Accurately diagnosing antibiotic allergies

Drug allergy, as defined by the British Society for Allergy and Clinical Immunology, is an adverse drug reaction with an established immunological mechanism^[1]
. All forms of natural and semisynthetic penicillins or drugs with a similar structure, such as cephalosporins or carbapenems, can cause allergy. These drugs, which have a beta-lactam ring, are recognised as one of the most frequent causes of immediate (within 1 hour) and delayed (after 72 hours) drug reactions. 

Correct diagnosis can be difficult, but true allergy is more likely if the patient has recognised allergic symptoms (e.g. rash, rash involving hives, wheezing, or swelling of the skin or throat), or has previously experienced a similar reaction to the same agent or another agent in the same class. Gastrointestinal symptoms alone do not represent a true allergy and alternative biological or pharmacological explanation should be excluded before a diagnosis of allergy is made^[2]
.

Adverse reactions to penicillin have been reported in 0.2–5.0% of individuals per course of treatment^[3]
. However, fewer than 10% of those believed to be allergic to penicillin are truly allergic^[2]
. As a result, antibiotics can be withheld unnecessarily, affecting clinical outcomes, increasing healthcare costs and potentially contributing to the development of drug-resistant bacteria^[3]
. Accurate diagnosis and antimicrobial stewardship practices are therefore key^[2]
.

Delabelling — the removal of inappropriate patient allergy labels — allows first-line treatment options to be used. These are often more effective antibiotics with fewer side effects, have a narrower spectrum of activity and are more cost efficient^[4]
. Using effective narrow-spectrum agents where possible is one of the major aims of antimicrobial stewardship and ensures the right drug is being used for the right patient. Current practice of delabelling allergy status in the UK varies and may depend on whether there is an established local allergy service or whether pharmacists have the capacity to assess patients’ allergies on admission to hospital.

Pharmacists, working as part of the wider multidisciplinary team, are able to effectively manage infections in patients with antibiotic allergies by delabelling their inappropriate allergy status on admission^[5]
(see Box 1). The process of optimising medication on admission to hospital currently falls under the remit of doctors and pharmacists^[6]
.

This article describes the possible consequences of labelling a patient with an antibiotic allergy; the subsequent impact on patient outcomes; how pharmacists can undertake appropriate investigations and delabel patients; and when to refer patients to specialist allergy services. This article complements a previous article on the identification and management of penicillin allergy^[3]
.

Mechanisms of action and prevalence

Beta-lactams

Allergic reactions to beta-lactam antibiotics can be attributed to their drug chemistry and metabolism. The beta-lactam ring (thiazolidine ring in penicillins and dihydrothiazine ring in cephalosporins) and the side chains are immunogenic (see Figure 1). Drug metabolites containing these components act as haptens (incomplete antigens) that can covalently bind to larger proteins to elicit an immune response, resulting in the production of immunoglobulin E (IgE) antibodies^[7]
.

Once administered, benzylpenicillin will spontaneously undergo degradation to form reactive intermediate products, such as benzylpenicilloyl, the major antigenic determinant to which most patients with allergies react. This metabolite binds to proteins and is used in commercially available skin testing kits in the UK. The remaining penicillin metabolites can also act as haptens (called minor determinants) and account for around 16% of allergic reactions^[7]
. Beta-lactam allergy has been identified as a possible independent risk factor for other classes of antibiotic allergy^[8]
.

Non-beta-lactams

Allergic reactions to non-beta-lactam antibiotics occur less frequently and IgE-mediated anaphylaxis is rare^[9]
.

Anaphylaxis to fluoroquinolones is reported to occur with an incidence of 1.8–2.3 per 10 million treatment days and accounts for 4.5% of all drug-induced anaphylaxis^[8],[10]
. Moxifloxicin is associated with more severe allergic reactions than ciprofloxacin and levofloxacin^[10]
. Up to 70% of allergies reported with moxifloxacin are immediate hypersensitivity reactions^[8]
; it is restricted for use by the Medicines and Healthcare products Regulatory Agency owing to an association with non-allergic, life-threatening hepatic reactions^[11]
. Ciprofloxacin is most commonly associated with delayed skin reactions^[8]
. 

Sulphonamide allergy rates are reported at around 2% and anaphylaxis is rare, but patients with HIV/AIDS have an increased risk of a reaction^[10]
. They are associated with more severe life-threatening reactions, such as drug rash with eosinophilia and systemic symptoms (DRESS), which typically appear between two and ten weeks after initiation^[10]
.

Vancomycin is associated with the pseudo-allergy ‘red man syndrome’ (RMS), an IgE-independent histamine reaction with a worldwide incidence rate of 3.7–47%. Slowing infusion rates to 500mg/hour greatly reduces the risk of RMS. Anaphylaxis is possible but rare, and DRESS, Stevens–Johnson syndrome and toxic epidermal necrolysis have also been observed. Anaphylaxis to teicoplanin has been reported at an incidence of 1 in around 1,600 perioperative patients^[12]
. 

