https://orcid.org/0000-0002-7430-2744
Figures
Abstract
Introduction
Patients with a penicillin allergy label (PAL) use more and broader-spectrum antibiotics, have worse health outcomes and cost more to treat than patients without a PAL. A significant proportion of penicillin allergy labels are incorrect; here we review the published evidence on the costs, health-related quality of life, and cost-effectiveness of penicillin allergy testing.
Methods
We conducted a systematic review of published economic evaluations of penicillin allergy testing in accordance with Cochrane guidelines. We searched Medline, Embase, Scopus, Web of Science, EconPapers (RePeC) and the International HTA Database (INAHTA) and included reports of full or partial economic evaluations of costs and/or health benefits of penicillin allergy testing strategies. The outcomes of interest were healthcare resource use, medical costs, and health-related quality of life for patients with a penicillin allergy label and for patients with the label removed, and cost-effectiveness. We evaluated the methodological quality of the studies using a published checklist designed for systematic reviews. The review followed a narrative synthesis.
Results
Thirty-six studies met the inclusion criteria. Most studies analysed the effect of testing on the costs of antibiotic use among patients admitted to hospital with a PAL. Studies measured costs of testing (n = 19); antibiotic medication use (n = 23); adverse reactions with penicillin use (n = 4), alternative antibiotic drugs (n = 3); length of hospital stay (n = 5); subsequent health care use episodes (n = 4); and antibiotic medication use in subsequent care episodes (n = 3). The median cost of skin testing plus oral challenge across six primary costing studies was USD 246 (range: 164, 514), which contrasts with the USD 42–258 range of antibiotic cost savings during the initial hospital admission. Two studies presented evidence that penicillin allergy testing is cost-saving in an outpatient setting over 3.5–4.5 years. One model-based study reported that testing in inpatient settings is cost-saving. No reports on the effect of penicillin allergy testing on health-related quality of life were found and the two cost-effectiveness studies that accounted for this outcome employed the opinion of healthcare professional or an assumption of a common generic value for adverse reactions.
Citation: Mujica-Mota RE, Yang M, King N, Ahmed S, Powell N, Pavitt S, et al. (2025) The cost-effectiveness of penicillin allergy testing: Evidence and gaps from a systematic review. PLoS One 20(12): e0337131. https://doi.org/10.1371/journal.pone.0337131
Editor: Muhammad Shahzad Aslam, Xiamen University - Malaysia Campus: Xiamen University - Malaysia, MALAYSIA
Received: September 6, 2024; Accepted: November 4, 2025; Published: December 19, 2025
Copyright: © 2025 Mujica-Mota et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The detail data extracted from the studies included in the review are now included in Supplementary File 4, and Tables S1-S4.
Funding: This work was part of the ALABAMA project funded by the UK National Institute of Health Research Programme Grants for Applied Research (RP-PG-1214-20007), which sought to conduct a randomised controlled trial and cost-effectiveness analysis of testing for penicillin allergy. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Penicillin and other beta-lactam antibiotics are the most frequent cause of medication induced anaphylaxis [1–3] and 6% of people in England have a record of penicillin allergy [4]. However, at least 9 out of 10 people who believe they are penicillin-allergic are found not to be when tested, and so could safely take penicillins [5,6]. These patients may be receiving less effective antibiotic treatments with additional long-term health risks. Penicillin-allergy records (or label) drive prescribing towards alternative broad-spectrum antimicrobials that contribute to increased antimicrobial resistance (AMR) and may result in poorer patient outcomes. Research has found that macrolide, tetracycline, cephalosporin and quinolone prescribing were all more common in patients with a record of penicillin allergy, compared to those without, and that antimicrobial prescriptions were almost twice as frequent in patients with a penicillin allergy label [4,7].
Antibiotic allergies, including penicillin allergy, have been associated with suboptimal antibiotic therapy, increased antimicrobial resistance, increased length of stay, increased antibiotic-related adverse effects such as Clostridioides difficile infection, intensive care unit (ICU) admission, death, as well as increased treatment cost [8]. Antibiotic regimens deviate from the standard of care in approximately 40% of patients who report a penicillin allergy [9,10]. The costs of the consequences of sub-optimal antibiotic therapy due to reported penicillin allergy are likely significant, reaching far beyond the actual differences in antibiotic costs [11]. Given the significant proportion of incorrect penicillin allergy labels and their impact on healthcare costs and outcomes, it is critical to evaluate the economic implications of penicillin allergy testing.
The standard assessment for penicillin allergy involves skin testing followed by intradermal tests and oral drug provocation testing (oral challenge test). Several recent studies have reported safe and effective testing for low-risk penicillin allergy labels using a direct oral challenge in children and increasingly in adults [12–15]. Due to the prevalence of penicillin allergy labels and the limited capacity of existing specialist allergy clinics, there is increasing interest in expanding access to penicillin allergy testing through provision of non-allergy specialist delivered testing for low-risk patients. Evidence on the costs and benefits of different models of penicillin allergy testing services that helps inform policy decisions and service design and planning is limited, as existing reviews have looked at the associated costs of penicillin allergy labels rather than assessments of testing models and predate the advent of penicillin allergy testing models for low risk patients by non-allergists [16].
The aim of this study was to review the published evidence on the cost-effectiveness of penicillin allergy testing and impact on healthcare costs and health-related quality of life associated with removing a penicillin allergy label. This evidence is intended to inform sustainable service delivery models for increasing access to penicillin allergy testing and de-labelling of non-allergic patients.
