Abstract
Background:
Frontline providers frequently make time-sensitive antibiotic choices, but many feel poorly equipped to handle antibiotic allergies.
Objective:
We hypothesized that a digital decision support tool could improve antibiotic selection and confidence when managing beta lactam allergies.
Methods:
A digital decision support tool was designed to guide non-allergist providers in management of patients with beta lactam allergy labels. Non-allergists were asked to make decisions in clinical test cases without, then with, the tool. These decisions were compared using paired t-tests. Users also completed surveys assessing their confidence in managing antibiotic allergies.
Results:
The tool’s algorithm was validated by confirming its recommendations aligned with that of five allergists. Non-allergist providers (n=102) made antibiotic management decisions in test cases, both with and without the tool. Use of the tool increased the proportion of correct decisions from 0.41 to 0.67, a difference of 0.26 (95% CI: 0.22–0.30, p<0.001). Users were more likely to give full-dose antibiotics in low-risk situations, give challenge doses in medium-risk situations, and avoid the antibiotic and/or consult Allergy in high-risk situations. Ninety-eight (96%) users said the tool would increase their confidence when choosing antibiotics for patients with allergies.
Conclusion:
A point-of-care clinical decision tool provides allergist-designed guidance to non-allergists and is a scalable system for addressing antibiotic allergies, irrespective of allergist availability. This tool encouraged appropriate antibiotic use in low- and medium-risk situations and increased caution in high-risk situations. A digital support tool should be considered in quality improvement and antibiotic stewardship efforts.
Keywords: Beta lactam allergy, penicillin allergy, antibiotic allergy, digital tool
Introduction
Beta lactam antibiotics are first-line treatment for many common infections and for surgical prophylaxis. Beta lactam allergy labels often result in use of alternative antibiotics that are unnecessarily broad in spectrum, less effective, and associated with negative outcomes.1,2 Patients with beta lactam allergy labels utilize healthcare systems more frequently and are prescribed more expensive antibiotics.3,4 A penicillin allergy label is associated with multidrug-resistant organism colonization and infections such as Clostridium difficile, methicillin-resistant Staphylococcus aureus, and vancomycin-resistant Enterococcus species.5 Patients with antibiotic allergy labels also have increased risk of intensive care unit admissions, longer hospital stays, longer time to defervesce, increased number of antibiotics used, and higher rate of treatment failure.3,6 Surgical patients with a penicillin allergy label had 50% increased odds of developing a surgical site infection due to the receipt of second-line perioperative antibiotics.7 Similarly, pregnant patients with unverified penicillin allergy labels who received non-beta lactam antibiotics had increased rates of cesarean deliveries, days in hospital, and morbidity.8
Most patients who carry beta lactam allergy labels can be delabeled after a single dose challenge, sometimes after penicillin skin testing. Delabeling has been shown to change prescription behavior and result in overall health care savings.9 Beta lactam allergy delabeling can reduce outpatient visits, emergency department visits, and hospitalizations.10
Although beta lactam allergy delabeling is routinely done by allergists, non-allergist providers may not routinely address drug hypersensitivities. Frontline providers frequently need to make time-sensitive antibiotic choices but may not have sufficient training to optimize antibiotic management in patients with allergy labels. The lack of drug allergy education, incomplete allergy documentation, and inaccurate labeling of beta lactam allergies have prompted many institutions to create decision support measures aimed at improving antibiotic allergy management. These have included visual algorithmic guidelines, mobile-friendly applications, electronic health record-incorporated tools and alerts, pharmacist-led penicillin skin testing, involvement of various healthcare providers to obtain allergy histories, and other educational programs.11–16 Some of these interventions have been shown to improve providers’ knowledge of drug allergies.11 Other implementations have resulted in more skin testing and challenges to penicillin and improved electronic documentation of allergies.12,17–19 Such stewardship efforts have increased the number of beta lactam prescriptions and decreased use of broad-spectrum antibiotics for patients who previously reported penicillin allergies.12,18,19
Non-allergist providers must often make time-sensitive decisions on antibiotic selection but may not feel comfortable classifying reactions, have time to review educational materials, or have allergy consultant services available. We therefore developed an easily accessible and interactive tool using the survey function of Research Electronic Data Capture (REDCap) to apply branching logic to antibiotic reaction histories. After a user enters the antibiotic of interest and prior drug reaction details, our Antibiotic Allergy Tool (AAT) produces a recommendation on whether the patient may receive the desired antibiotic, and if so, how it should be administered. We hypothesized that this novel, digital point-of-care decision support tool designed for non-allergists could augment adherence to drug allergy guidelines by helping them identify not only low-risk situations in which a beta lactam could be used, but also high-risk situations in which beta lactam use should be avoided. Our aim was to assess how the AAT changed frontline providers’ antibiotic selection behavior and confidence using test cases representative of clinical scenarios. We hoped that these findings could support implementation of the AAT in a live clinical setting.