Macrolide IgE-mediated hypersensitivity reactions are drug-specific, not class-specific, and are uncommon, occuring in 0.4–3% of patients treated^[10]
. Allergic reactions to tetracyclines are rare, although more common with minocycline. They are associated with anaphylaxis and life-threatening skin reactions (e.g. DRESS)^[10]
. Drug-induced lupus has been reported with minocycline and may appear two to six years after initiation^[13]
. 

Aminoglycoside allergic reactions are extremely rare, although case reports of anaphylaxis and contact dermatitis have been reported with topical preparations^[10]
.

Perioperative anaphylaxis

The recent National Audit Project on perioperative anaphylaxis declared that antibiotics were the main cause of anaphylaxis in the perioperative period in the UK, accounting for 46% of reported cases with an identified causal agent^[12]
. The authors found the incidence of antibiotic anaphylaxis to be 4 per 100,000 administrations, with teicoplanin, co-amoxiclav and vancomycin being the most prevalent. The most common clinical feature of antibiotic anaphylaxis was hypotension, occurring in 42% of cases, and the onset of anaphylaxis was within 5 minutes in 74% of cases, within 10 minutes in 92% of cases, and in all cases anaphylaxis occurred within 30 minutes after drug exposure^[12]
.

Disappearance of allergies

Avoidance of the drug that initially triggered anaphylaxis results in reduced IgE antibody concentrations that can disappear completely over time^[14]
. Around 80% of patients with a true allergy to penicillin, confirmed with skin testing and who avoided penicillin after diagnosis, were subsequently found to have a negative skin test after 10 years^[15]
. Resensitisation to penicillin is possible following re-exposure, but rates are very low, with levels similar to patients with no previous allergy^[16]
. However, patients with severe delayed allergic reactions (e.g. Stevens–Johnson syndrome, toxic epidermal necrolysis and DRESS) should not receive the causative agent again, as the risk of these reactions recurring does not decrease over time^[14],[15]
.

Penicillin allergy labels

Penicillin-based antibiotics are often used as first-line treatment for infections. Therefore, patients previously diagnosed as allergic to penicillin, whose notes are marked with ‘penicillin allergy’ labels, are usually prescribed non-penicillin agents that are often more costly, more toxic and may be less effective. See the National Institute for Health and Care Excellence’s (NICE) Clinical Guideline 183 for more information on its identification and management^[2]
.

Use of broad-spectrum penicillin alternatives in incorrectly labelled patients increases the risk of future infections with resistant bacteria, such as methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus and Clostridioides difficile (formerly known as Clostridium difficile)^{[17],[18],[19],[20],[21],[22],[23]}
. In addition, patients are exposed to a greater number of antibiotics, including quinolone and carbapenem agents, further contributing to selection pressures for resistance^{[17],[23],[24]}
.

Incorrectly labelled patients may also be denied appropriate treatment with newer beta-lactam derivatives unnecessarily. The joint Infectious Diseases Society of America and Society for Healthcare Epidemiology of America guideline on antimicrobial stewardship recommends the promotion of allergy assessment and testing, where appropriate, owing to the potential impact this could have on the utilisation of first-line agents^[25]
.

An email survey of 55 UK hospitals revealed there to be little evidence that hospitals are tackling the issue of incorrect penicillin allergy labelling^[26]
. This may be owing to the lack of allergy services available, meaning patients with incorrect penicillin allergy labels cannot be completely reviewed and confirmed by allergy services alone. However, many low-risk patients (e.g. those who only experience gastrointestinal symptoms or give a history of non-severe delayed skin reactions) can be safely delabelled without the need for skin allergy testing^[27],[28]
.

Pharmacists working with a multidisciplinary team to undertake allergy assessments, as outlined in Box 1, have been shown to reduce non-beta-lactam consumption and increase beta-lactam first-line therapy^[23],[29]
. There is a role for ward-based pharmacy technicians to contribute to this process during the medication reconciliation process.

Primary care pharmacists can ensure documentation of allergy status is as detailed as possible and highlight inappropriate allergy labelling to GPs. Community pharmacists can ensure patients understand the difference between true allergy and intolerance when discussing allergy status, as well as directing patients for update or assessment of allergy status.

Delabelling ‘penicillin allergic’ patients

Although delabelling programmes with pharmacist involvement have been successfully delivered in US hospitals, the introduction of penicillin allergy delabelling programmes is a complex intervention that requires changes in the behaviour of both healthcare professionals and patients^{[23],[30],[31]}
. Healthcare professionals need to feel confident using decision tools to delabel those incorrectly labelled as penicillin allergic and the process of delabelling needs to be supported by NHS management^[32]
. Fear of patient harm through lack of understanding around which allergy labels are inappropriate may prevent healthcare professionals delabelling inappropriate allergy status in patients^[32]
. Delabelled patients and parents of delabelled children need to trust the test and the advice that they are given and that it is safe to take penicillin in the future if deemed non-allergic. Patients and healthcare professionals may still avoid penicillin, even after the delabelling has taken place, if they are not confident in the delabelling process^[33]
. Many patients retain their allergy status in their medical notes despite a negative oral challenge test^[34]
.

