Evaluating the PEN-FAST Clinical Decision-making Tool to Enhance Penicillin Allergy Delabeling

A call to address penicillin allergy labels in patients with hematopoietic stem cell transplants: How to avoid rash decisions

Among patients with hematopoietic stem cell transplants, infections, particularly multidrug-resistant infections, pose a grave threat. In this setting, penicillin allergy labels are both common and harmful. Though the majority of patients who report penicillin allergy can actually tolerate penicillin, penicillin allergy labels are associated with use of alternative antibiotics, which are often more broad spectrum, less effective, and more toxic. In turn, they are associated with more severe infections, multidrug-resistant infections, Clostridium difficile, and increased mortality. Evaluating penicillin allergy labels can immediately expand access to preferred therapeutic options, which are critical to care in patients with recent hematopoietic stem cell transplants. Point-of-care assessment and clinical decision tools now exist to aid the nonallergist in assessment of penicillin allergy. This can aid in expanding use of other beta-lactam antibiotics and assist in risk-stratifying patients to determine a testing strategy. In patients with low-risk reaction histories, direct oral challenges can be employed to efficiently delabel patients across clinical care settings. We advocate for multidisciplinary efforts to evaluate patients with penicillin allergy labels prior to transplantation.

Predictive models and applicability of artificial intelligence-based approaches in drug allergy

PURPOSE OF REVIEW

Drug allergy is responsible for a huge burden on public healthcare systems, representing in some instances a threat for patient's life. Diagnosis is complex due to the heterogeneity of clinical phenotypes and mechanisms involved, the limitations of in vitro tests, and the associated risk to in vivo tests. Predictive models, including those using recent advances in artificial intelligence, may circumvent these drawbacks, leading to an appropriate classification of patients and improving their management in clinical settings.

RECENT FINDINGS

Scores and predictive models to assess drug allergy development, including patient risk stratification, are scarce and usually apply logistic regression analysis. Over recent years, different methods encompassed under the general umbrella of artificial intelligence, including machine and deep learning, and artificial neural networks, are emerging as powerful tools to provide reliable and optimal models for clinical diagnosis, prediction, and precision medicine in different types of drug allergy.

SUMMARY

This review provides general concepts and current evidence supporting the potential utility of predictive models and artificial intelligence branches in drug allergy diagnosis.

Efficacy of the PEN-FAST score in a French cohort of patients with reported allergy to penicillins

Introduction

Various clinical decision-making tools for penicillin allergy have been developed to guide delabeling strategies.

Objective

To evaluate the penicillin allergy PEN-FAST decision score in a retrospective cohort of patients, adults and children, with penicillin-reported allergy.

Methods

This monocentric retrospective cohort included patients with penicillin-reported allergy. All patients underwent penicillin allergy testing using skin tests and/or drug challenge. The PEN-FAST score sensitivity, specificity, negative (NPV) and positive (PPV) predictive values, and the area under the receiver operating characteristics curve (AUC) were calculated.

Results

Two hundred and fourteen patients were included (64 children and 150 adults). Allergy was confirmed in 52 cases (24%). A PEN-FAST score <3 points showed a poor discrimination capacity for the whole population (AUC = 0.66; 95% CI: 0.58-0.75), while it demonstrated a better discrimination capacity in the adults group (AUC = 0.71; 95% CI: 0.63-0.80). The sensitivity to identify penicillin allergy using this cutoff of less than 3 points was 0.67 (95% CI: 0.52-0.80); specificity, 0.58 (95% CI: 0.48-0.68); PPV, 0.43 (95% CI: 0.32-0.55); and NPV, 0.78 (95% CI: 0.68-0.87).

Conclusions

Although our data confirm a rather good discrimination value of a PEN-FAST score <3 points, its low negative predictive value (78%) did not advocate for its use as an accurate, simple and cost-effective clinical decision-making tool to effectively reduce the number of penicillin skin tests required before direct oral challenge. Further studies are required to improve the predictive capacity of the PEN-FAST score.

Comparative Performance of 4 Penicillin-Allergy Prediction Strategies in a Large Cohort

BACKGROUND

A safe and pragmatic guide for labelling and delabelling patients with suspected penicillin allergy is mandatory.

OBJECTIVE

To compare the performance of 4 penicillin-allergy prediction strategies in a large independent cohort.

METHODS

We conducted a retrospective study for subjects presenting between January 2014 and December 2021 at the University Hospital of Montpellier, with a history of hypersensitivity to penicillins. The outcome targeted by the study was a positive penicillin-allergy test.

RESULTS

Of the 1,884 participants included, 382 (20.3%) had positive penicillin-allergy tests. The ENDA (European Network on Drug Allergy) and Blumenthal strategies yielded relatively high sensitivities and low specificities and, by design, did not misclassify any positive subjects with severe index reactions. The PEN-FAST <3 score had a negative predictive value of 90% (95% confidence interval [95% CI] 88%-91%), with a sensitivity of 66% (95% CI 62%-71%) and a specificity of 73% (95% CI 71%-75%), and incorrectly delabelled 18 subjects with anaphylaxis and 15 with other severe nonimmediate reactions. For the adapted Chiriac score, the specificity corresponding to 66% sensitivity was 73% (95% CI 70%-75%). Conversely, at a 73% specificity threshold, the sensitivity was 65% (95% CI, 61%-70%). Attempts to improve these prediction algorithms did not substantially enhance performance.