Methods
AAT design and validation
The AAT is accessible on any smartphone, tablet, laptop, or desktop computer via a REDCap weblink. Because the AAT is intended for use by non-allergists who have not received specialized training, it can be used in its entirety without accessing other educational materials. It does, however, provide optional links to additional resources, such as videos and documents about the tool’s algorithm and other information on antibiotic allergies.
The AAT elicits from the user their role and medical specialty, the needed antibiotic, the antibiotic listed as an allergy, details of the prior reaction, and the user’s assessment of the patient’s clinical stability (Figure E1). The AAT ultimately produces a recommendation for the provider to give a full dose of the antibiotic, give a test dose of the antibiotic (e.g., graded challenge), refer to outpatient Allergy clinic, consult the inpatient Allergy service, and/or choose an alternative antibiotic. We tailored these recommendation options to the available resources at our institution’s inpatient and outpatient settings. For example, we have developed nursing and pharmacy procedures and protocols for inpatient and outpatient graded challenges at our hospitals, and the Allergy service is available for inpatient consultation. Drug skin testing, however, cannot always be done within hours of a consultation, as we do not have a dedicated inpatient penicillin skin testing service.
We used Section VII A of the 2010 Drug Allergy Practice Parameter to design the AAT’s branching logic in determining whether the needed beta lactam and listed allergy have side chain similarity (Table E1).20 If there is side chain similarity, the AAT will result in a more cautious recommendation given the current limited data on the relevance of side chain similarity and cephalosporin reactions. The prior reaction details are classified by the AAT based on type of hypersensitivity reaction. We require the user to assess the patient’s clinical status: if a patient is unstable or unable to communicate, cannot safely receive epinephrine, cannot be monitored, or cannot discontinue a beta blocker for graded challenge, then the AAT will recommend inpatient Allergy service consult regardless of side chain similarity or reaction type, in the interest of patient safety. If a user has insufficient information to answer any question, an “unknown” response can be selected, which leads to a more conservative recommendation by the AAT.
To both validate and test the AAT, we created six sample clinical cases drawn from representative patient scenarios (Table E2). We included six cases as this number allowed for representation of varying scenarios while minimizing the time requested of participants. We assigned each case a risk level based on the index antibiotic allergy reaction classification and cross reactivity with the needed antibiotic. If participants used the AAT correctly for the two low-risk cases, the AAT recommends giving a full dose of the antibiotic. The medium-risk case should result in the AAT recommending giving a test dose, e.g. graded challenge. The three high-risk cases correspond to reactions for which the risk of reintroduction is high, such that the AAT should recommend consulting the Allergy service and/or choosing an alternative antibiotic. We included three high-risk scenarios to verify that participants would be able to identify situations in which there is considerable risk for an adverse reaction if the patient receives the desired antibiotic. These high-risk cases were disproportionately represented in the interest of patient safety, as the AAT is intended for use by non-allergists. Both adult and pediatric case versions were created with identical antibiotics and reaction histories, and study participants accessed the version most applicable to their practice expertise.
We validated the AAT’s logic by comparing its recommendations to those of allergists. Five board-certified allergists not otherwise involved in this study were asked to recommend a clinical action for each of the pediatric and adult versions of the cases without using the AAT. To ensure safety in a potential live clinical setting, the recommendation by the AAT was deemed satisfactory if it was the same or more conservative than the allergists’ recommendation.