Methods
This systematic review was conducted according to the Cochrane guidelines [17], and is reported in line with PRISMA 2020 [18] (S1 File). The review protocol is registered in PROSPERO (CRD42021231848).
Search strategy
The searches were designed and run by an information specialist in November 2020, then rerun in full on 14th November 2023 in Medline, Embase, Scopus, Web of Science, EconPapers (RePeC) and the International HTA Database (INAHTA). Text words and database subject headings were used for two search concepts, penicillin allergy and economic evaluations. The searches were not limited by date or language and were peer-reviewed by a second information specialist using the PRESS checklist [19]. See S2 File for full details of the search strategies.
Eligibility criteria
We included all studies that had: i) patients with a recorded or self-reported penicillin allergy label; ii) used a test to delabel the penicillin allergy; iii) contained one or more comparator groups; iv) measured both costs and health outcomes (full economic evaluation) or either costs or quality of life outcomes (partial economic evaluation). Only peer-reviewed publications written in English were included.
We excluded studies that did not perform a penicillin allergy test, conference abstracts, short notes, comments, editorials and study protocols. Reviews that did not synthesise new cost effectiveness estimates were excluded; however, their reference lists were screened for additional records.
Outcomes
The outcomes of interest were healthcare resource use, medical costs, and health-related quality of life for patients with a penicillin allergy label and for patients with the label removed. In addition, reports on cost-effectiveness of interventions for safely de-labelling individuals with an incorrect penicillin allergy label were reviewed.
Study selection
Two reviewers independently screened the title and abstract of each record using Rayyan software [20]. Full texts of potentially relevant records were then reviewed against the eligibility criteria for study inclusion. Disagreements were resolved by discussion with a third reviewer.
Data extraction
Study characteristics were extracted from each included study, including publication details, the type of study conducted, the population included in the study, the intervention and comparators. Any results reported on health outcomes and cost measures, total or incremental costs were also extracted. Data was independently extracted by two reviewers, any discrepancies were resolved by a third reviewer.
Quality and risk of bias assessment
In a slight deviation from the protocol, we evaluated the methodological quality of the included economic evaluations using a published checklist designed for systematic reviews [21] instead of the CHEERS statement [22], which is a reporting checklist as opposed to a quality assessment tool. Each study was appraised independently by two reviewers, with conflicts resolved by discussion. In addition, the ROBINS-I tool was used to assess risk of bias for each healthcare cost and health outcome reported in observational studies [23].
Data synthesis
The narrative synthesis was structured according to study setting (primary or community, hospital outpatient and inpatient, other), study design (prospective, retrospective or model based), and study type (costing study, cost of illness, cost comparison/minimisation analysis, cost effectiveness analysis). We also recorded whether the study was 1) a cohort study with concurrent or retrospective controls, 2) a before-and-after study of provider organisation policy change or implementation, or 3) an individual counterfactual analysis of case series.
The analysis was based on data as reported in the original publications, except when sufficient data were reported to calculate statistics for unreported outcomes of interest, e.g., mean cost differences on a per patient basis. Only complete case analysis was conducted; i.e., no attempt was made to impute or otherwise adjust for missing data.
Cost estimates were presented in 2024 USD, using the latest purchasing power parities [24] and the US consumer price index [25,26] for converting original published figures reported in different years and currencies.
Intervention effect on healthcare cost estimates from primary data reported by more than one independent study were converted to percentage change units and summarised in terms of minimum, median and maximum mean reported values across studies. Effect estimates for health-related quality of life and health outcomes from multiple independent primary studies were summarised in their original units using the same statistics.
Results
Identified records and included studies
The search produced 2,117 records, with 1,312 unique records after de-duplication. Screening the title and abstracts led to the exclusion of 1,227 records, with 79 records included for full-text screening. This yielded 35 unique articles reporting on 36 studies (unlike the rest of the articles, which reported one economic evaluation study each, one published article reported two economic evaluation studies) meeting the final inclusion criteria. Four review articles were identified [27–30]; we hand searched the references lists of these reviews but no additional articles were identified (Fig 1; see S3 File for list of articles excluded at the full-text screening stage with reasons for exclusion).
Of the 36 included studies, 28 were observational studies and 8 were model-based evaluations. The majority of the observational studies were cost impact analyses of two or more testing strategies (n = 17), followed by studies limited to measuring the costs of testing (n = 6), and cost of therapy (‘cost of illness’) studies (n = 5 studies). Four of the modelling studies were cost-effectiveness studies comparing costs and health outcomes of two or more testing strategies, with the remaining four comparing only costs. The descriptive summary characteristics of observational studies are presented in Table 1 and those of modelling studies in Table 2, and their reported cost measures in S1 Table.
Study quality
The median score across the 36 studies was 9 (range: 3–14) out of 19 items (Table 3). Whilst most studies clearly described the population, the testing strategies under comparison and their study question, 50% (n = 18) of studies adopted a time horizon that was too limited to capture relevant costs and health outcomes to their study question (e.g., until discharge for studies of inpatients or until initial antibiotic course was complete). Only 11% (n = 4) and 44% (n = 16) of studies adequately measured health outcomes and costs respectively (e.g., average antibiotic costs per day of regimen were reported as opposed to average per patient costs over a defined follow-up length), and 67% (n = 24) clearly presented an analysis of costs and/or benefit differences. In 75% (n = 27) of studies, no sensitivity analyses of assumptions used in measuring or analysing costs or benefits were presented, and 36% (n = 13) of studies did not discuss the generalisability of their findings beyond their local study setting. Almost two-thirds of studies reported conflicts of interest or presented no information on whether such conflict existed (Table 3).
