Article Image

Have Allergy Lists Lost Their Way?

Op-Med is a collection of original articles contributed by Doximity members.

Allergy lists are ubiquitous in medicine. But anyone who works in health care knows that they are often inaccurate. Real allergic reaction postings are nearly always mixed in with dubious data such as poorly-characterized childhood rashes or patient preference against certain medications. 

This is largely because there is not a current accepted standard of how to reconcile the allergy list. To some clinicians, reconciliation means listing out patient stated allergies. To others, it means reviewing past allergies already listed in the chart. Some may even go the extra mile and do both, but give little attention to the validity of the reaction itself. And what about all of the other reactions — shouldn’t they be included too? The result is a foundational medical tool used with great consistency, but low fidelity. 

Below I share the three most common issues I see for the allergy list, and offer a solution for how to reclaim this important tool. 

1) Allergy Over-Listing

A study of Kaiser Permanente’s health records (the nation’s largest) reported that approximately 20% of patients have an active drug allergy listed. However, another study revealed that only 1 in 3 listed allergies were truly immune-mediated reactions with well-documented supporting data. Another third of listings included non-allergic reactions (i.e., common side effects, patient preference), and the final third had no associated data whatsoever. While not surprising, this lack of data makes it nearly impossible for the clinician to truly use this tool, leading to expanding allergy lists with little chance for reconsideration. Indeed, the same study above found that over its seven years of data collection, nearly 700,00 more allergies were logged than deleted. These runaway, unreliable lists do more than just crowd the chart, however. They can have real world consequences.

“My mom always told me I was allergic to penicillin,” my patient told me as I finished examining her tender sinuses and congested nose. “Ever since I was little I’ve avoided it. But I keep getting these sinus infections and these antibiotics just don’t seem to be working.” I scrolled over her profile and saw a recent prescription for levofloxacin filled last month. I glanced at the sidebar and saw “Penicillin” in bright yellow underneath the allergies section of her summary. As I hovered over the icon, I noticed there was no reaction listed. “Well, what was your reaction?” I asked. “Honestly, I have no idea and no one has ever asked me. Most people just go along with it,” she replied. 

The rising rates of improperly classified allergies have significant impacts on patient health. Penicillin antibiotics are among the most frequent culprits. According to the CDC, nearly 10% of U.S. adults report having a prior reaction to penicillin. However, less than 1% of the population has a true allergy, and nearly 80% of these patients will lose their sensitivity after 10 years. Yet, these patients rarely receive confirmatory testing and often are prescribed broader-spectrum antibiotics to treat common infections. This leads to increased health care costs and higher rates of antibiotic resistance. 

2) Lack of Accurate Allergy Data

Improper documentation within the allergy list is another major problem in American medical centers. One study of safety events at a large tertiary medical center found that nearly 40% of inpatient allergic reactions were caused by inaccurate documentation within the EHR. This was almost always due to a failure to appropriately reconcile the information with the patient directly, or to recording the information in the written record but failing to add a listing to the EHR. While these oversights could partially be due to the significant number of tasks required for routine patient care, alarmingly, another study showed that nearly 90% of allergy alerts populated in the EHR are overridden by prescribers. This speaks to a fundamental disregard of this tool as critical to patient care. 

Such inappropriate documentation is serious. Not long ago, I received a message from a nurse colleague: “Hey, I think the patient is having a reaction to this medicine.” I gathered my papers and made my way down the hall reviewing what I knew about the patient from handoff earlier that morning: 12-month-old male admitted for cellulitis and tolerated IV ceftriaxone overnight; transitioned to oral amoxicillin this morning with plan to discharge later today. As I arrived at the bedside I saw an irritable infant with a diffuse red rash covering his chest, back, and face. “He’s allergic to amoxicillin!” his grandmother moaned. “This has happened before!” Sweating, I grabbed the computer and noted that he did indeed have an amoxicillin reaction listed: ‘GI upset’ — a far cry from what I saw before me. “Did anyone ask you what reaction he had?” I asked tentatively. “No,” she replied. “The doctor overnight just told me it was already listed and never asked what had happened.” 

3) Important Drug Information, No Place to Record

Allergy lists are a convenient way for clinicians to prevent certain medications from reaching patients. It is not uncommon to see lists containing “allergies” to NSAIDs, heparin, and other drugs that patients may not be able to tolerate due to their concomitant illnesses. Even more so, allergy lists often contain common adverse drug reactions — think nausea, cough, diarrhea, etc. — for medications like opioids, ACE inhibitors, and metformin that patients may be unable to tolerate. This information is important for patient care and should be included to prevent patients from having needless re-exposure to poorly tolerated medications. However, labeling these reactions as “allergies” reinforces patient misconceptions. 

This is related in part to the EHR design. All three major EHR vendors (Epic, Cerner, and Meditech) list these types of reactions under the “Allergy” heading. Additionally they do not require users to categorize these reactions, and the user interface is cumbersome to use when trying to provide full details. For this reason, the American Academy of Allergy, Asthma, and Immunology (AAAAI) has called for redesigning this section of the EHR to be designated as the “Drug Reaction” section so that this critical data can be captured without reinforcing false labeling among patients and other health care professionals. 

Clearly, our current approach to the allergy list isn’t working. We need a solution that not only restores trust in the list, but is usable for the variety of credentialed staff who reconcile the list and that functions seamlessly in EHR platforms. 

Fortunately, in 2022, the AAAAI proposed a solution in the form of a novel schema. The schema holds that an accurate allergy posting requires four main descriptive domains: 

  1. The approximate date of reaction should be included to help stratify the likelihood of recurrence. 
  2. The symptoms exhibited, as well as their severity, should be written out to help risk-stratify in cases of medication challenge. 
  3. A narrative description of the event should be included to determine the time course of the reaction, sequence of symptoms, and ultimate action taken to resolve the event. 
  4. All allergy postings should be classified as either a true immune-mediated hypersensitivity, or a non-immune adverse drug event by either the clinician of record, or an allergy specialist if confirmatory testing is required. 

Implementing this simple, yet comprehensive method would be a game-changer for patients and health systems. By knowing these key points we can quickly and reliably debunk false allergies, re-trial low risk medications, and avoid harmful exposure to known offenders. Doing so not only helps our system lower cost, but allows us to provide more individualized, and safer, care to our patients. 

The potential of the allergy list to advance patient care cannot be doubted. In its current configuration, however, it is at best incomplete and at worst can actively facilitate patient harm. To ensure that it does not continue to fail us, we must wake up and decide to take this bedrock tool more seriously. 

What in your specialty has "lost its way"? Share in the comments!

Dr. Matthew Scott is a third-year internal medicine-pediatrics resident in Richmond, VA. He is passionate about advocacy, engaging with his community, and his newborn daughter. Dr. Scott is a 2023–2024 Doximity Op-Med Fellow.

Image by elenabsl / Shutterstock

All opinions published on Op-Med are the author’s and do not reflect the official position of Doximity or its editors. Op-Med is a safe space for free expression and diverse perspectives. For more information, or to submit your own opinion, please see our submission guidelines or email

More from Op-Med