Study design and data collection
Ten internal medicine residents were invited by email to pilot the AAT in December 2019. Their qualitative feedback was used to refine the AAT prior to this study. From January to July 2020, we invited internal medicine, pediatrics, and anesthesiology residents, infectious disease specialists, and pharmacists via email to group sessions. We aimed to include 100 individuals in the study. Each participant partook in one session. Using their mobile devices or laptop computers, they completed an initial survey (Table E3), chose decisions for the six clinical cases without the AAT, chose decisions for the cases with the AAT, and completed a final survey (Table E4). The surveys were developed using Qualtrics. Participants were given a unique user code to link responses from each section. For each clinical case, users could choose to give a full dose of the antibiotic, give a test dose of the antibiotic (graded challenge), refer to outpatient Allergy clinic, choose an alternative antibiotic, and/or consult inpatient Allergy service. Those who completed the session received a $15 gift card.
Our primary outcome was the proportion of users’ correct decisions, defined as decisions concordant with AAT recommendations, for the cases with versus without the AAT. Our secondary outcomes included changes in the proportion of correct decisions based on case risk-level with versus without the AAT. We aimed to depict how users’ decisions and patterns of behavior changed with and without the tool. We also measured alterations in users’ self-reported confidence of management of patients with antibiotic allergies with versus without the AAT. We asked users to report on the applicability and usability of the AAT in the final survey.
Statistical analysis
We evaluated our primary outcome using paired t-tests and 95% confidence intervals. We also performed a subgroup analysis based on case-risk level, as well as medical specialty and level of training. We did a sensitivity analysis comparing the proportion of equally-as or more-conservative decisions with versus without the AAT. For this, we grouped decisions into the following categories, ranked by conservativeness: most conservative (choose alternative antibiotic and/or consult inpatient Allergy, or refer to outpatient Allergy clinic), less conservative (give challenge dose of antibiotic), or least conservative (give full dose of antibiotic). We evaluated these secondary outcomes using 95% confidence intervals.
We expected the difference in proportion of cases correctly classified would be at least 13% with and without the AAT; therefore, a sample size of 100 users was required for our study to be adequately powered. All statistical tests were two-sided and performed at the 0.05 significance level. Analyses were conducted using R version 3.6.2.21–23
Results
Validation of the AAT
The branching logic of the AAT was validated by a comparison of its recommendations to board-certified allergists’ recommendations. For four cases, including the two low-risk and one medium-risk scenarios, allergists made the same recommendation as the AAT (Figure 1). For one high-risk case, one allergist made the less conservative recommendation to give a test dose of the antibiotic than the AAT’s recommendation to avoid the antibiotic. For another high-risk scenario, four allergists made the less conservative recommendation to give a test dose of the antibiotic than the AAT’s recommendation to avoid the antibiotic. These results met our criteria for validation in that none of the allergists recommended more conservative actions than the AAT recommendations.
Figure 1:
Allergist recommendations compared to AAT recommendations. Each horizontal bar represents one of six clinical cases, with asterisks (*) indicating the three high-risk cases. Each bar is separated with colors indicating the percentage of allergists who recommended the same action (orange) or a less conservative action (gray) than the AAT.
Participants
Our analysis included 102 users who completed a full session and excluded 11 individuals who completed only portions of a session. The users included 24 internal medicine, 12 pediatric, and 8 anesthesiology residents, 19 infectious disease specialists, and 39 pharmacists. In total, we conducted 17 user sessions, each including two to 20 participants per session. Each session was 25 to 45 minutes long, and they were done in-person from January through February 2020, then virtually starting in March 2020 to adhere to pandemic-related distancing precautions.
Clinical cases without versus with the AAT
Use of the AAT significantly increased the proportion of correct decisions, defined by concordance with AAT recommendations, for the clinical cases compared to without the AAT by 63%, from 0.41 to 0.67, a difference of 0.26 (95% CI: 0.22–0.30, p<0.001) (Figure 2). Use of the AAT also increased the proportion of equally-as or more conservative decisions compared to without the AAT by 19%, from 0.78 to 0.93, a difference of 0.15 (95% CI: 0.12–0.19, p<0.001).
Figure 2:
Correct decisions without and with the AAT. The two left-sided bars indicate the proportion of decisions matching AAT recommendations. The two right-sided bars indicate the proportion of decisions matching or being more conservative than AAT recommendations. Gray and orange bars represent the proportion of correct decisions before and after use of the AAT, respectively. Error bars represent 95% confidence intervals.
