Antibiotic Allergies in Children: Clinical Impact and Emerging Strategies in Management

The prevalence of self-reported antibiotic allergy is highest among adults; however, a label of antibiotic allergy is commonly acquired during childhood, and the incidence of pediatric antibiotic allergy labels (AALs) has been increasing over the past decade.1 On average 5%–10% of children report an antibiotic allergy, more commonly in older than in younger children.2 Parental recollection of the reaction is commonly lacking in detail and clinician documentation of allergies and adverse reactions is often inadequate, making an assessment of the AAL in adult life difficult.2,3 Furthermore, referral pathways for children who experienced an adverse event while being treated with antibiotics for infection are not well-established, and pediatric specialists in drug allergy are sparse, leading to long-term maintenance of allergy labels instead of prompt assessment. Most pediatric AALs are associated with beta-lactam antibiotics, predominantly penicillins and cephalosporins with 6% of children reporting an AAL to this class.2,4 Reactions to non-beta-lactam antibiotics are less common, reported in approximately 1%–3% of children. The literature regarding non-beta-lactam antibiotics is sparse, with most published data focusing on the management of beta-lactam allergies in children. Therefore, this review will specifically examine beta-lactam allergy in the pediatric population. Where an allergy to a member of the beta-lactam family is suspected, often the whole group is avoided, due to concerns about cross-reactivity. However, in many cases, this is not required and more detailed advice after assessment can be given. For example, historically, patients with an allergy to penicillin were being told that they were also being allergic to first-generation cephalosporins due to cross-reactivity. However, recent evidence indicates that this is rare, with the current understanding being that shared identical side chains between penicillins and cephalosporins determine immunologic cross-reactivity between these antibiotic groups.5 Cross-reactivity due to similar, though not identical, side chains within the cephalosporin group is less understood. However, cross-reactivity between amoxicillin, which shares a similar side chain with cephalexin and cefaclor, has been studied, with rates ranging between 1% and 20%.6 (Figure, Supplemental Digital Content 1, https://links.lww.com/INF/F649).7 For the small percentage of children with a confirmed penicillin or cephalosporin allergy, studies have shown that cross-reactivity between the beta-lactam groups is low and the majority of children only have to avoid the culprit drug and closely structurally related drugs, for example, drugs carrying an identical side chain. In children, like adults, over 90% of reported antibiotic allergies are not true persisting allergies, with "allergic" reactions not reproducible when rechallenged with the same drug.3 This suggests that true allergy is rare and often incorrectly diagnosed, resulting in unnecessary avoidance of key antibiotics. Significant initiatives are emerging to streamline the assessment of AALs to effectively delabel patients with AALs and optimize antibiotic selection in our healthcare system.3,8 Within pediatrics, the rate of serious bacterial infections is low, but the rate of antibiotic use is comparable to adults.9 However, the majority of information available pertains to guidelines and consensus statements for adults.8 Hypersensitivity reactions are generally classified by Gell and Coombs. The mechanisms and clinical manifestations of the 4 reaction types are described in Table 1.10 TABLE 1. - Gell and Coombs Classification of Hypersensitivity Reactions Classifications Mechanism Clinical Manifestations Timing of Reactions Type I (IgE mediated) Allergen IgE binds to mast cells with immune mediators released Anaphylaxis, urticaria, angioedema, rhinitis, bronchial asthma and vomiting Minutes to hours Type II (cytotoxic) IgG and IgM bind to allergen on target cell, complement mediated Neutropenia, thrombocytopenia and hemolytic anemia Variable Type III (immune complex) Antibody binds to soluble antigen, leading to cellular destruction by complement Serum sickness, vasculitis and glomerulonephritis 1–3 wk post exposure Cutaneous manifestations, such as a maculopapular rash or urticaria, are the most prevalent and most reported. Macular, papular or morbilliform rashes occurring several days after the commencement of treatment are more common than immediate reactions. However, these symptoms together with other seemingly benign indications of an allergic reaction in children can often be due to a bacterial or viral infection, leading to the potential misdiagnosis of an allergy to the antibiotic prescribed.11 Delayed reactions commonly occur in patients with intercurrent Epstein-Barr virus and are often not reproducible with deliberate challenge when the patient is well.12 Severe, life-threatening adverse drug reactions such as Steven-Johnson syndrome, toxic epidermal necrolysis, acute generalized exanthematous pustulosis and drug reactions with eosinophilia and systemic symptoms are extremely rare in children representing <0.07%.13 IMPACT OF ANTIBIOTIC ALLERGY LABELS IN CHILD HEALTH While the impact of antibiotic allergy has been extensively studied in adults, there is limited research examining its effects on children's health.1 Within adult populations, approximately 50% of patients with an antibiotic allergy have poorer adherence to prescribing patterns by clinicians, compared to those without such allergies, leading to increased usage of restricted antibiotics.14 Although data are limited in pediatrics, 3 studies reported on this