Imagine if only 10% of all patients labelled diabetic actually had diabetes. Or if 90% of card-carrying asthmatics didn't actually have asthma. Would you still feel comfortable recommending medications (and their side effects) that may be unnecessary or even harmful?
1 in 10 of people report a penicillin allergy, but true penicillin allergy occurs in only 1 in 100 people (even more rarely in children), and severe or life-threatening reactions occur about once in every 10,000 penicillin exposures. Even for true Type 1 hypersensitivity (more on that below...), 80% of people will naturally outgrow their allergy after 10 years.
So making a patient retain an allergy label from childhood into elderly life is like forcing them to wear the same 1990s boy-band T-shirt for the rest of their life, never acknowledging that allergies (like prepubescent musical tastes) evolve with time. Correctly identifying and preventing severe reactions is of course essential, but is it worth mislabelling and mistreating 90% of all patients reporting a penicillin allergy? The answer is a very emphatic NO.
Isn't it better to "just be on the safe side"?
Over-cautious allergy labelling is decidedly not safer for patients. Those labelled with a penicillin allergy are more likely to receive broad-spectrum antibiotics, are at higher risk of infections like MRSA, VRE and C. difficile, have longer hospital stays, and are more likely to die during hospitalisation. This is because, for many infections, penicillins are amongst the most effective and well-tolerated antibiotics. Inappropriate allergy labelling can mean using an inferior drug (e.g. vancomycin for severe staphylococcal or streptococcal infection) or unnecessarily broad-spectrum drugs (e.g. ciprofloxacin or meropenem for a GI infection), which are associated with greater disruption of the patient's microbiome.
So what is penicillin allergy?
"Penicillin allergy" is a bit of a misnomer. Penicillins (and their useful cousins, cephalosporins and carbapenems) are all beta-lactams. Only about 10% of true allergies are directed at the drug's beta-lactam ring itself - the rest are due to the antibiotic's side chains. This is important, as allergic cross-reaction between beta-lactams is much less common than you might think. True allergies are most common with aminopenicillins (e.g. amoxicillin), and cross-reactions in this case are most common with some first generation cephalosporins (e.g. cefalexin, cefaclor). So, even in proven aminopenicillin allergy, using drugs like ceftriaxone, cefuroxime and meropenem is usually safe.
Generally, it's helpful to split drug reactions into three groups:
1) Immediate (type 1 hypersensitivity)
These IgE-mediated reactions occur within an hour of exposure, often following the second or subsequent lifetime exposure.
Caveats: reactions can occur after the first dose, and can be delayed by up to 6 hours (e.g. if drug taken with food).
Reactions include itching, urticaria, angioedema, bronchospasm, abdominal distress, hypotension, and stridor.
Anaphylaxis includes several of these (multi-system) and is rapidly-evolving and potentially fatal if untreated.
Risk factors include frequent courses of the same antibiotic, intravenous (rather than oral use), and age 20-50 years. Type 1 reactions are less common in children.
80% of people with immediate reactions become desensitised naturally (i.e. outgrow their allergy) after 10 years.
2) Delayed (mostly type 4 hypersensitivity)
Includes most mild delayed skin reactions, such as diffuse maculopapular rashes. Onset is hours to weeks after taking the last dose, and symptoms can persist after stopping the drug. These are mostly pretty harmless, except for...
SCARs: severe acute cutaneous reactions, including SJS, DRESS, TEN and AJEP. Look out for fever, pain, blistering or exfoliation, and involvement of mucous membranes (mouth and genitals). These are serious systemic reactions requiring hospitalisation. They are far more likely to recur than Type 1 reactions, and patients with a history of SCARs should never receive penicillins (or related drugs without senior allergy advice).
3) Non-allergic side effects
Mild nausea? Headache? Thrush? A family history of allergy without any personal history? Yup, NONE of these are allergies, and reporting them as allergies is inaccurate and even potentially harmful.
The trouble with kids
The majority of penicillin allergy labels are bestowed in childhood, despite true allergy being even less common in kids than in adults. The reason? Most kids (my preschooler included) are snot-goblins experiencing back-to-back mostly-mild viral infections, many of which cause a rash. If these children receive antibiotics (as many woefully do), this probably-just-a-viral-rash may lead to a life-long penicillin allergy label. Fast forward 40 years and the patient has no memory of the reaction, and the label goes unchallenged "just to be on the safe side". Surely there has to be another way ?! Read on, intrepid prescribers...
And what can YOU do about it?
1) Take an allergy history
The drug - name, dose, route, time of first and most recent dose; any concurrent new medication
The reaction - timing of symptoms and resolution relative to first and most recent drug dose
Any high risk features - Type 1 or SCARs
Any alternative explanation for reaction - e.g. viral illness (especially for childhood rashes), underlying medical conditions, another concurrent drug
Any past history of drug reactions
2) Update the allergy record
Consider these three alternative allergy alerts on a patient's electronic records:
a) "Penicillin allergy"
b) "Vomiting with co-amoxiclav (2010). Has since tolerated flucloxacillin and amoxicillin. Not consistent with penicillin allergy"
c) "Delayed rash to oral amoxicillin (2008). No severe cutaneous reactions or type 1 features"
Now consider this case: A 65 year-old diabetic is admitted with severe right leg cellulitis. He is tachycardic and febrile, with a CRP of 200 and stage 2 acute kidney injury. Each of the three allergy alerts above might lead to a different outcome...
a) If not questioned, the blanket "Penicillin allergy" label might lead to this patient receiving second-line treatment, such as vancomycin, for his cellulitis. A week later, his acute kidney injury is worse, he's in a side room as his rectal swab grew VRE, and his cellulitis isn't improving much... In contrast:
b) With no concern of penicillin allergy, he was treated empirically with first-line antibiotics. A week later, he's already home completing an oral course. Job done.
c) Even with past delayed hypersensitivity, he is able to receive IV cefazolin (no side-chain cross-reactivity with any licensed beta-lactam). He was last seen down the pub celebrating with the patient in scenario b).
So, if you take only ONE thing from this post, let it be this: YOUR updates to the allergy record can make a real difference to patient care. Be specific (drug, date, reaction), and state if a previously-recorded alert is not valid. [TIP: this is more effective than just deleting the allergy label, as future clinicians may mistake this for an omission and add it back in!]
3) Explore direct antibiotic challenge ("test dose")
For low risk in-patients (see orange box above), an observed antibiotic test dose can be used to rule out immediate hypersensitivity. Patients that tolerate a test dose can then have their penicillin allergy label altered or removed.
Due to issues with accessing formal allergy input (e.g. skin testing), hospital-based de-labelling by non-allergists has the potential to improve patient care in a resource-efficient and pragmatic way.
The BSACI (British Society for Allergy and Clinical Immunology) guidelines are fab, as are those by SAPG (Scottish Antimicrobial Prescribing Group), including advice on when antibiotic challenge ISN'T safe.
Evidence-based clinical scoring systems such as PEN-FAST also provide a quick way to assess risk of pen allergy and suitability for antibiotic challenge.
4) Close the loop
If you successfully de-label a patient, rejoice! But remember, the patient may have been carrying that allergy label around for decades, and may be nervous about letting it go. Reassure them, and explain the risks of unnecessary allergy labelling (as outlined above). And document it all clearly, so the GP is kept in the loop.
I hope these top tips help you re-think your approach to managing patients reporting penicillin allergy. And for trickier cases, be sure to call your friendly local Microbiologist - but first, download the free guide on "How to make the perfect call to Micro"!