Implementing a sustainable penicillin allergy delabelling service in NHS hospitals will require understanding and expertise regarding the influences on behaviour and human factors. A guide on how to achieve delabelling in the hospital setting is provided in Box 1.

This approach to penicillin allergy clarification and delabelling, without the use of skin allergy testing, is reported to delabel between 16% and 50% of patients, enabling the use of first-line penicillin therapy^{[19],[23],[29],[35],[36],[37],[38]}
.

The two case studies (see Panel 1) illustrate the approaches to managing penicillin allergy clarification and delabelling, and when skin allergy testing would be required.

Referral to specialist allergy services

NICE recommends that patients should be referred to allergy specialists, where available, if they have a possible allergy to beta-lactams and are likely to need antibacterial therapy in the future; for example, patients with recurrent infections (e.g. urinary tract infections or cystic fibrosis), patients at increased risk of recurrent infections (e.g. bronchiectasis), or patients generally at an increased risk of infection (e.g. if they have asplenia [absence of normal spleen function] or other immunodeficiencies) (see Box 2)^[2]
.

At time of writing, there are 33 allergy centres across the UK providing allergy services for adults, with 3 centres in Scotland, and 1 each in Wales and Northern Ireland^[12]
.

Available tests

Skin-prick testing, intradermal testing, and oral and intravenous drug challenges can be used for patients requiring further allergy testing. These tests are critical where the history describes serious allergic reactions; risk assessments based on historical descriptions alone are not reliable.

Skin-prick testing involves using a panel of reagents. For penicillin testing, this includes major and minor determinants, amoxicillin and benzylpenicillin, plus amoxicillin/clavulanic acid if this was the offending drug. Other agents can also be used depending on the drug in question^[7]
.

Testing should take place in a specialist allergy centre to ensure correct interpretation of results and management of the patient if a severe reaction occurs. These tests should be undertaken relatively quickly after the initial reaction, as positive test results are less likely to be determined as the time period increases^[39]
. Standard practice in the UK is to wait six weeks after the initial allergic reaction took place to test for allergies. However, there is little evidence to support this and testing may be carried out earlier if required. Early negative test results may lead to the patient being retested at a later date^[7]
.

Positive results

Histamine is used as a positive control (weal size can be used as a comparison) and sodium chloride acts as a negative control^[7]
. A weal diameter of a least 3mm greater than the negative control is considered a positive result. Flare and itch also supports a positive result^[7]
. Based on the history of the initial reaction, patients are observed for 15–20 minutes for an immediate response and up to 72 hours for delayed responses^[28]
.

Patients who test positive to the oral challenge may need to undergo desensitisation if alternative antibiotics cannot be used or when a specific drug is necessary for treatment^[7]
.

If any of the results to the tests described are positive, the patient, GP and relevant referring clinician should be informed of the outcome to ensure consistent documentation (see Box 3). NICE also make recommendations about patient information and support (see Box 4).

Negative results

Negative skin-prick and intradermal test results mean that the patient can receive a graded oral challenge to confirm they can tolerate the antibiotic. Oral testing is not required for patients with positive skin tests and is not recommended for patients that have had previous delayed severe cutaneous reactions, have severe asthma or are taking beta-blockers^[40]
. Beta-blockers may be temporarily withheld for up to 48 hours, based on the decision of the GP or cardiologist.

A negative skin-prick result would indicate that intradermal testing should be undertaken. Results are assessed at different intervals, usually at 20 minutes, 24 hours, 48 hours and 72 hours^[7]
.

If intradermal skin tests are negative, the patient and clinician can decide whether oral and intravenous drug challenges are suitable. An example of an oral penicillin provocation test includes 25mg, 100mg, 500mg and 1g doses of amoxicillin given at 30-minute intervals, with standard patient observations (i.e. pulse, blood pressure, oxygen saturations and peak expiratory flow rate) at baseline, before the next dose and 1hour after the final dose^[7],[40]
. If the oral and intravenous tests are uneventful, a full treatment course of the antibiotic (usually five days) is given as per local guidelines and the patient is contacted at the end of the course to establish clinical tolerance. A thorough management plan with emergency treatment should be provided in case of allergic reactions that occur at home^[7]
.

The role of pharmacy professionals

Pharmacists should be able to clarify the nature of an allergy, the significance of reactions to antibiotics and determine whether these and other class-related antibiotics need to be avoided in future. Pharmacists can ensure clear documentation of allergy status and community pharmacists in particular should be able to discuss allergies and the difference between allergy and intolerance with patients; for example, before they purchase over-the-counter medication or when they are dispensing prescriptions for antibiotics^[2]
. Although often considered the remit of the GP, pharmacists can also offer advice on which patients should be referred to specialist allergy clinics.