Costs measured in the studies
Costs of evaluating Beta-lactam allergy.
Nine studies reported on the costs of testing, of which six evaluated skin tests with oral challenge (ST + OC) [32,34,38,39,42,54](apart from two [32,42], these studies used intradermal injection after negative skin prick testing), and six evaluated direct drug provocation (DPC) testing [31,32,34,35,41,54]. The median cost of ST with oral challenge at an outpatient allergy clinic was USD 246 (range: 164, 514), whilst DPC in an outpatient allergy clinic was USD149 (range: 71, 253; S4 File). A prospective study in children reported that the cost of DPC delivered by a nurse and supervised by a consultant amounted to USD 800 [31]. A second report of DPC delivered by non-allergy specialists reported costs of USD 172 for low-risk patients tested by trained nursing, pharmacy or medical staff in Australia [41] (Table 4; S4 Table).
Three studies investigated patient travel and work absenteeism costs associated with seeking and receiving testing, the costs of which were as large as those of testing itself to the health system [38,39,41]. Two of these studies also reported variation in costs according to test result, which was negligible. They also reported a cost comparison between patients with immediate and patients with delayed reactions, with the latter being 50% larger than the former in adults [39] and negligible in children [38].
Low-risk penicillin allergy de-labelling in an inpatient setting was found to cost less than attending an outpatient clinic following discharge from hospital, whether only the direct healthcare costs or also patient and carer travel costs were included in the analysis [41]. This was due to the high proportion of patients missing their de-labelling appointments (11%) and having separate visits for assessment and testing (17%) in the outpatient setting, whereas inpatients had their allergy assessment, testing and communication of results within the index admission without needing further clinical appointments.
Costs of antibiotic medication use.
Antibiotic costs were the most commonly investigated outcome (reported in 19 of the included studies), with all studies documenting cost savings [7,33,36,40–47,49–56]. The ability to compare their outcomes is low as studies varied widely in their methodology. In cohort studies of the index admission with concurrent or historical controls [41,42,52], the median reduction of antibiotic costs associated with testing was 57% (range: 53,66), and 52% (range 20,66) when extending the study set to include uncontrolled cohort studies [43,46] (Table 5; S4 File). Per patient antibiotic cost savings in these studies, excluding the study in aztreonam users [42], ranged between USD 42–258 (S3 Table).
In contrast, uncontrolled studies of ‘potential impact’ compared the observed antibiotic costs in a group of inpatients undergoing penicillin allergy testing with the counterfactual costs that would have been observed if those patients had not had their PAL tested [7,51,55,56]. The median reported projected reduction in antibiotic costs with skin testing plus oral challenge in the inpatient setting [7,51,55] was 57% (range: 29, 83).
Three studies [36,40,43] reported the effects on costs of antibiotic use in the community over a 1-year period after testing, resulting in a median reduction of 52% (range 42,66). One study reported antibiotic cost savings per patient of USD 22 associated with removing a PAL in 81 low-risk children attending the emergency department [40]. A second study compared the total number of antibiotic courses in the year before with the year after intervention in 250 patients at the imputed cost of the preferred prescribed antibiotic, which resulted in an antibiotic cost per day per patient that was 2.5 times greater in confirmed penicillin allergic patients than de-labelled patients [43]. A third study in 236 patient who obtained a prescription medication from a health plan pharmacy reported USD 48 in cost savings [36].
Costs of healthcare resource utilisation.
Four studies measured the impact of penicillin allergy de-labelling on hospital length of stay (LOS), three were limited to initial admission LOS [33,41,53] and one looked at LOS for admissions up to 6 months after testing [52]; two of these reported sufficient information to estimate the effect on costs [41,52]. Adult inpatients with a type B penicillin allergy diagnosis requiring penicillin antibiotic treatment who were tested with DPC or ST + OC had a reduced LOS (23%, p < 0.05) [52], whilst the costs of admission (emergency department and acute including antibiotics) were reduced by 47% (p = 0.002) among admitted adults who were low-risk and had a DPC [41]. The respective cost savings of these studies were USD 4602 and 8012 (95% CI: 13,050–2975; S4 Table). One of the studies that presented no cost data found no statistically detectable differences in LOS during index hospitalisation nor in length of ICU stay [53].
In a retrospective study of insurance claim records, individuals with a penicillin allergy history attending an outpatient allergy consultation for skin testing plus an oral challenge were subsequently observed to experience 0.55 fewer days in hospital and 0.09 fewer outpatient visits than matched controls annually over 3.6 years [37]. In an economic model of outpatient penicillin allergy testing [63], downstream cost savings offset testing costs, and were driven by an effect of de-labelling of 3.05 (95% CI: 2.3, 3.8) fewer annual primary care visits over 4.5 years. This value was obtained from a published estimate of the excess number of primary care visits in an observational study of a group of patients with a PAL relative to matched controls without a PAL [65]. The estimated minimum de-labelling rate for skin testing with oral challenge to be cost-saving was 30–60% for inpatient and 10–50% for outpatient testing settings, and larger than for DPC (15–30% and 5–30%, respectively [63]).
Two publications [46,54] (three evaluated service models) reported the costs of PAL testing and antibiotic costs but lacked a control or comparable group and were based on hypothetical savings, resulting in a median net costs of USD 39 (range, −39, 167). One matched controlled study [37] reported sufficient data to calculate the costs of testing for PAL and healthcare use over a 3.6 follow-up period after testing, resulting in cost savings of USD 8,811 (Table 6).
