HealthAugust 14, 2018

Optimizing EHR drug allergy screening: Concepts, entry options, and picklists


Any clinician that has a role in medication therapy is affected by allergen screening.

Common scenarios in which potential allergen interactions would need to be considered include:

  • A physician entering a new medication
  • A pharmacist filling a prescription
  • Nurses administering medications

However, allergy screening is not limited to specific sectors. It impacts a wide range of healthcare roles and areas including hospice, home healthcare, imaging/radiology, just to name a few.

Until the advent of electronic health records (EHR), allergen screening was performed manually. With the widespread adaptation of EHRs, embedded allergy screening has been well incorporated and has significantly improved medication safety.

In some scenarios, along with an increase in patient safety, we have also seen an increase in undesirable alerts leading to alert fatigue. One cause of this rise in irrelevant EHR notifications is that non-specific or broad drug concepts, that can include unintended ingredients, are being used during the allergen selection process. By employing best practices and optimizing the capabilities of drug data and EHR software, the sensitivity of allergen screening can be improved, thus, enhancing patient outcomes and reducing alert fatigue.

Challenges to effective screening

Two common medication concepts that are used in practice and lead to alert fatigue are NDC and Drug Name.

Using the NDC, or packaged drug concept, to post patient allergens can lead to several issues:

  • The NDC concept contains all ingredients in the medication, including inactive ingredients. If a newly prescribed medication shares inactive ingredients with the allergen, the system will generate unintended alerts involving inactive ingredients, such as dyes or fillers.
  • NDCs do not stay persistent. After a NDC is removed from the market, it will also be removed from databases after a specified amount of time, and an expected allergy alert will not fire. Using NDCs to post allergens requires constant maintenance to identify and replace NDCs as they become obsolete.
  • Some NDCs are recycled by manufacturers. If a recycled NDC remains on a patient’s allergy profile and is not properly maintained, it may cause the system to miss appropriate alerts or lead to false alerting.
  • Medications removed from the market due to safety reasons will have no replacement for obsolete NDCs. This could lead to missing cross-sensitivity reactions to similar medications still marketed. Some recently removed examples include Vioxx and Darvocet. 
  • Since a medication is typically represented by multiple NDCs, searching by this concept to post patient allergens will result in many duplicative or similar results and place an undue burden on the clinician during the selection process. 

The Drug Name concept was designed as a grouping concept used to identify collections of drugs or health-related items. It is independent of route, strength, and dosage form and not directly linked to a specific active ingredient. Whereas the Drug Name concept seems to be a good choice when posting allergens, its use can also lead to alert fatigue:

  • The concept will screen against all packaged drugs containing the Drug Name including their active and inactive ingredients. Because some drugs exist in many dosage forms (e.g., oral suspension, oral tablets, intravenous solutions, eye drops), the list of inactive ingredients included in allergy screening can be numerous (e.g., aspartame, dyes, polysorbate, propylene glycol).
  • These inactive ingredients are also contained in many drug products, leading to a considerable amount of unwanted allergy alerts related to a common inactive ingredient.

Is there a better way?

There are best practices organizations can employ when setting up allergen screening that can help improve its effectiveness.

Allergen picklists

An allergen picklist provides clinicians medication names that are linked to the best concept for allergen screening. It is also designed to give clinicians a concise search that reduces duplicative or confusing drug names. The most effective allergen picklists, like the one developed by Medi-Span, contain common generic names with allergenic potential, select trade names associated with minimal inactive ingredients, select allergen classes, and some inactive ingredients if they are proven to have significant allergic potential (e.g., peanut-containing drugs).

EHR inputs

Some inputs incorporated into the EHR may be helpful in addition to screening. These could include:

  • Customer-defined severity
  • Identity of the EHR user recording the allergen
  • Date/time stamp
  • Person reporting the allergen (i.e., patient, caregiver, other institution’s medical chart)
  • Reaction
  • Allergy vs ADE designation

Medical conditions can be associated with allergens to record reactions and give further context to clinicians. A clinician-assigned designation of allergy versus adverse drug reaction (ADE) or intolerance can also be included. For example, anaphylaxis is an allergic reaction, compared to nausea, which a patient may call an allergic reaction, but is more appropriately defined as an ADE or intolerance.

