the classification of cephalosporins relates to their antimicrobial activity rather than their chemical structure.

Cefalexin (cephalexin)*

Cephaloridine

Cephalothin

Cefazolin

Cefradine (cephradine)*

Cefadroxil *

Cefaclor*

Cefamandole

Cefuroxime*

Cefixime*

Cefotaxime*

Cefpodoxime*

Ceftazidime*

Ceftriaxone*

Cefepime

Cefpirome

Fourth generation cephalosporins

*Listed in the British National Formulary.

Local anaesthetics

Anaphylactic reactions to local anaesthetic drugs are very uncommon. Local anaesthetic esters are more likely than amides to provoke a Type lV allergic reaction. Preservatives such as methyl-paraben or metabisulphites may be responsible in some cases. It has been suggested that inadvertent intravascular injection of a local anaesthetic or the systemic absorption of adrenaline may be responsible for many of the reported reactions. Reactions occurring in the dental chair may also be associated with idiopathic angioedema or latex allergy.

Opioids

Opioids are an uncommon cause of anaesthesia-related anaphylaxis. Diagnosis is difficult and the true incidence is unknown. Morphine, pethidine and codeine are well-known to cause non-specific histamine release which precludes diagnostic skin testing. The diagnosis of opioid anaphylaxis often rests on a careful history and the exclusion of other possibilities. Challenge testing may be appropriate in some cases but may be performed only in specialist centres.

Anaesthetic induction agents

Anaphylaxis to propofol is very uncommon. The antigenic determinant may be the isopropyl groups. It has been stated that patients with egg allergy or soy allergy should avoid propofol but there is no strongly-supportive evidence. It has been suggested that propofol anaphylaxis is more likely if lignocaine is added to reduce pain on injection: there is no evidence to support this suggestion. Anaphylaxis to thiopental has become extremely uncommon, probably reflecting the decline in its use. A small number of cases of midazolam anaphylaxis have been reported.

Non-steroidal anti-inflammatory drugs (NSAIDs)

Several mechanisms may be responsible for reactions to NSAIDs. Inhibition of the PGE2 pathway leads to excessive leukotriene synthesis and subsequent mediator release, causing urticaria or bronchospasm. IgE-mediated reactions may also occur in relation to some NSAIDs. Fatal anaphylaxis has been described after oral administration of NSAIDs.

Halogenated volatile anaesthetics

There are no published reports of anaphylaxis to halogenated volatile anaesthetics. The rare fulminant form of hepatitis associated with halothane has an immune basis which is unrelated to anaphylaxis.

Colloids

Intravenous colloids are responsible for approximately 4% of all peri-operative anaphylactic reactions. In one study, gelatin solutions were responsible for 95% of the intravenous colloid reactions. It has been stated that the incidence may be greater with the urea-linked gelatins compared with the modified fluid gelatins. Intravenous gelatin solutions should be avoided in patients with a history of allergy to gelatin-containing vaccines. Anaphylaxis to intravenous dextrans has been reported, with an incidence similar to the modified fluid gelatins. Anaphylaxis to hydroxyethyl-starch is rare.

Antiseptics and disinfectants

Reactions to chlorhexidine have come into greater prominence in recent years. There appears to be a significant difference in incidence between countries [13]. Reactions range from contact dermatitis to life-threatening anaphylaxis. Anaphylaxis has occurred when chlorhexidine was used as an antiseptic for urological and gynaecological procedures as well as insertion of central venous and epidural catheters. The chlorhexidine coating of certain central venous catheters has been implicated in such reactions. It is prudent to allow skin disinfectant to completely dry before beginning an invasive procedure. Anaphylaxis to polyvinylpyrrholidine as povidone-iodine or as an excipient for oral medicines occurs but is rare.

Miscellaneous agents

Many agents to which patients may be exposed during anaesthesia may be associated with anaphylaxis, including aprotinin, protamine, heparins, radiological contrast material, dyes and oxytocin. Anaphylaxis to glycopyrronium and neostigmine may occur very rarely.

Recognition and clinical diagnosis of anaphylaxis

In a closely monitored patient such as during anaesthesia, when a fall in blood pressure, change in heart rate or difficulty with ventilation are noticed, anaphylaxis is so rarely the cause that inevitably treatment is given for another diagnosis before anaphylaxis is recognised; the response to this treatment has commonly delayed or even prevented recognition of the true cause. It is therefore probable that many grade I and II acute allergic reactions to anaesthetic drugs are missed. Conversely, a cause other than allergic anaphylaxis seems more likely in around one-third of the patients referred to specialist centres for investigation of suspected anaphylaxis during anaesthesia.