Health outcomes.
Two studies reported health outcomes [52,53]. In a retrospective study of adults undergoing inpatient hematopoietic stem cell transplantation in a single centre, no detectable difference in 90-day post-transplant incidence of Clostridioides difficile nor mortality was observed (11% before and 7% after, p = 0.34, in both outcomes) after mandatory implementation of ST + OC [53](Risk of bias: Severe, before-cohort vs after-cohort study without concurrent control). In a prospective case series study of 70 ST + OC tested adult inpatients requiring penicillin, their rate of 6-month hospital readmission was 43% versus 61% in matched controls (p < 0.05), and the rate of newly acquired multi-resistant organisms was 4 and 7% (p = 0.50), respectively [52](Risk of bias: Moderate, cohort study with concurrently controls).
Health-related Quality of life.
No study was found that reported evidence on the health-related quality of life of individuals with a PAL. Only a couple of studies, both based on decision tree models, reported results in terms of quality-adjusted life years (QALYs). One of them elicited health state utility values of treatment cure and failure in Staphylococcus aureus endocarditis under skin testing with and without adverse reaction to penicillin or toxicity from second-line therapy (vancomycin), and under no testing with and without vancomycin toxicity, from a nurse providing care for such inpatients [58]. The other was a study of skin testing in S. aureus bacteremia which obtained values for a ‘post-septic episode with no other issue’, ‘disutility for adverse event/adverse drug reaction’ independent of antimicrobial regimen and ‘disutility for readmission’ from previous studies in patients with bacteremia [60].
Cost-effectiveness of de-labelling individuals with an incorrect penicillin allergy label.
Four studies reported results combining differences in costs and health benefits in cost-effectiveness analyses, all of them using decision trees. One study compared skin prick testing (SPT) with no SPT over a 1-year time horizon, including the inpatient and outpatient costs of antibiotic therapy and care for complicated infections and health-related quality of life losses associated with readmissions and adverse events [60]. A second study modelled outcomes of SPT-guided management of patients with S. aureus infective endocarditis followed up until hospital discharge and found it to have lower expected costs and higher health-related quality of life (utilities) than the alternative of not testing and treating patients with vancomycin [58]. In contrast, SPT-guided vancomycin prophylaxis for cardiovascular surgery patients with a PAL had incremental costs per anaphylactic case avoided above USD 400,000 and were not recommended for routine use instead of oral history-based management [62]. The fourth study evaluated ST + OC-guided intrapartum Group B Streptococcus (GBS) prophylaxis in pregnant women, in which the costs were included for penicillin allergy and GBS sensitivity testing and antibiotics, resulting in a cost per additional patient appropriately treated of USD 1360 relative to usual treatment according to the PAL [64].
Discussion
Of the 36 studies included in this review, the majority met less than half the methodology quality appraisal items required to be considered a robust cost-effectiveness analysis. Most were cost analysis studies, focusing on antibiotic and direct healthcare costs before and after testing. Only a few studies evaluated the impact of testing and de-labelling outside a hospital setting, and most of these included small inpatient populations, were confined to single specialties (e.g., joint replacement surgery [61], and breast surgery [44]), and typically conducted in a single centre. Most studies were conducted over a short period of time with only a few looking at impacts beyond a 1-year time period.
This review highlights several potential economic benefits of penicillin allergy testing, primarily its impact on antibiotic acquisition cost during the index episode of care. Three studies explored the impact of allergy testing on LOS, and two report associated cost reductions to the healthcare sector of 23–48% due to LOS reductions over an acute admission alone or including repeat episodes over a 6 months period after testing. If confirmed by independent studies, this magnitude of effects on costs may offset the costs of the test, especially among low-risk patients timely screened for DPC after admission. There were no studies that reported on the patient benefits of an earlier discharge from hospital, e.g., due to switching from complex antibiotics to simpler oral penicillin administration [49,55].
A couple of studies documented the cost of immediate and delayed reactions to the test, but there were no studies reporting the economic consequences of any potential reduction in the incidence of adverse events associated with using narrower spectrum agents due to de-labelling, e.g., C. difficile infection. The impact of penicillin allergy delabelling on antimicrobial acquisition costs in the community was measured in three studies; one followed children for a median of 12 months after being de-labelled at the ED [40]; another recorded costs of adult inpatients 12 months after testing [43]; and a third followed an outpatient cohort of health plan members over a 1-year period (20% loss to follow-up) after testing [36]. Neither the costs of any downstream health care service use for subsequent or relapsing episodes of infection, nor their associated health impacts were measured.
The median cost of skin testing plus oral challenge across six primary costing studies was USD 246 (range: 164, 514), which was more than the associated median antibiotic acquisition cost reductions during the index hospital episode of care (i.e. USD 74, range: USD 42–258, excluding a study in the highly selective population of patients on aztreonam [42]). This finding is also evident in the three studies that have measured the cost of testing and antibiotic use during the index hospital admission. Our study also summarises the reported unit costs of different testing strategies according to whether testing is led by an allergy specialist or non-allergy specialists in an inpatient or outpatient setting. Our review reflects the currently evolving recognition of the safety and feasibility of providing non-allergy specialist testing for penicillin allergy using DPC led by trained nurses or pharmacists [66]. Although the information presented in published studies did not always allow us to determine whether tested patients were of low, moderate or high risk, we found several reports involving high-risk populations, for example those undergoing hematopoietic stem cell transplant or the immunocompromised, where SPT and oral challenge was done by allergists. Staicu et al. [56] reported a potential saving of $152 if the testing was done by a non-allergist. A multi-test strategy of skin test plus an oral challenge was used in 18 out of 27 testing strategies, with nine of those having a three-step process (skin test, intradermal test, and an oral challenge). Skin testing alone was used in one study, and some studies estimated the cost of testing being carried out by non-allergists [31,41]. While the summarised unit costs of DPC are lower than skin test plus an oral challenge overall, there is one report of low-risk patients tested by nurse-led DPC plus remote doctor consultation (telemedicine) at a cost of USD 747, which is above most of the other estimates of skin test plus an oral challenge reported by primary costing studies and suggests results driven by differences in costing methodolgy rather than true costs.