Timing and overrides

Best practices also include performing allergen screening early in the medication order entry process. This design allows the clinician to be alerted before taking further, time-consuming steps.

Allergy alerts should be allowed to be overridden with the opportunity or requirement for the clinician to document a reason. If possible, this override message should be displayed to subsequent users. For example, if a physician overrides an allergy in an EHR, the recorded override reason should be passed to the pharmacist at order verification and the nurse at administration.


An option to enter No Known Drug Allergies, or “NKDA,” should be easily accessible to the end user. One convenient method to achieve this is adding a one-click NKDA radio button close to the search bar.

Allowing other options, such as “Unknown” or “Not available at this time” is debatable. It may be beneficial in some scenarios, such as an emergency room where a patient may be admitted unconscious or confused. However, offering a prompt to the clinician to update allergens when additional information is available has the potential to create obstacles.

Free-text allergen input should only be allowed after all search options have been exhausted. Free-text allergens will not screen because they are not linked to an ingredient in the system’s data. Therefore, our team always recommends incorporating extra steps for the user before they are permitted to enter a free-text allergen. Also, free-text entries should be easily identifiable in the patient’s profile. This might include displaying the allergen in a different colored text or marking it with an indicator such as an asterisk. These easily recognizable free-text entries should also include a general disclaimer so all clinicians are aware they are unscreenable allergens and manual screening is required.

The Medi-Span common allergen list

In developing Medi-Span drug data at Wolters Kluwer Health, we are building new content and technology on a foundation of more than 45 years of legacy solutions. In our pursuit of the best practices I described, we recognized that some of our legacy screening systems originally allowed limited concept types – such as Drug Name and/or NDC – for allergen posting. As technology and content evolved, we were able to incorporate better concepts for screening. Additionally, we saw a need to improve allergen screening and selection, so we created an Allergen Picklist that started to get adapted around 2012. The picklist is an editorially-reviewed subset of allergens intended to provide healthcare professionals with a user-friendly listing from which allergens can be easily selected and posted to a patient’s health record.

The Allergen Picklist is known by two names: the Medi-Span Allergen Picklist is a flat data file version, whereas the Common Allergen List is incorporated into our Medi-Span Clinical Drug Allergy API (application programming interface). For users of the more advanced, flexible Medi-Span Clinical platform, the Common Allergen List is automatically included in the drug allergy screening API with no additional licensing. The API also includes a filter to exclude insignificant inactive ingredients that are not known to elicit an allergic reaction. These two features, when used together, help to reduce alert fatigue and make for a more user-friendly allergy selection process.

Although the Allergen Picklist/Common Allergen List is compatible with our current allergen screening solutions, we still see NDC and Drug Name being used to post allergens in some EHRs because it remains as a legacy practice from older versions of software or screening resources. In some cases, healthcare organizations and professionals don’t even realize that they aren’t taking full advantage of all the options and capabilities of their drug allergen screening solutions. It is worthwhile to speak with your implementation specialist or customer support department for more information about ways you could optimize drug allergy data and screening within your EHR.

Please note: All these recommendations apply to drug allergen screening in the U.S. International best practice recommendations may differ. Consult your implementation specialist for additional considerations.

Rachael Fisher, PharmD, is a Senior Clinical Implementation Analyst for Wolters Kluwer Clinical Effectiveness. She is a pharmacist with a background in many areas of pharmacy practice including hospital, retail, long-term acute care, and informatics. She serves as a clinical resource for the implementation team and customers of Wolters Kluwer.

Evidence-based drug reference solution used in the workflow and on-the-go
Clinicians choose Lexicomp® for evidence-based drug information to support smart, safe medication decisions. Studies show that Lexicomp is the preferred drug reference solution for care teams.