Previous history

Any clue that appropriately raises anticipation of anaphylaxis will be helpful. A previous history of reaction to anaesthetic drugs, antibiotics, other drugs, chlorhexidine or latex should obviously lead to appropriate avoidance. Because it is always possible the wrong trigger factor was identified, the exact details of exposure leading to any previous reaction should be sought during the pre-operative assessment. Cross-reactivity between non-anaesthetic drugs such as pholcodine or environmental chemicals containing quaternary ammonium groups such as cosmetics and detergents has been proposed as the source of sensitivity to opioids and muscle relaxants: a history of cutaneous sensitivity to cosmetics or rashes from cough medicines should raise caution.

Anaphylaxis to amoxicillin and cephalosporins is commonest in asthmatic smokers who have had multiple courses of these antibiotics without reacting: the symptoms of anaphylaxis in such a patient are likely to be initially misinterpreted (quite reasonably) as those expected from anaesthesia in an asthmatic smoker.

Clinical features

Although anaesthesia-related anaphylaxis usually results from intravenous drug administration, administration by other routes, for example cutaneous, mucosal/vesical peritoneal, intra-articular or intramuscular may be responsible. Clinical features include hypotension, tachycardia or bradycardia; cutaneous flushing, rash or urticaria; bronchospasm; hypoxia; angioedema and cardiac arrest. If an adverse event such as hypotension or bronchospasm occurs during anaesthesia it is appropriate to suspect anaphylaxis unless there exists a significantly more likely cause.

Tachycardia is not invariable: bradycardia is seen in approximately 10% of patients with allergic anaphylaxis during anaesthesia. Hypotension is the sole clinical feature in approximately 10% of patients and is likely to be exaggerated in patients undergoing anaesthesia with neuraxial blockade. Widespread flushing or urticaria is seen in the majority of patients but the absence of cutaneous signs does not exclude anaphylaxis. Bronchospasm may be more common in patients with pre-existing asthma.

The clinical features usually occur within a few minutes but may be delayed by up to an hour. Substances to which the clinical reaction may be delayed include latex, antibiotics, intravenous colloids and Cidex OPA (used to disinfect surgical instruments). However, an immediate response does not exonerate these substances. The clinical features of anaphylaxis to neuromuscular blocking agents usually develop rapidly. Deflation of a surgical tourniquet may induce anaphylaxis if the allergen has been sequestered in the limb.

Immediate management

These management guidelines presuppose that the patient is in the care of an appropriately-trained anaesthetist and full resuscitation facilities and appropriate vital signs monitors are available.

There is a wide spectrum of severity and combinations of clinical features. Although management should be tailored to the individual patient, there is consensus that adrenaline should be given as early as possible. In addition to having alpha-agonist activity, adrenaline is a valuable beta-agonist which is inotropic and a bronchodilator, and reduces further mediator release.

Other causes of hypotension or difficulty in ventilation should be excluded, for example a misplaced tracheal tube or equipment failure.

Use the ABC approach (Airway, Breathing, Circulation). Team-working enables several tasks to be accomplished simultaneously.

Remove all potential causative agents (including IV colloids, latex and chlorhexidine) and maintain anaesthesia, if necessary, with an inhalational agent.

Call for help and note the time.

Maintain the airway and administer oxygen 100%. Intubate the trachea if necessary and ventilate the lungs with oxygen.

Elevate the patient’s legs if there is hypotension.

If appropriate, start cardiopulmonary resuscitation immediately according to Advanced Life Support Guidelines.

Administer adrenaline intravenously. An initial dose of 50 μg (0.5 ml of 1 : 10 000 solution) is appropriate (adult dose). Several doses may be required if there is severe hypotension or bronchospasm.

If several doses of adrenaline are required, consider starting an intravenous infusion of adrenaline (adrenaline has a short half-life).

Administer saline 0.9% or lactated Ringer’s solution at a high rate via an intravenous cannula of an appropriate gauge (large volumes may be required).

Administer chlorphenamine 10 mg IV (adult dose).

Administer hydrocortisone 200 mg IV (adult dose).

If the blood pressure does not recover despite an adrenaline infusion, consider the administration of an alternative intravenous vasopressor according to the training and experience of the anaesthetist, for example metaraminol.

Treat persistent bronchospasm with an intravenous infusion of salbutamol. If a suitable breathing-system connector is available, a metered-dose inhaler may be appropriate. Consider giving intravenous aminophylline or magnesium sulphate.

