Commentary: De-labeling penicillin allergies just the beginning

A study published in Annals of Allergy, Asthma & Immunology found that almost one-third of patients who had their penicillin allergies de-labeled might decline penicillin prescriptions or remained unsure about their allergy status.

When the researchers contacted patients a year after their penicillin allergies were de-labeled, 72.9% said they would take penicillin if it were prescribed to them. But 13.6% said they would not, and 8.5% said they were not sure because they did not understand their allergy status. Also, 5.1% said they would be hesitant to accept a penicillin antibiotic and would prefer an alternative.

The role of human nature

I think these findings highlight something that we have known for a while. It is often difficult to agree on who “owns” the penicillin allergy label. Generations of physicians have been trained to err on the conservative side of leaving allergy labels alone and not ever touching them, because they were taught that this is how they should protect their patients and themselves.

However, the emerging evidence base shows that most people outgrow events that used to be thought of as penicillin allergies, many of which were never allergies to begin with. Also, the evidence shows that most penicillin allergy labels (95% or more) can be safely disproven with testing. However, these concepts will take a while to disseminate.

The hard part is that this evidence flies in the face of the natural impulse to avoid things that we suspect have caused us a problem in the past. This impulse is how cavemen survived in a world full of wonders and dangers.

In the modern world, facing a previously experienced adverse drug event head on can feel irresponsible and dangerous, and this impulse is stronger in some people than others. A heightened awareness that people are growing out of most of these events is often the most helpful piece of information that I can share.

The rub is that unnecessary avoidance of unproven allergens can cause unintended consequences that may end up being more harmful to your patient. It is especially true of penicillin allergies, where the testing is safe, almost all the patients who were told they were allergic are not actually allergic, and alternative drugs often do not work as well to treat or prevent an infection.

But this is not just true of penicillin and medication allergies. Food allergies are another big category where a patient’s nutrition can suffer greatly in the face of too many “assumed but never proven” allergies.

In practice

The results of this study are very similar to experiences in our own drug allergy practice.

To improve our ability to de-label patients in multiple settings, we make sure to share records on all the testing we perform with their referring providers. But we also provide the patients themselves with written handouts, summaries of testing, letters and even wallet cards in some instances, so they can advocate for themselves and ask for their disproven penicillin allergies to be removed from their charts.

However, if patients and their other doctors are not highly motivated to remove the allergy label, despite having had the negative test, inertia can kick in and the allergy label can persist. The experience of testing is quite positive for most patients.

But as time goes by, doubts can begin to creep in for some folks. “Did I really have enough testing? Can I really have this allergy removed from my chart?” they may ask. In that setting, the patient’s penicillin allergy label might be removed from the electronic health record where the allergy doctor works, but nowhere else.

When patients feel comfortable enough to return and express these doubts to us, we go back to what is known to be true about penicillin allergy testing. If you have tolerated a dose of the drug that you previously reacted to in an observed setting, either with or without preceding skin testing, you go back to being just like everyone else in terms of your overall risks of a future reaction, and you are not currently at major risk for anaphylaxis.

The main risk for you, as a newly returned member of the general population, is rashes, which happen in about 2% to 5% of all antibiotic treatments. These rashes are mostly random and are kind of like hitting the bankrupt slice on Wheel of Fortune. The more antibiotics you use, the more chances you have of that rash. The more you spin the wheel, the more chances you have to hit one of the “uh oh” slices. Knowing these facts is often the reassurance that patients need.

The label can also creep back into our records if we are not being careful and precise when taking a history in subsequent encounters or when patients only feel “somewhat confident” in the testing they received, such that they keep reporting the reaction but don’t talk about the negative testing. I call these “zombie” labels. You keep shooting them down with testing and evidence and faxing lots of records, but they keep rising from the dead to try and bite us again.
Beyond reservations that the patients might have, there is also the need to get buy in from the patient’s medical providers when removing disproven allergy labels. My primary care colleagues who see the value in having their antibiotic options restored to them for at least one patient — ie, they have an experience where the patient immediately benefits from testing — are the ones who really “catch the bug” and will start referring patients to us for penicillin allergy testing and de-labeling as a routine preventative measure.

These providers realize that it will not only help their patients acutely in some instances where there is a treatment need for penicillin, but also that the testing and de-labeling might help them with a treatment option for many other infections these patients might subsequently experience for the rest of their lives. This value proposition is very lopsided in favor of testing and de-labeling, but I know we need to explain it better.

For the future, we also need to go back to our medical schools and healthcare training programs and revamp our teaching of medication allergies, especially the assumption that all penicillin allergy labels represent a hard, lifelong “stop sign.” Instead, we need to think of these possible penicillin allergy events as events that can usually be resolved with careful questions, maybe some temporary avoidance to give time for the air to clear, and then allergy testing.

It is to the benefit of patients and their doctors when we do so and reframe these events as phenomena that are unconfirmed and that the odds are almost always in favor of “not an allergy” for the long run.

At a minimum, we clearly need to move away from what a lot of my patients have experienced, which is often a lifetime of people (including their doctors) telling them they will “die if they ever take penicillin again” without ever having had the testing to confirm whether they are actually allergic. This is especially problematic for events that were not even consistent with an allergy to begin with.

I think that this injection of fear with every mention of penicillin is becoming an increasingly irresponsible thing to do in light of the available evidence. Adding to the fear of a penicillin allergy without knowing for sure whether a patient is truly allergic does not seem like a good way to practice.

Next steps

This research by Pinto et al highlights that the next big frontier is overhauling our allergy education for all kinds of health care providers so that we can introduce the necessary level of precision and skepticism about events that have been historically interpreted as penicillin allergies.

With new data from the last couple of decades in hand, the times in which we tell someone that they will need lifelong avoidance of a penicillin are going to be really rare, and this kind of advice probably shouldn’t be given to a patient without a specialty consultation and possibly testing.

Along those lines, I think disseminating the evidence on the emerging best practices for penicillin allergy evaluations throughout the health care system should be a full-time effort for lots of people at this point. Pinto et al also highlight how more research needs to be done on the optimal communication of drug allergy testing with a patient’s other health care providers.

Finally, this study highlights how primary care physicians are really busy. It might surprise people to know how enormous the amounts of paperwork are that their PCPs have to digest and complete for every single patient encounter. Helping patients manage their allergies is only one of the hundreds of things we are asking them to do.

Therefore, my final thought is that we clearly need to figure out how to implement our new knowledge of penicillin allergy risk stratification, testing and management into our routine care pathways without adding to the burdens of our primary care doctors.

For more information:

Cosby A. Stone Jr., MD, MPH, is assistant professor of medicine in the division of allergy, pulmonology and critical care medicine within the department of medicine at Vanderbilt University Medical Center and VUMC Drug Allergy Research. He can be reached at cosby.a.stone@vumc.org.