Our study adds to a previous systematic review of the costs associated with a self-reported penicillin allergy label [16], by focusing instead on studies evaluating the effect of penicillin allergy testing interventions. Like the previous review, we found that the majority of studies are observational in nature and focused on inpatient populations. Our study adds new evidence that de-labelling may reduce costs of hospital stay during the index episode of care, but the magnitude of this effect may still not fully offset the costs of penicillin allergy testing and studies that track outcomes over 3.5 years after testing may be required to capture important cost savings from hospital stays and outpatient attendances [37].
The heterogeneous findings across studies regarding the effect of penicillin allergy testing on the costs of antibiotic use partly reflects the varying quality in this literature. For example, in measuring the impact of testing, study designs vary from measuring antibiotic use in uncontrolled case series of tested patients or controlled tested cohorts. Several studies report the difference between the costs of the observed antibiotic use in de-labelled patients and the ‘theoretical’ amount of antibiotic that would have been used by these patients had they retained their PAL (i.e., cost avoidance studies). Furthermore, some of these only measured costs for the subset of patients who are switched to a preferred penicillin or beta-lactam after testing as opposed to the whole sample of patients undergoing testing [45,49].
We found a number of key areas for future research. In addition to health-related quality of life, there is lack of evidence on long-term benefits of appropriate de-labelling to patients from avoiding delayed treatment of serious infections, e.g., sepsis or meningococcal meningitis. Further, none of the studies identified in this review sought to account for the population health benefits of penicillin allergy delabelling from reducing antimicrobial resistance; those that recorded antibiotic prescriptions over a one year or longer follow-up period post-testing found results consistent with reductions of 6% (p = 0.5 [43]), 14% (p not available [37]), and 28% (p = 0.0001 [36]), which suggest such potential benefits. Measurement of costs associated with rare events leading to ICU care was almost absent from the reviewed studies, reflecting the single-site nature of most of them. Multicentre controlled studies are needed to generate evidence on or adequate surrogates of these key outcomes, which would then serve to inform cost-effectiveness and cost-utility studies that account for outcomes relevant to patients as opposed to single provider institutions as in most of the received literature.
There are limitations of this review. First, we did not consider studies published in languages other than English. In view of the growing awareness of the impact of penicillin allergy labels on patient and public health we are likely to miss important emerging international evidence. Second, we excluded studies that reported quantities of resource use, such as length of hospital say, without providing the associated costs. Future reviews may seek to expand the criteria for inclusion to non-economic studies that report key outcomes, such as length of hospital stays, and apply a notional unit cost to derive more precise and informative measures of economic impact than presented here. Third, there is heterogeneity in reporting and costing methods used across studies, with some studies of ST only costing the test kit, whilst other studies accounted for the costs of staff performing the test, test kits and materials for in vivo and in vitro (intradermal) tests, cost of allergist and consultant time and capital costs of using the consultation room. We have mitigated against biases by excluding studies that did not account for the costs of consultation but observed variation across studies that may reflect differences in methods as opposed to variation in a consistent measure of unit cost of skin testing and an oral challenge or DPC test.
Despite the limitations, the reviewed evidence suggests that the economic case for penicillin allergy delabelling may be better made by carefully selecting patients on high cost antibiotics, such as admitted patients on aztreonam. Similarly, outpatients may be selected for testing among frequent users of antibiotics.
Our study findings are also consistent with the view that efficient service delivery models that rely on trained non-allergy specialists are required for the adoption of penicillin allergy testing to appeal to service managers and payers. Such adoption would be easier and less costly for hospitals with robust electronic health record systems, telemedicine services and established networks with allergy testing centres. Whilst pharmacists may have the training most suitable to lead in this role of expanding access to testing services, they would cost more than suitably trained nurses or healthcare assistants.
Conclusions
There is limited evidence that penicillin allergy testing results in antibiotic costs savings, as published studies are observational, often uncontrolled and poorly reported. Whilst there is evidence of reduced antibiotic consumption in the community, little is known about the associated cost impact of de-labelling. Penicillin allergy testing is unlikely to recoup its costs within the first year after testing. There is emerging observational evidence to suggest significant cost savings to the health care system from reduced outpatient attendances and inpatient stays three to four years after testing. No evidence exists on the health-related quality of life impact of penicillin allergy testing. In order to identify optimal penicillin allergy testing models and prove cost-effectiveness, randomised controlled trials with sufficiently long follow-ups and power to detect meaningful impacts to patients and national health services are urgently required, particularly in high risk and resource-constrained settings. Non-allergy specialist delivery models may offer an affordable way to expand allergy testing service beyond the limited capacity of allergy testing centres.