We analyzed user decisions based on case risk-levels to identify the impact of the AAT in different situations (Figure 3) and show exact responses in each case (Figure 4). In the two low-risk cases, though most users were initially correct without the AAT, use of the AAT nonetheless increased the proportion of users opting to give the full dose of the needed antibiotic by 14%, from 0.80 to 0.91, a difference of 0.11 (95% CI: 0.04–0.18, p = 0.004) (Figure 3). Only one user in chose an alternative antibiotic after using the AAT, suggesting the AAT’s role in encouraging preferred beta lactam use in low-risk situations (Figure 4).
Figure 3:
Correct decisions without and with the AAT based on case risk-level. There were three high-risk, one moderate-risk, and two low-risk cases. Gray and orange bars represent the proportion of correct decisions before and after use of the AAT, respectively, and are grouped according to case risk level. Error bars represent 95% confidence intervals.
Figure 4:
User decisions without and with the AAT. For each case, the risk level and AAT recommendations are as follows: Case 1, medium-risk, give a graded challenge dose; Case 2, high risk, choose alternative antibiotic and consult inpatient Allergy; Cases 3 and 4, high-risk, consult inpatient Allergy; Cases 5 and 6, low-risk, give a full dose.
In the medium-risk case, the proportion of users who correctly chose to give a test dose of the antibiotic without the AAT was only 0.17. Of those initially incorrect, 28% of users had selected the less conservative decision to give a full dose of the antibiotic, while 72% had selected one of the more conservative decisions to avoid the antibiotic (Figure 4). With the use of the AAT, the proportion of users selecting the correct decision increased by 300%, from 0.17 to 0.68, a difference of 0.51 (95% CI: 0.40–0.62, p<0.001) (Figure 3). Use of the AAT increased the likelihood of challenging the patient with the preferred antibiotic rather than avoiding the beta lactam.
For the high-risk cases, use of the AAT increased the proportion of correct decisions by 44%, from 0.62 to 0.88, a difference of 0.26 (95% CI: 0.20–0.33, p<0.001) (Figure 3). Use of the AAT increased caution in these situations compared to without the AAT in that fewer users chose to give a full or test dose of the antibiotic (Figure 4).
Across all case risk levels, many participants chose to give an alternative antibiotic without allergist consultation without the AAT, but subsequent use of the AAT reduced this behavior (Figure 4). For the low and medium-risk cases, the proportion of users opting to give an alternative antibiotic alone decreased from 0.27 without the AAT to 0.01 with the AAT, a difference of 0.26 (95% CI: 0.17– 0.34, p<0.001). With the AAT, more users instead chose to administer the beta lactam antibiotic, either by regular dosing or graded challenge. The proportion of users who chose to give an alternative antibiotic without Allergy consultation in the high-risk cases decreased from 0.60 without the AAT to 0.15 with the AAT, a difference of 0.45 (95% CI: 0.34– 0.56 p<0.001). Users in these cases shifted towards allergist consultation when using the AAT.
To identify a specific group of potential users for whom the AAT might have a greater impact, we analyzed decisions by users’ roles. The greatest impact was observed among resident physicians, for whom use of the AAT increased the proportion of correct decisions compared to without the AAT by 89%, from 0.37 to 0.70, a difference of 0.33 (95% CI: 0.27–0.40, p<0.001) (Figure 5). Among infectious diseases specialists, use of the AAT increased the proportion of correct decisions by 70%, from 0.43 to 0.73, a difference of 0.30 (95% CI: 0.21–0.40, p<0.001). Use of the AAT by pharmacists increased the proportion of correct decisions by 33%, from 0.45 to 0.59, a difference of 0.14 (95% CI: 0.10–0.21, p<0.001).
Figure 5:
Correct decisions without and with the AAT based on user role. Residents include internal medicine, pediatrics, and anesthesiology residents. Infectious Disease (ID) specialists and pharmacists include both adult and pediatric practitioners. Gray and orange bars represent the proportion of correct decisions before and after use of the AAT, respectively. Error bars represent 95% confidence intervals.