issue, with significantly higher use of second-line antibiotics such as macrolides, quinolones, lincosamides and metronidazole and lower use of beta-lactam antibiotics for antibiotic allergy patients (Table 2).1–3 Such prescribing in adults has been correlated with increased resistance, and it has been demonstrated that this alternative use of antimicrobials that are linked to the development of infections with antibiotic-resistant organisms such as methicillin-resistant Staphylococcus aureus and Clostridium difficile infections. While pediatric studies have not yet reported these findings, the alignment with similar prescribing in adults suggests the possibility of such occurrences in children being reported in the future. TABLE 2. - Summary of Current Pediatric Studies Assessing Antibiotic Allergy in Children as Described in This Review Author (Year)Country Study Design Sample Size/Participants/Age Type of Reported AAL Key Results Diagnostic Assessment Catalano et al (2022)3 Australia (Melbourne) Retrospective cohort study N = 938 children admitted to a tertiary children's hospitalAge: 0–18 years Beta-lactam AAL increased the use of inappropriately prescribed restricted antibiotics. Hospital length of stay was longer for AAL group. None Jones et al (2021)1 USA (Utah) Retrospective cohort study N = 38,906 pediatric patients 30 hospitalized between 2007 and 2017Age: 1 month–17 years Beta-lactam AAL patients were significantly more likely to receive alternative broad-spectrum antibiotics. AAL patients also had higher antimicrobial costs but no differences in costs of hospitalization. None Lucas et al (2018)2 Australia (Western Australia) Retrospective cohort study N = 1672 patients admitted to a tertiary children's hospital over a 1-year periodAge: 0–18 years All antibiotics Prevalence of antibiotic allergy increased with age. Oncology or other specialties were more likely to have AALs than those in general medical or surgical. AALs significantly increased the use of alternative antimicrobial therapy and increased hospital length of stay. None Ibanez et al (2018)15 Spain (Madrid) Prospective multicenter study N = 732 children presenting to allergy departments with a reported penicillin allergyAge: <14 years Penicillin Penicillin allergy was confirmed in 35 (4.8%) of patients, amoxicillin was the trigger in 96.9% of all reactions. Skin testing and sIgE were poor predictors of penicillin allergy. 91% of allergic patients were negative. Direct provocation challenges can be considered safe in nonsevere penicillin allergy in children. SPT, IDT, sIgE, DPT Krusenstjerna-Hafstrom and Rubak (2020)16 Denmark (Aarhus N) Retrospective cohort study N = 141 children with a clinical history of a suspected penicillin class allergyAge: 0–15 years Penicillin Only 4 (2.8%) of patients had a confirmed allergy. There was no correlation between positive skin testing, drug challenge or sIgE. No patients with a positive drug challenge reacted with rash alone on their index reaction. SPT, IDT, sIgE and DPT Labrosse et al (2018)17 Canada (Montreal) Prospective double-blinded study N = 158 children with a reported penicillin allergy referred to a tertiary-care allergy center Penicillin 5 (3.1%) children had an immediate or accelerated (6–48 h) reaction to DPT, only one had a positive skin test. 15 (10.1%) patients who had positive skin tests tolerated the DPT. Routine use of skin testing is unlikely to significantly alter the pretest probability of reaction to challenge. IDT, DPT Mill et al (2016)18 Canada (Montreal) Retrospective and prospective cohort study N = 818 children assessed for a suspected amoxicillin allergy and children followed up post a negative DPTAge: 1–4 years Amoxicillin 770 (94.1%) children tolerated the graded challenge, 17 (2.1%) had mild immediate reactions and 31 (3.8%) nonimmediate reactions. 55 children had subsequently received amoxicillin on follow-up, 49 (89.1%) tolerated it. Graded DPTs provide an accurate and safe confirmatory test for skin-related reactions to amoxicillin. DPT Vezir et al (2016)19 Turkey (Mulga) Prospective cohort study N = 119 children referred to a pediatric allergy clinic with symptoms suggestive of a nonimmediate mild cutaneous reaction to beta-lactamsAge: 2–7.5 years Beta-lactam 4 patients (3.4%) had an urticarial reaction at provocation challenge, no severe reactions were experienced. Omitting skin testing before oral provocation challenge in children with a mild history of reaction, can decrease the burden on allergy clinics and alleviate the discomfort of children. DPT Garcia Rodrigues et al (2019)20 Spain (Ciudad Real) Retrospective observational study N = 97 children referred for a suspected nonimmediate reaction to beta-lactamsAge: <14 years Beta-lactam 14 children (14.4%) had positive reactions, 3 (21.4%) immediate and 8 (57.1%) delayed following initial provocation, 1 (7.1%) immediate and 2 (14.2%) delayed on prolonged course. Provocation testing, if negative should be followed by OPC. DPT, extOPC Jaoui et al (2019)21 France (Paris) Retrospective cohort study N = 456 children referred with suspected mild nonimmediate reactions to beta-lactam antibiotics Beta-lactam 39 patients (8.5%) were positive for DPT. 1 (2.6%) patient had an immediate reaction, 38 (97.4) nonimmediate. 