Supporting information
S3 File. Studies excluded at full text screening stage with reasons.
https://doi.org/10.1371/journal.pone.0337131.s003
(DOCX)
S2 Table. Data extraction of costs of testing.
https://doi.org/10.1371/journal.pone.0337131.s006
(DOCX)
S3 Table. Data extraction of antibiotic costs.
https://doi.org/10.1371/journal.pone.0337131.s007
(DOCX)
Acknowledgments
We want to thank Declan Kohl for thorough and diligent research assistance. Declarations: This work was part of the ALABAMA project, which sought to conduct a randomised controlled trial and cost-effectiveness analysis of testing for penicillin allergy.
References
- 1. Renaudin J-M, Beaudouin E, Ponvert C, Demoly P, Moneret-Vautrin D-A. Severe drug-induced anaphylaxis: analysis of 333 cases recorded by the Allergy Vigilance Network from 2002 to 2010. Allergy. 2013;68(7):929–37. pmid:23741979
- 2. Dhopeshwarkar N, Sheikh A, Doan R, Topaz M, Bates DW, Blumenthal KG, et al. Drug-induced anaphylaxis documented in electronic health records. J Allergy Clin Immunol Pract. 2019;7(1):103–11. pmid:29969686
- 3. Worm M, Moneret-Vautrin A, Scherer K, Lang R, Fernandez-Rivas M, Cardona V, et al. First European data from the network of severe allergic reactions (NORA). Allergy. 2014;69(10):1397–404. pmid:24989080
- 4. West RM, Smith CJ, Pavitt SH, Butler CC, Howard P, Bates C, et al. “Warning: allergic to penicillin”: association between penicillin allergy status in 2.3 million NHS general practice electronic health records, antibiotic prescribing and health outcomes. J Antimicrob Chemother. 2019;74(7):2075–82. pmid:31225607
- 5. Borch JE, Andersen KE, Bindslev-Jensen C. The prevalence of suspected and challenge-verified penicillin allergy in a university hospital population. Basic Clin Pharmacol Toxicol. 2006;98(4):357–62. pmid:16623858
- 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;31(8):742–7. pmid:21923600
- 7. Blumenthal KG, Lu N, Zhang Y, Li Y, Walensky RP, Choi HK. Risk of meticillin resistant Staphylococcus aureus and Clostridium difficile in patients with a documented penicillin allergy: population based matched cohort study. BMJ. 2018;361:k2400. pmid:29950489
- 8. Krah NM, Jones TW, Lake J, Hersh AL. The impact of antibiotic allergy labels on antibiotic exposure, clinical outcomes, and healthcare costs: a systematic review. Infect Control Hosp Epidemiol. 2021;42(5):530–48. pmid:33059777
- 9. Arnold A, Coventry LL, Foster MJ, Koplin JJ, Lucas M. The burden of self-reported antibiotic allergies in health care and how to address it: a systematic review of the evidence. J Allergy Clin Immunol Pract. 2023;11(10):3133-3145.e3. pmid:37352931
- 10. Powell N, West R, Sandoe JAT. The impact of penicillin allergy de-labelling on the WHO AWaRe antibiotic categories: a retrospective cohort study. J Hosp Infect. 2021;115:10–6. pmid:33895164
- 11. Macy E, Adkinson NF Jr. The evolution of our understanding of penicillin allergy: 1942-2022. J Allergy Clin Immunol Pract. 2023;11(2):405–13. pmid:36116763
- 12. Rose MT, Slavin M, Trubiano J. The democratization of de-labeling: a review of direct oral challenge in adults with low-risk penicillin allergy. Expert Rev Anti Infect Ther. 2020;18(11):1143–53. pmid:32662696
- 13. Mustafa SS, Conn K, Ramsey A. Comparing direct challenge to penicillin skin testing for the outpatient evaluation of penicillin allergy: a randomized controlled trial. J Allergy Clin Immunol Pract. 2019;7(7):2163–70. pmid:31170542
- 14. Chua KYL, Vogrin S, Bury S, Douglas A, Holmes NE, Tan N, et al. The Penicillin allergy delabeling program: a multicenter whole-of-hospital health services intervention and comparative effectiveness study. Clin Infect Dis. 2021;73(3):487–96. pmid:32756983
- 15. Copaescu AM, Vogrin S, James F, Chua KYL, Rose MT, De Luca J, et al. Efficacy of a clinical decision rule to enable direct oral challenge in patients with low-risk penicillin allergy: the PALACE randomized clinical trial. JAMA Intern Med. 2023;183(9):944–52. pmid:37459086
- 16. Mattingly TJ 2nd, Fulton A, Lumish RA, Williams AMC, Yoon S, Yuen M, et al. The cost of self-reported penicillin allergy: a systematic review. J Allergy Clin Immunol Pract. 2018;6(5):1649-1654.e4. pmid:29355644
- 17.
Lefebvre C, Glanville J, Briscoe S, Featherstone R, Littlewood A, Metzendorf MI. Searching for and selecting studies. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, eds. Cochrane handbook for systematic reviews of interventions. Cochrane; 2024.
- 18. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. pmid:33782057
- 19. McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS peer review of electronic search strategies: 2015 guideline statement. J Clin Epidemiol. 2016;75:40–6.
- 20. Ouzzani M, Hammady H, Fedorowicz Z, Elmagarmid A. Rayyan-a web and mobile app for systematic reviews. Syst Rev. 2016;5(1):210. pmid:27919275
- 21. Evers S, Goossens M, de Vet H, van Tulder M, Ament A. Criteria list for assessment of methodological quality of economic evaluations: consensus on Health Economic Criteria. Int J Technol Assess Health Care. 2005;21(2):240–5. pmid:15921065
- 22. Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D. Consolidated health economic evaluation reporting standards (CHEERS) statement. BMJ. 2013;346.