Survey responses
We sought to understand how the AAT affected users’ confidence in antibiotic allergy knowledge and management. We asked users about their confidence in determining when antibiotics may be related, such as with shared beta lactam ring or side chain structural similarity. At baseline prior to using the AAT, 53 users (52.0%) were moderately confident in making this determination, and 31 (30.3%) were slightly or not at all confident (Figure 6). With the use of the AAT, the number of users who reported being very or extremely confident at this task rose from 18 (17.6%) to 71 (70.0%). After using the AAT, 66.7% and 24.5% of users reported increased and no change in confidence, respectively. The remaining nine (8.9%) users reported a decrease in confidence after using the AAT. These nine users are overrepresented in the high baseline confidence groups: four were moderately confident and five were very or extremely confident at baseline prior to using the AAT.
Figure 6:
Users’ confidence in determining antibiotics’ relatedness without versus with the AAT. Each horizontal bar represents users (n) grouped by their baseline level of confidence. Each bar is separated using colors to show user confidence levels after using the AAT: extremely (gray), very (orange), moderately (blue), slightly (green), and not at all confident (yellow).
We also queried users’ confidence in taking antibiotic allergy histories and how the AAT would change their confidence. At baseline, 15 (14.7%), 54 (53.0%), and 33 (32.3%) users reported being very or extremely, moderately, and slightly or not at all confident, respectively, in taking an antibiotic allergy history (Figure 7). Use of the AAT increased confidence in taking an allergy history for 92 (90.2%) users, while only one user said its use would decrease confidence.
Figure 7:
Users’ confidence in taking antibiotic allergy histories at baseline versus change in confidence at this task with the AAT. Each horizontal bar represents users (n) grouped by baseline level of confidence. Each bar is separated using colors to show user change in confidence after using the AAT: strongly increase (gray), somewhat increase (orange), no change (blue), and somewhat decrease (green).
We measured whether the structured questions in the AAT would encourage users to take more detailed antibiotic allergy histories. Seventy-four (72.5%) users stated that their baseline practice is to confirm the details of an allergy history often or always, and use of the AAT modestly increased this number to 90 (88.2%) users.
One of the goals in creating the AAT was to encourage providers to give beta lactams to patients with low-risk histories of a reaction to beta lactams. We surveyed users on their confidence in giving a needed antibiotic to a patient with a listed allergy to a possibly related antibiotic: 12 (11.8%), 57 (56.0%), and 25 (24.5%) of users were very or extremely, moderately or slightly, and not at all confident, respectively. Ninety-eight (96.1%) users reported the AAT would increase their confidence in this task.
We also surveyed users’ opinions of the AAT (Table 1). Seventy percent of users found the AAT easy to use. All 102 (100%) users reported they would use the AAT in an inpatient setting, while 72 (70.6%) and 53 (52.0%) of users said they would use the AAT in an outpatient and emergency department setting, respectively. We did not ask whether the users, who were recruited from inpatient services, provide care in other settings such as the emergency department, so we cannot draw conclusions about the lower rates of projected emergency department utilization. Before using the AAT, 82 (80.4%) users indicated they would prefer a digital decision support tool to assist with antibiotic allergy assessment and antibiotic selection over other resources, such as a collection of documents or educational sessions providing cross-reactivity charts and other information. After using the AAT, 88 (86.3%) of users continued to state their preferred resource would be a tool such as the AAT.
Table 1:
Users’ assessment of the AAT
| All | Resident | ID Specialist | Pharmacist | |
|---|---|---|---|---|
| Number of users | 102 | 42 | 21 | 39 |
| AAT is easy to use (%) | 72 (70.6) | 27 (64.3) | 18(85.7) | 27 (69.2) |
| Would use AAT in clinical practice (%) | 97 (95.1) | 39 (92.9) | 20 (95.2) | 38 (97.4) |
| AAT is the preferred resource (%) | 88 (86.3) | 37 (88.1) | 20 (95.2) | 31 (79.5) |
| Settings in which user would use AAT* (%) | ||||
| Inpatient | 102 (100.0) | 42 (100.0) | 21 (100.0) | 39 (100.0) |
| Outpatient | 72 (70.6) | 33 (78.6) | 19 (90.5) | 20 (51.3) |
| Emergency Department | 53 (52.0) | 25 (59.5) | 9 (42.9) | 19 (48.7) |
| Other | 6 (5.9) | 4 (9.5) | 1 (4.8) | 1 (2.6) |
| Most useful modality to access AAT (%) | ||||
| Link in the electronic medical record | 56 (54.9) | 20 (47.6) | 13 (61.9) | 23 (59.0) |
| Link on the pharmacy intranet or housestaff manual | 14 (13.8) | 2 (4.8) | 0 (0.0) | 12 (30.8) |
| Mobile application | 26 (25.5) | 16 (38.1) | 8 (38.1) | 2 (5.1) |
| Other | 6 (5.9) | 4 (9.5) | 0 (0.0) | 2 (5.1) |
Users could choose one or more settings.