Arrange transfer of the patient to an appropriate Critical Care area.

Take blood samples (5–10 ml clotted blood) for Mast Cell Tryptase as follows:

Initial sample as soon as feasible after resuscitation has started – do not delay resuscitation to take the sample.

Second sample at 1–2 h after the start of symptoms.

Third sample either at 24 h or in convalescence (for example in a follow-up allergy clinic). This is a measure of baseline tryptase levels as some individuals have a higher baseline level.

Ensure that the samples are labelled with the time and date.

Liaise with the hospital laboratory (see Appendix lll: Mast cell Tryptase).

Drug doses in children

Adrenaline

Intramuscular

> 12 years: 500 μg IM (0.5 ml of a 1 : 1000 solution)

300 μg IM (0.3 ml of a 1 : 1000 solution) if the child is small

6–12 years: 300 μg IM (0.3 ml of a 1 : 1000 solution)

Up to 6 years: 150 μg IM (0.15 ml of a 1 : 1000 solution)

Intravenous

Intravenous adrenaline may be used in children in acute areas such as operating theatres or intensive care units by those familiar with its use and if IV access is already available. Great care should be taken to avoid dose errors when preparing drug dilutions. Prepare a syringe containing 1 ml of 1 : 10 000 adrenaline for each 10 kg body weight (0.1 ml.kg−1 of 1 : 10 000 adrenaline solution = 10 μg.kg−1. Titrate to response, starting with a dose of one-tenth of the contents of the syringe, i.e. 1 μg.kg−1. Often a child will respond to as little as 1 μg.kg−1. In smaller children, further dilution may be needed to allow dose titration (check carefully for decimal point and concentration errors). The intramuscular route is preferred where there is no venous access or where establishing venous access would cause a delay in drug administration.

Hydrocortisone

> 12 years: 200 mg IM or IV slowly

6 to 12 years: 100 mg IM or IV slowly

6 months to 6 years: 50 mg IM or IV slowly

< 6 months: 25 mg IM or IV slowly

Chlorphenamine

> 12 years: 10 mg IM or IV slowly

6 to 12 years: 5 mg IM or IV slowly

6 months to 6 years: 2.5 mg IM or IV slowly

< 6 months: 250 μg.kg−1 IM or IV slowly

Later investigations to identify the causative agent

Any patient who has a suspected anaphylactic reaction associated with anaesthesia should be investigated fully, but investigations should not interfere with the immediate treatment of the patient. The anaesthetist who administered the anaesthetic or the consultant anaesthetist in charge of the patient is responsible for ensuring that the reaction is investigated.

The patient should be referred to a specialist Allergy or Immunology centre that has appropriate experience of investigating this type of problem. Anaesthetic departments should identify a lead anaesthetist for anaesthetic anaphylaxis. Referral pathways should be agreed prospectively with the appropriate specialist centre.

Criteria for referral to a specialist centre for investigation

Patients in whom allergic or non-allergic anaphylaxis is suspected should be referred to a specialist clinic for investigation. A patient should be referred if there is any of the following:

Unexplained cardiac arrest during anaesthesia.

Unexplained, unexpected hypotension (for example a decrease of mean arterial pressure of more than 30 mmHg) which requires active treatment.

Unexplained, unexpected bronchospasm, particularly if the bronchospasm is severe, causes a significant decrease in oxygen saturation and is relatively resistant to treatment.

Widespread rash, flushing or urticaria.

Angioedema.

A detailed analysis of events surrounding the suspected anaphylactic reaction must be undertaken. The time of onset of the reaction in relation to induction and other events is the most important information. All drugs and other agents to which the patient was exposed before and during the anaesthetic as well as their timing in relation to the reaction must be recorded.

Details sent to the specialist centre along with a letter of referral must include:

A legible photocopy of the anaesthetic record

A legible photocopy of the recovery room chart

Legible photocopies of drug charts

A description of the reaction and time of onset in relation to induction

Details of blood tests sent and their timing in relation to the reaction

Contact details of the surgeon and the general practitioner

A standard referral proforma is useful (see Appendix S1 in Supporting Information).

Tests performed at the specialist centre

Skin tests

Patients should be referred for skin testing as soon as possible after the clinical event. Skin testing can be performed as soon as the patient has made a full clinical recovery and the effects of any antihistamine given to treat the reaction have fully worn off. Skin tests can provide confirmation of sensitisation to a specific drug but must be interpreted within the clinical context. There are a variety of techniques and several different criteria are used for positivity [1-4] so skin testing must be performed by suitably trained and experienced personnel.