- 23. Sterne JA, Hernán MA, Reeves BC, Savović J, Berkman ND, Viswanathan M, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919.
- 24.
Purchasing power parities (PPP). Conversion rates. OECD; 2017. https://doi.org/10.1787/1290ee5a-en
- 25. Ha J, Kose MA, Ohnsorge F. One-stop source: a global database of inflation. J Intern Money Finan. 2023;137:102896.
- 26.
Bureau of Labor Statistics. Consumer Price Index. All Urban Consumers (CPI-U). Summary of annual and semi-annual indexes. Washington, DC: US Department of Labor, Bureau of Labor Statistics; 2025. https://www.bls.gov/regions/mid-atlantic/data/consumerpriceindexannualandsemiannual_table.htm
- 27. Parikh P, Patel NC, Trogen B, Feldman E, Meadows JA. The economic implications of penicillin allergy. Ann Allergy Asthma Immunol. 2020;125(6):626–7. pmid:32768634
- 28. Liu MY, Challa M, McCoul ED, Chen PG. Economic viability of penicillin allergy testing to avoid improper clindamycin surgical prophylaxis. Laryngoscope. 2023;133(5):1086–91. pmid:35904127
- 29. Lee G-C. CORR Insights(®): is vancomycin-only prophylaxis for patients with penicillin allergy associated with increased risk of infection after arthroplasty?. Clin Orthop Relat Res. 2016;474(7):1607–9. pmid:26818598
- 30. Gugkaeva Z, Crago JS, Yasnogorodsky M. Next step in antibiotic stewardship: pharmacist-provided penicillin allergy testing. J Clin Pharm Ther. 2017;42(4):509–12. pmid:28504314
- 31. Allen HI, Gillespie P, Vazquez-Ortiz M, Murphy AW, Moylett EM. A cost-analysis of outpatient paediatric penicillin allergy de-labelling using telemedicine. Clin Exp Allergy. 2021;51(3):495–8. pmid:33170984
- 32. Blumenthal KG, Li Y, Banerji A, Yun BJ, Long AA, Walensky RP. The cost of penicillin allergy evaluation. J Allergy Clin Immunol Pract. 2018;6(3):1019-1027.e2. pmid:28958738
- 33. Englert E, Weeks A. Pharmacist-driven penicillin skin testing service for adults prescribed nonpreferred antibiotics in a community hospital. Am J Health-Syst Pharm. 2019;76(24):2060–9.
- 34. Ferré-Ybarz L, Salinas Argente R, Gómez Galán C, Duocastella Selvas P, Nevot Falcó S. Analysis of profitability in the diagnosis of allergy to beta-lactam antibiotics. Allergol Immunopathol (Madr). 2015;43(4):369–75. pmid:25087091
- 35. Jaoui A, Delalande D, Siouti S, Benoist G, Sève E, Ponvert C, et al. Safety and cost effectiveness of supervised ambulatory drug provocation tests in children with mild non-immediate reactions to beta-lactams. Allergy. 2019;74(12):2482–4. pmid:31087349
- 36. Macy E. Elective penicillin skin testing and amoxicillin challenge: effect on outpatient antibiotic use, cost, and clinical outcomes. J Allergy Clin Immunol. 1998;102(2):281–5. pmid:9723673
- 37. Macy E, Shu Y-H. The effect of penicillin allergy testing on future health care utilization: a matched cohort study. J Allergy Clin Immunol Pract. 2017;5(3):705–10. pmid:28366717
- 38. Sobrino M, Muñoz-Bellido FJ, Macías E, Lázaro-Sastre M, de Arriba-Méndez S, Dávila I. A prospective study of costs associated with the evaluation of β-lactam allergy in children. J Pediatr. 2020;223:108-113.e2. pmid:32532647
- 39. Sobrino-García M, Muñoz-Bellido FJ, Moreno E, Macías E, Gracia-Bara MT, Laffond E, et al. A Comprehensive prospective study of the costs associated with evaluation of ß-Lactam allergy. J Investig Allergol Clin Immunol. 2021;31(1):52–7. pmid:31599727
- 40. Vyles D, Chiu A, Routes J, Castells M, Phillips EJ, Kibicho J. Antibiotic use after removal of penicillin allergy label. Pediatrics. 2018;141(5).