Discussion
In this study, we describe the testing of a digital decision support tool that uses branching logic to go beyond risk-stratifying penicillin allergy by classifying prior reactions and determining antibiotic cross reactivity. Use of the AAT changed providers’ antibiotic allergy management in test cases by increasing alignment with allergists’ recommendations: users were more likely to give beta lactams in low- and medium-risk situations and practice caution and consult Allergy in high-risk situations. In all the test cases, AAT use reduced unfavorable beta lactam avoidance without allergist consultation. Because allergists were less conservative than the AAT in the high-risk cases, allergist consultation in these scenarios could result in more beta lactam administration in practice. The AAT also increased the confidence of most users, supporting its implementation in clinical practice for multiple provider types, including pediatricians, internists, infectious disease specialists, and pharmacists.
A limitation of this study is the small number of cases. This number was chosen to make participant recruitment and session completion feasible. Out of concern for patient safety, we intentionally overrepresented high-risk scenarios to verify that users’ decisions would either match or be more conservative than the recommendation of the AAT.
A digital decision support tool such as the AAT could expedite antibiotic decisions when immediate allergist consultation might be impractical. The AAT is designed for prescribing providers to swiftly determine the safety of a needed antibiotic, and allergy delabeling may not occur. Full delabeling could, however, be done as an investment in patients’ longitudinal care at a time when it would not delay antibiotic administration. The AAT provides point-of-care guidance for time-sensitive antibiotic management, which complements traditional antibiotic allergy delabeling initiatives.
Barriers to the creation of an institution-specific digital tool can be low, as the REDCap platform is available at many institutions. The AAT’s format allows algorithms to be tailored according to institution-specific resources and allergist practices. Its format also allows for nimble algorithm updates as guidelines are updated and new data emerge, particularly around cephalosporin cross reactivity and non-beta lactam antibiotic allergies.24
In conclusion, we demonstrate that users with AAT access were more likely to choose preferred behaviors in simulated cases. Use of the AAT resulted in reduced beta lactam avoidance in low-risk situations, increased use of test doses in medium-risk situations, and increased caution and allergist consultation in high-risk situations. As most front-line providers may not keep abreast of the evolving evidence base for antibiotic allergies, the availability of an AAT has the potential to positively impact point-of-care antibiotic selection by non-allergists, promote adherence to best-choice antibiotics, and improve antimicrobial stewardship. We plan to implement the AAT into clinical care at our institution and prospectively assess its real-world benefits in ongoing pilot studies.
Supplementary Material
Highlights.
What is already known about this topic?
Most patients labeled with a penicillin allergy can safely receive beta lactams. The benefits of delabeling are well-established. However, most front-line, non-allergist providers are not equipped with the necessary training to assess and manage antibiotic allergies.
What does this article add to our knowledge?
Non-allergist providers made clinical decisions in test cases of beta lactam allergy encounters with a digital, allergist-validated decision support tool. This tool increased non-allergist adherence to allergist recommendations and increased their confidence in antibiotic management.
How does this study impact current management guidelines?
A digital decision support tool for non-allergist providers can improve time-sensitive antibiotic administration for patients with antibiotic allergy labels. Hospitals and providers should consider implementation of accessible and adaptable digital antibiotic allergy decision support tools.