A histamine solution and physiological saline are used as positive and negative controls respectively. Drugs with anti-histamine activity must be discontinued a few days before testing. There is no need to discontinue oral or inhaled steroids.

Skin prick tests are usually done on the volar surface of the forearm. A drop of the drug is placed on the skin and the skin pricked through the drop with a lancet. The results are read after 15–20 min.

Intradermal tests may be performed where skin prick tests are negative. Although they are more sensitive, they are less specific, more difficult to interpret and more likely to precipitate a systemic reaction. Intradermal tests are usually performed on the forearm or back. 0.02 to 0.05 ml of a dilute solution is injected intradermally and the results read after 20–30 min.

Both skin prick tests and intradermal tests are widely used in the diagnosis of IgE mediated allergic reactions though the diagnostic value of a positive test for most drugs is limited due to lack of subsequent challenge data. If the mechanism is unknown a negative test is unreliable.

All drugs used during anaesthesia may be tested and lists of non-irritant dilutions have been compiled for both skin prick tests and intradermal tests [5]. Skin tests are most useful for latex, beta lactam antibiotics and NMBAs. They are also useful for induction agents, protamine [6] and chlorhexidine. Intradermal testing may be more reliable for propofol [2]. Skin tests are not usually performed for opioids because false positive results are common. However, skin prick testing appears to be informative for the synthetic opioids fentanyl and remifentanil. Skin tests are not useful for NSAIDs, dextrans or iodinated radiological contrast media [6] because anaphylaxis to these agents is not usually IgE-mediated.

Patch tests are the mainstay of diagnosis for contact allergic reactions and may be useful for diagnosing exanthematous drug rashes but are not helpful in eliciting the cause of suspected anaphylactic reactions occurring during anaesthesia.

Blood tests

Appropriate further blood investigations depend on the drug in question, and the experience of the specialised laboratory service. Blood samples for specific IgE tests may be taken at the time of the reaction, or soon afterwards during that hospital admission. If the results are negative these tests should be repeated when the patient is seen at the specialist centre in case the initial results might be falsely negative due to possible consumption of relevant IgE antibodies during the reaction. The most commonly used test for allergen-specific IgE uses target allergen or drug bound onto a three-dimensional sponge-like solid matrix or ‘CAP’ and a fluorescent detection system. The Radio-AllergoSorbent Test (RAST) uses a radioactive detection system: it is now rarely used but the name has persisted in common use.

Specific IgE antibodies against succinylcholine (thiocholine ester) can be assayed in serum but the sensitivity is relatively poor (30–60%). Tests for serum IgE antibodies against other neuromuscular blocking drugs are not available in the UK. Specific IgE antibodies against several antibiotics can be assayed. These include amoxicilloyl, ampicilloyl, penicilloyl G, penicilloyl V and cefaclor, however the diagnostic value of these tests is less well defined. Tests for specific antibodies against latex, chlorhexidine and bovine gelatin are available.

The presence of drug-specific IgE in serum is evidence of allergic sensitisation in that individual and provides a possible explanation of the mechanism and specific drug responsible for the reaction. It is not in itself proof that the drug is responsible for the reaction. It is important to emphasise that attribution of cause and effect is a judgement made after considering all the clinical and laboratory information relevant to the reaction.

Experimental tests

There is much research into ways of improving the in vitro elucidation of anaphylaxis. In addition to tissue-bound mast cells, circulating basophils are also involved in anaphylaxis. During anaphylaxis, proteins such as CD63 and CD203c become newly or increasingly expressed on the surface of basophils. These can be detected by flow cytometry which forms the basis of experimental drug-induced basophil stimulation tests.

Clinical features	Allergic anaphylaxis (n = 518), n (%)	Non-allergic anaphylaxis (n = 271), n (%)
Cardiovascular	387 (74.7)	92 (33.9)
Arterial hypotension	90 (17.3)	50 (18.4)
Cardiovascular collapse	264 (50.8)	30 (11.1)
Bradycardia	7 (1.3)	2 (0.7)
Cardiac arrest	31 (5.9)	–
Bronchospasm	207 (39.8)	52 (19.2)
Cutaneous signs	374 (71.9)	254 (93.7)
Angioedema	64 (12.3)	21 (7.7)

Clinical features of allergic anaphylaxis and non-allergic anaphylaxis occurring during anaesthesia in France between 1st Jan 1999 and 31st Dec 2000. Mertes PM et al.Anesthesiology 2003; 99: 536–45.