- 41. Brusco NK, Bury S, Chua KYL, Vogrin S, Holmes NE, Trubiano JA. Penicillin Allergy Delabeling Program: an exploratory economic evaluation in the Australian context. Intern Med J. 2023;53(1):74–83. pmid:34523209
- 42. Chen JR, Tarver SA, Alvarez KS, Wei W, Khan DA. Improving aztreonam stewardship and cost through a penicillin allergy testing clinical guideline. Open Forum Infect Dis. 2018;5(6):ofy106. pmid:29977963
- 43. du Plessis T, Walls G, Jordan A, Holland DJ. Implementation of a pharmacist-led penicillin allergy de-labelling service in a public hospital. J Antimicrob Chemother. 2019;74(5):1438–46. pmid:30753497
- 44. Fan B, Udeh B, Quinn N, Bernard SL, Grobmyer SR, Valente SA. Delabeling penicillin allergy in breast surgery patients: a cost analysis. Am Surg. 2020;86(2):e75–8. pmid:32106918
- 45. Foolad F, Berlin S, White C, Dishner E, Jiang Y, Taremi M. The impact of penicillin skin testing on aztreonam stewardship and cost savings in immunocompromised cancer patients. Open Forum Infect Dis. 2019;6(10):ofz371. pmid:31660339
- 46. Forrest DM, Schellenberg RR, Thien VV, King S, Anis AH, Dodek PM. Introduction of a practice guideline for penicillin skin testing improves the appropriateness of antibiotic therapy. Clin Infect Dis. 2001;32(12):1685–90. pmid:11360207
- 47. Harmon S, Richardson T, Simons H, Monforte S, Fanning S, Harrington K. The clinical and financial impact of a pharmacist-driven penicillin skin testing program on antimicrobial stewardship practices. Hosp Pharm. 2020;55(1):58–63. pmid:31983768
- 48. Heil EL, Bork JT, Schmalzle SA, Kleinberg M, Kewalramani A, Gilliam BL, et al. Implementation of an infectious disease fellow-managed penicillin allergy skin testing service. Open Forum Infect Dis. 2016;3(3):ofw155. pmid:27704011
- 49. Jones BM, Bland CM. Penicillin skin testing as an antimicrobial stewardship initiative. Am J Health Syst Pharm. 2017;74(4):232–7. pmid:28179249
- 50. Jones BM, Avramovski N, Concepcion AM, Crosby J, Bland CM. Clinical and economic outcomes of penicillin skin testing as an antimicrobial stewardship initiative in a community health system. Open Forum Infect Dis. 2019;6(4):ofz109. pmid:30968057
- 51. King EA, Challa S, Curtin P, Bielory L. Penicillin skin testing in hospitalized patients with β-lactam allergies: Effect on antibiotic selection and cost. Ann Allergy Asthma Immunol. 2016;117(1):67–71. pmid:27211057
- 52. Li J, Shahabi-Sirjani A, Figtree M, Hoyle P, Fernando SL. Safety of direct drug provocation testing in adults with penicillin allergy and association with health and economic benefits. Ann Allergy Asthma Immunol. 2019;123(5):468–75. pmid:31419490
- 53. Modi AR, Majhail NS, Rybicki L, Athans V, Carlstrom K, Srinivas P, et al. Penicillin allergy skin testing as an antibiotic stewardship intervention reduces alternative antibiotic exposures in hematopoietic stem cell transplant recipients. Transpl Infect Dis. 2019;21(6):e13175. pmid:31539459
- 54. Ramsey A, Mustafa SS, Holly AM, Staicu ML. Direct challenges to penicillin-based antibiotics in the inpatient setting. J Allergy Clin Immunol Pract. 2020;8(7):2294–301. pmid:32156611
- 55. 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. J Hosp Med. 2013;8(6):341–5. pmid:23553999
- 56. Staicu ML, Holly AM, Conn KM, Ramsey A. The use of telemedicine for penicillin allergy skin testing. J Allergy Clin Immunol Pract. 2018;6(6):2033–40. pmid:29751152
- 57. Bragg JT, Sudah SY, Moverman MA, Puzzitiello RN, Pagani NR, Menendez ME. Preoperative allergy testing for patients reporting penicillin and cephalosporin allergies is economically justified in preventing infection after total shoulder arthroplasty. J Shoulder Elbow Surg. 2023;32(1):186–91. pmid:36108882
- 58. Dodek P, Phillips P. Questionable history of immediate-type hypersensitivity to penicillin in Staphylococcal endocarditis: treatment based on skin-test results vers-us empirical alternative treatment--A decision analysis. Clin Infect Dis. 1999;29(5):1251–6. pmid:10524971
- 59. Lee OC, Cheng DC, Paul JL, Ross BJ, Hawkins BJ, Sherman WF. Economic burden of patient-reported penicillin allergy on total hip and total knee arthroplasty. J Arthroplasty. 2021;36(9):3067–72. pmid:34053750
- 60. Mattingly TJ 2nd, Meninger S, Heil EL. Penicillin skin testing in methicillin-sensitive staphylococcus aureus bacteremia: a cost-effectiveness analysis. PLoS One. 2019;14(1):e0210271. pmid:30615655
- 61. Pagani NR, Moverman MA, Puzzitiello RN, Menendez ME, Barnes CL, Kavolus JJ. Preoperative allergy testing for patients reporting penicillin and cephalosporin allergies is cost-effective in preventing infection after total knee and hip arthroplasty. J Arthroplasty. 2021;36(2):700–4. pmid:32933797
- 62. Phillips E, Louie M, Knowles SR, Simor AE, Oh PI. Cost-effectiveness analysis of six strategies for cardiovascular surgery prophylaxis in patients labeled penicillin allergic. Am J Health Syst Pharm. 2000;57(4):339–45. pmid:10714971
- 63. 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. Clin Infect Dis. 2021;72(6):924–38. pmid:32107530
- 64. Thao V, Sharpe EE, Dholakia R, Ahn HH, Moriarty JP, Borah BJ, et al. Evaluating the cost-effectiveness of testing pregnant women for penicillin allergy. PLoS One. 2023;18(1):e0280151. pmid:36662778
- 65. Su T, Broekhuizen BDL, Verheij TJM, Rockmann H. The impact of penicillin allergy labels on antibiotic and health care use in primary care: a retrospective cohort study. Clin Transl Allergy. 2017;7:18. pmid:28593040
- 66. Powell N, Stephens J, Kohl D, Owens R, Ahmed S, Musicha C, et al. The effectiveness of interventions that support penicillin allergy assessment and delabeling of adult and pediatric patients by nonallergy specialists: a systematic review and meta-analysis. Int J Infect Dis. 2023;129:152–61. pmid:36450321