Acknowledgments
The authors wish to thank the following individuals for their support in this study: Laura Bio, PharmD, BCPS, BCIDP, Department of Pharmacy, Infectious Diseases, Stanford Children’s Health; Hayden Schwenk, MD, Division of Infectious Diseases, Department of Pediatrics, Stanford University School of Medicine; Lauren Kushner, MD, Division of Infectious Diseases, Department of Pediatrics, Stanford Children’s Health; Emily Mui, PharmD, BCPS, Department of Pharmacy, Infectious Diseases, Stanford Health Care; Denise Kwong, PharmD, BCPS, Department of Pharmacy, Medicine, Stanford Health Care.
This study was reviewed by the Stanford University School of Medicine Human Research Committee and determined to be exempt/not defined as human subjects research as its primary purpose is institutional quality improvement (Protocols 54216, 54480, 56809).
The Stanford REDCap platform (http://redcap.stanford.edu) is developed and operated by the Stanford Medicine Research IT team. The REDCap platform services at Stanford are subsidized by a) Stanford School of Medicine Research Office, and b) the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through grant UL1 TR001085⤉.
Funding:
This work was supported by the Stanford Children’s Hospital Quality Improvement Grant and in part by the TADA-BSSR training grant from the NIH National Heart, Lung, and Blood Institute (NHLBI, grant number 1T32HL151323).
Abbreviations used:
- CI
Confidence Interval
- AAT
Antibiotic Allergy Tool
Footnotes
The authors have no conflicts to disclose.
The Qualtrics survey tool is a web-based tool for creating and conducting surveys online.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Trubiano JA, Chen C, Cheng AC, Grayson ML, Slavin MA, Thursky KA. Antimicrobial allergy ‘labels’ drive inappropriate antimicrobial prescribing: lessons for stewardship. J Antimicrob Chemother. 2016. Jun;71(6):1715–22. [DOI] [PubMed] [Google Scholar]
- 2.Lee CE, Zembower TR, Fotis MA, Postelnick MJ, Greenberger PA, Peterson LR, et al. The Incidence of Antimicrobial Allergies in Hospitalized Patients: Implications Regarding Prescribing Patterns and Emerging Bacterial Resistance. Arch Intern Med. 2000. Oct 9;160(18):2819. [DOI] [PubMed] [Google Scholar]
- 3.Sousa-Pinto B, Blumenthal KG, Macy E, Pereira AM, Azevedo LF, Delgado L, et al. Penicillin Allergy Testing Is Cost-Saving: An Economic Evaluation Study. Clinical Infectious Diseases. 2021. Mar 15;72(6):924–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rimawi RH, Cook PP, Gooch M, Kabchi B, Ashraf MS, Rimawi BH, et al. The impact of penicillin skin testing on clinical practice and antimicrobial stewardship: Impact of Penicillin Skin Testing. J Hosp Med. 2013. Jun;8(6):341–5. [DOI] [PubMed] [Google Scholar]
- 5.Macy E, Contreras R. Health care use and serious infection prevalence associated with penicillin “allergy” in hospitalized patients: A cohort study. Journal of Allergy and Clinical Immunology. 2014. Mar;133(3):790–6. [DOI] [PubMed] [Google Scholar]
- 6.Charneski L, Deshpande G, Smith SW. Impact of an Antimicrobial Allergy Label in the Medical Record on Clinical Outcomes in Hospitalized Patients. Pharmacotherapy. 2011. Aug;31(8):742–7. [DOI] [PubMed] [Google Scholar]
- 7.Blumenthal KG, Ryan EE, Li Y, Lee H, Kuhlen JL, Shenoy ES. The Impact of a Reported Penicillin Allergy on Surgical Site Infection Risk. Clinical Infectious Diseases. 2018. Jan 18;66(3):329–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Desai S Morbidity in Pregnant Women Associated with Unverified Penicillin Allergies, Antibiotic Use, and Group B Streptococcus Infections. TPJ [Internet]. 2017. [cited 2021 Jul 14];21(1). Available from: http://www.thepermanentejournal.org/issues/2017/6296-streptococcus.html [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vyles D, Chiu A, Routes J, Castells M, Phillips EJ, Kibicho J, et al. Antibiotic Use After Removal of Penicillin Allergy Label. Pediatrics. 2018. May;141(5):e20173466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Macy E, Shu Y-H. The Effect of Penicillin Allergy Testing on Future Health Care Utilization: A Matched Cohort Study. The Journal of Allergy and Clinical Immunology: In Practice. 2017. May;5(3):705–10. [DOI] [PubMed] [Google Scholar]
