MMHA 6600 Week 7 Allergy Alert Fatigue Case Example

The Rash and the Reaction Look the Same on Screen: An Alert Fatigue Case Traced to How the Allergy List Is Built

Student Name

Walden University

MMHA 6600: Healthcare Informatics and Technology Management

Instructor Name

Month Day, Year

The Rash and the Reaction Look the Same on Screen: An Alert Fatigue Case Traced to How the Allergy List Is Built

The Case

On a medical unit at Bramble Creek Hospital, the composite hospital used for this course's assignments, a hospitalist ordered cefazolin before a minor procedure for a man whose allergy list read "penicillin." The record fired an interruptive cross-sensitivity alert. The hospitalist overrode it, as she had overridden the same alert many times that month, and wrote "tolerated before" in the override reason. The patient did well. A week later, on the same unit, a different prescriber overrode an identical-looking alert for a woman whose penicillin entry recorded hives and throat tightness years earlier. She also did well, after a nurse questioned the order and the dose was given with monitoring.

Both overrides looked the same in the alert log. The difference between the two patients, a stomach upset in one and a history consistent with true allergy in the other, was visible to nobody at the moment of decision. This case analysis traces the habit of overriding to that invisibility and proposes changes to how the allergy list is built.

How Common Overrides Are

Overriding drug allergy alerts is the norm rather than the exception. A study of a decade of alerts at two academic centers found that override rates for drug allergy alerts rose over time and that most overrides were judged appropriate, frequently because the patient had tolerated the drug before or the listed reaction was not allergic (Topaz et al., 2016). Earlier work on drug safety alerts in general found override rates between about half and almost all alerts, with prescribers citing irrelevance and repetition (van der Sijs et al., 2006). At Bramble Creek, the pharmacy's review of the past quarter found that 91 percent of penicillin cross-sensitivity alerts on adult units were overridden. That figure is illustrative for this sample paper, but it sits within the published range.

Root Cause: One Field for Two Kinds of Entry

A review of 200 adult penicillin entries on the allergy list found the pattern summarized in Table 1. The list stores intolerances, side effects, family history and true allergies in the same field, with the same severity default, and the alert logic treats them all alike.

This mix is consistent with the wider literature. When people carrying a penicillin label are formally tested, the large majority turn out to tolerate the drug, and the label does damage of its own: labeled patients receive broader-spectrum antibiotics and have higher rates of resistant infections and longer stays (Macy & Contreras, 2014; Blumenthal et al., 2019).

Table 1

Recorded Reactions on 200 Penicillin Allergy Entries (Illustrative)

Fewer than one entry in three described a reaction consistent with true immediate allergy, yet every entry produced the same interruptive alert.

Why Prescribers Learn to Override

An alert teaches through repetition. When most alerts concern entries that are not true allergies, prescribers learn that the alert is usually wrong, and the learned response, override, is applied to the rare alert that matters. The hospitalist in the first case was behaving reasonably given her experience. The prescriber in the second case was doing the same thing. Nothing in the alert distinguished the patient with hives and throat tightness from the patient with an upset stomach.

In sociotechnical terms, the failure belongs to the interaction between content, interface and people, not to any single part (Sittig & Singh, 2010). Retraining prescribers to read alerts more carefully would ask them to work harder against a system that keeps teaching the opposite lesson.

Proposed Changes

Four changes address the list rather than the prescriber.

First, split the entry. Require a reaction type when an allergy is entered, chosen from a short list: intolerance or side effect, low-risk reaction, possible immediate allergy, severe reaction. Second, tie alert behavior to the type. Severe and possible immediate reactions keep an interruptive alert; low-risk reactions show a passive notice; intolerances show nothing at ordering. Third, clean the existing list. A pharmacist-led review of penicillin entries on admission, using the patient interview, can reclassify most entries and identify patients for allergy testing. Fourth, record tolerance. When a patient receives a drug in the same class without reaction, the record should prompt the nurse to note it, so the next prescriber sees evidence rather than a bare label.

Table 2

Changes, Owners and Measures

Each measure can be pulled from existing reports, so progress can be shown within one quarter.

Risks

Fewer alerts could mean a missed true allergy if entries are misclassified. The safeguard is that uncertain or blank entries default to the possible immediate allergy tier until reviewed. The review workload for pharmacy is real; starting with patients who are about to receive a beta-lactam limits it. A third risk is cultural: some prescribers may distrust a passive notice at first and ask for the old alert back. The pharmacy and therapeutics committee should publish the first quarter's results, including any reaction in a patient whose entry had been downgraded, so the change is judged on evidence. If a downgraded entry is ever followed by a true reaction, the case should go through the same review as any other safety event, and the tier rules adjusted if needed.

Conclusion

The overrides at Bramble Creek are not careless. They are the predictable response to an alert that cannot tell a rash from a reaction because the allergy list records both in the same way. Building the list to carry reaction type, and letting alerts follow that type, gives the rare dangerous entry the attention it needs and stops teaching prescribers to ignore every alert.

References

Blumenthal, K. G., Peter, J. G., Trubiano, J. A., & Phillips, E. J. (2019). Antibiotic allergy. The Lancet, 393(10167), 183-198. https://doi.org/10.1016/S0140-6736(18)32218-9

Macy, E., & Contreras, R. (2014). Health care use and serious infection prevalence associated with penicillin "allergy" in hospitalized patients: A cohort study. Journal of Allergy and Clinical Immunology, 133(3), 790-796. https://doi.org/10.1016/j.jaci.2013.09.021

Sittig, D. F., & Singh, H. (2010). A new sociotechnical model for studying health information technology in complex adaptive healthcare systems. Quality and Safety in Health Care, 19(Suppl 3), i68-i74. https://doi.org/10.1136/qshc.2010.042085

Topaz, M., Seger, D. L., Slight, S. P., Goss, F., Lai, K., Wickner, P. G., Blumenthal, K. G., Dhopeshwarkar, N., Chang, F., Bates, D. W., & Zhou, L. (2016). Rising drug allergy alert overrides in electronic health records: An observational retrospective study of a decade of experience. Journal of the American Medical Informatics Association, 23(3), 601-608. https://doi.org/10.1093/jamia/ocv143

van der Sijs, H., Aarts, J., Vulto, A., & Berg, M. (2006). Overriding of drug safety alerts in computerized physician order entry. Journal of the American Medical Informatics Association, 13(2), 138-147. https://doi.org/10.1197/jamia.M1809