CONCLUSIONS

The ENDA and Blumenthal strategies are safe for high-risk subjects, but their delabelling effectiveness is limited, leading to unnecessary avoidance. Conversely, the PEN-FAST and Chiriac scores are performant in delabelling, but more frequently misclassify high-risk subjects with positive penicillin-allergy tests. Selection of the most appropriate tool requires careful consideration of the target population and the desired goal.

Predicting Penicillin Allergy: A United States Multicenter Retrospective Study

BACKGROUND

Using the reaction history in logistic regression and machine learning (ML) models to predict penicillin allergy has been reported based on non-US data.

OBJECTIVE

We developed ML positive penicillin allergy testing prediction models from multisite US data.

METHODS

Retrospective data from 4 US-based hospitals were grouped into 4 datasets: enriched training (1:3 case-control matched cohort), enriched testing, nonenriched internal testing, and nonenriched external testing. ML algorithms were used for model development. We determined area under the curve (AUC) and applied the Shapley Additive exPlanations (SHAP) framework to interpret risk drivers.

RESULTS

Of 4777 patients (mean age 60 [standard deviation: 17] years; 68% women, 91% White, and 86% non-Hispanic) evaluated for penicillin allergy labels, 513 (11%) had positive penicillin allergy testing. Model input variables were frequently missing: immediate or delayed onset (71%), signs or symptoms (13%), and treatment (31%). The gradient-boosted model was the strongest model with an AUC of 0.67 (95% confidence interval [CI]: 0.57-0.77), which improved to 0.87 (95% CI: 0.73-1) when only cases with complete data were used. Top SHAP drivers for positive testing were reactions within the last year and reactions requiring medical attention; female sex and reaction of hives/urticaria were also positive drivers.

CONCLUSIONS

An ML prediction model for positive penicillin allergy skin testing using US-based retrospective data did not achieve performance strong enough for acceptance and adoption. The optimal ML prediction model for positive penicillin allergy testing was driven by time since reaction, seek medical attention, female sex, and hives/urticaria.

Penicillin allergy reassessment for treatment improvement: A dental office tool to support appropriate penicillin allergy labeling

BACKGROUND

Dental appointments offer an opportunity to evaluate a documented penicillin (PCN) allergy and determine whether the patient might be a candidate for medical reassessment of their allergy. The authors gathered feedback on the Penicillin Allergy Reassessment for Treatment Improvement (PARTI) tool, designed to enhance dentist-patient communications regarding PCN allergies.

METHODS

From January 2022 through May 2023, the authors conducted a mixed-methods study, collecting focus group data from patients with PCN allergies and surveying health care workers (HCWs) regarding the PARTI tool. Feedback focused on reassessment procedures, patient-centered allergy information, and medical records updates. Thematic analysis was used for focus group data.

RESULTS

The study included 15 patients in focus groups and 50 HCW survey respondents representing diverse US regions. Patient demographic characteristics included varied races, the mean age was 52 years, and most of the patients were female (53.3%). Most patients had health care interactions within the preceding year, at which 86.6% of patients were asked about drug allergies. HCW respondents primarily consisted of pharmacists (30%) and dentists, dental hygienists, and dental assistants (28%). Feedback on the PARTI tool was constructive, with both patients and HCWs recognizing its potential benefits and providing insights for improvement. Many HCWs (68%) highlighted the importance of step 3 of the PARTI tool, that is, the section on PCN allergy testing. Feedback from participants was incorporated into the final PARTI tool.

CONCLUSIONS

Patient and HCW feedback on the PARTI tool was used to finalize a tool for the dental office to provide to patients who are candidates for PCN allergy reassessment. The feedback will also be used to inform an upcoming pilot study in US dental offices, focused on the process for PCN allergy reassessment and health record documentation.

PRACTICAL IMPLICATIONS

Deploying the PARTI tool in dental offices is pivotal, as mislabeling patients with PCN allergies could have severe consequences, such as hindering the prescription of lifesaving antibiotics for conditions like endocarditis, in the future. This implementation not only enhances communication between dentists and patients, but it is also crucial for ensuring improved patient safety and maintaining accurate medical records among health care settings.

Penicillin Allergy Evaluation Should Be Performed Proactively in Patients With a Penicillin Allergy Label

Penicillin allergy is the most common drug allergy in the US population. A penicillin allergy label is associated with poor patient outcomes including increased hospital length of stay, increased perioperative infections, and overall increased mortality. A penicillin allergy evaluation accurately identifies approximately 9 of 10 patients who, despite reporting a history of penicillin allergy, can receive penicillins safely. Penicillin allergy evaluations should be offered proactively to healthy patients during routine visits, including children and pregnant women, in advance of antibiotic need.