- 11.Blumenthal KG, Shenoy ES, Hurwitz S, Varughese CA, Hooper DC, Banerji A. Effect of a Drug Allergy Educational Program and Antibiotic Prescribing Guideline on Inpatient Clinical Providers’ Antibiotic Prescribing Knowledge. The Journal of Allergy and Clinical Immunology: In Practice. 2014. Jul;2(4):407–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Blumenthal KG, Shenoy ES, Wolfson AR, Berkowitz DN, Carballo VA, Balekian DS, et al. Addressing Inpatient Beta-Lactam Allergies: A Multihospital Implementation. The Journal of Allergy and Clinical Immunology: In Practice. 2017. May;5(3):616–625.e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Blumenthal KG, Shenoy ES, Varughese CA, Hurwitz S, Hooper DC, Banerji A. Impact of a clinical guideline for prescribing antibiotics to inpatients reporting penicillin or cephalosporin allergy. Annals of Allergy, Asthma & Immunology. 2015. Oct;115(4):294–300.e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ham Y, Sukerman ES, Lewis JS, Tucker KJ, Yu DL, Joshi SR. Safety and efficacy of direct two-step penicillin challenges with an inpatient pharmacist-driven allergy evaluation. allergy asthma proc. 2021. Mar 1;42(2):153–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Stukus D, Chapparo J, Hehmeyer J, Hussain C, Lecerf K, Macias C, et al. Nursing Administered Questionnaire to Identify Pediatric Inpatients Eligible for Dose Graded Penicillin Challenge. Journal of Allergy and Clinical Immunology. 2020. Feb;145(2):AB99. [Google Scholar]
- 16.Trubiano JA, Vogrin S, Chua KYL, Bourke J, Yun J, Douglas A, et al. Development and Validation of a Penicillin Allergy Clinical Decision Rule. JAMA Intern Med. 2020. May 1;180(5):745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Stone CA, Stollings JL, Lindsell CJ, Dear ML, Buie RB, Rice TW, et al. Risk-stratified Management to Remove Low-Risk Penicillin Allergy Labels in the ICU. Am J Respir Crit Care Med. 2020. Jun 15;201(12):1572–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wright A, Rubins D, Shenoy ES, Wickner PG, McEvoy D, Wolfson AR, et al. Clinical decision support improved allergy documentation of antibiotic test dose results. The Journal of Allergy and Clinical Immunology: In Practice. 2019. Nov;7(8):2919–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Chen JR, Tarver SA, Alvarez KS, Wei W, Khan DA. Improving Aztreonam Stewardship and Cost Through a Penicillin Allergy Testing Clinical Guideline. Open Forum Infectious Diseases. 2018. Jun 1;5(6):ofy106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Joint Task Force on Practice Parameters; American Academy of Allergy, Asthma and Immunology; American College of Allergy, Asthma and Immunology; Joint Council of Allergy, Asthma and Immunology. Drug Allergy: An Updated Practice Parameter. Ann Allergy Asthma Immunol. 2010;105(4): 2010;105(4):259–273.e78. [DOI] [PubMed] [Google Scholar]
- 21.Wickham H ggplot2: Elegant Graphics for Data Analysis. New York: Springer-Verlag; 2009. Available at: https://www.springer.com/gp/book/9780387981413. Accessed September 24, 2021. [Google Scholar]
- 22.R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria; 2016. Available at: https://www.r-project.org/. Accessed September 24, 2021. [Google Scholar]
- 23.Champely S, Ekstrom C, Dalgaard P, Gill J, Weibelzahl S, Anandkumar A, et al. Pwr: Basic Functions for Power Analysis; 2020. Available at: https://CRAN.R-project.org/package=pwr. Accessed September 24, 2021. [Google Scholar]
- 24.Khan DA, Banerji A, Blumenthal KG, Phillips EJ, Solensky R, White AA, et al. Drug Allergy: A 2022 Practice Parameter Update. Journal of Allergy and Clinical Immunology. 2022. Sep;S0091674922011861. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.