26 (66.6%) reactions occurred earlier or at the same time as the index reactions with the type of reactions were similar to the index reaction in 30 (76.9%) children. Supervised DPT with an extended course is considered safe. DPT, extOPC DPT indicates drug provocation test; extOPC, extended oral provocation challenge; IDT, intradermal testing; sIgE, serum-specific IgE test; SPT, skin prick testing. Alternate prescribing also results in other adverse clinical outcomes. A recent systematic review reported that adult patients with an allergy label experienced a longer duration of therapy, contributing to an increase in length of stay of between 1 and 4 days compared to nonallergic patients.22 Similar findings were observed in children with Catalano et al and Lucas et al both reporting an additional length of stay of 1–2 days for those with a reported allergy.1–3 In addition, higher readmission rates at 4 weeks and 6 months were reported in adults, although this was not reported in children22 ASSESSMENT AND MANAGEMENT OF ANTIBIOTIC ALLERGY LABELS IN CHILDREN It is important to assess drug allergies, as mild clinical symptoms have been shown to be unreliable predictors of genuine allergy.13 Antibiotic allergy delabeling is now recognized as a key component of antimicrobial stewardship programs where the concept is to reduce inappropriate antibiotic use, preserving therapeutic options and decreasing antibiotic resistance. In adults, appropriate and effective antibiotic delabeling, particularly for beta-lactam allergies, can decrease costs, patient length of stay and mortality.22 Developing strategies to address antibiotic allergies and offering ways to delabel these patients are crucial to guarantee optimal future care for patients within our health system.3,22 For both adults and children, delabeling has been extensively reported as a process that can result in revision of the allergy status in up to 95% of patients.22 Assessment of antibiotic allergies is limited in many institutions, with a scarcity of specialists available and allergies frequently only identified during a hospital admission where treatment is indicated.3 However, one study highlighted that even with a well-established drug allergy service, referral for allergy evaluation remained low,3 indicating a potential deficit in education and understanding of antibiotic allergy among clinicians and practitioners. While serum-specific IgE has been shown to be a poor predictor of drug allergy in children,15 measuring serum mast cell tryptase within 2–4 hours of an acute drug reaction is helpful. Mast cell tryptase, a serine protease released during IgE type 1 reactions, has high specificity and positive predictive value, aiding in further testing and management.23 Basophil activation tests, which test for antibiotic-specific IgE bound to basophils, are commercially available and have varying sensitivity and specificity for various drugs. For T cell-mediated reactions, lymphocyte transformation tests or enzyme-linked immunosorbent spot assays can help test for drug-specific T cells. These assays are currently predominantly used in the research setting and not offered for routine clinical assessment.24 During the last 3 years, there has been further development of delabeling strategies, moving from the original gold standard of skin testing and oral provocation challenges to newer approaches such as direct delabeling.25 This includes initiatives such as risk stratification, allowing individuals with a well-documented history of mild reactions to be delabeled by direct oral challenge, where a single or split dose is administered without the need for skin testing. Recent studies use such risk stratification tools to differentiate between patients with low or high risk of reacting to antibiotic challenges as outlined in Figure 1 and recommended by the Australasian Society of Clinical Immunology and Allergy. In high-risk categories such as Steven-Johnson syndrome-toxic epidermal necrolysis, acute generalized exanthematous pustulosis and drug reactions with eosinophilia and systemic symptoms, drug provocation testing should not be performed. However, where low risk is indicated, identified in Figure 1 as either family history only, a benign rash without mucosal involvement or systemic symptoms occurring over a year before assessment,12 the primary aim is to assess whether certain individuals can be safely delabeled through a direct oral challenge, followed by 1-hour observation, without the need for skin testing, or by no testing at all (direct delabeling).22FIGURE 1.: Beta-lactam allergy label: risk stratification and assessment for children 16 years or under.Accurate diagnostic evaluation and diagnosis are crucially important in children and obtaining a detailed history including antibiotic exposure history, detail of index reaction (chronology, symptoms, and timing), other suspected or confirmed allergies and family history of allergy is paramount before undergoing confirmatory testing.26 Evidence guiding management strategies for pediatric patients is evolving with recent data concluding that there is no clear evidence to support the use of skin testing in the investigation of antibiotic allergy in children and that direct oral provocation challenge with the culprit antibiotic in children with a rash can safely be performed (Table 2). However, there is limited evidence of the assessment of children with more severe reactions. One Australian study retrospectively reviewed challenges undertaken at a tertiary pediatric hospital, that is, skin testing and oral provocation challenge or direct oral provocation without skin testing, based on clinician preference alone. Skin testing was more likely to be omitted in younger children, suggesting that clinicians may favor omitting painful skin testing in younger children. Moreover, skin testing was not helpful in assessing antibiotic allergy, even in children with distant moderate to severe reactions because of a low positive predictive value.27 Three European studies concurred, reporting that conventional predictors such as skin testing performed unsatisfactory with a low positive predictive value in the range of 20%.15–17 Direct oral provocation penicillin challenges, without prior skin testing, in controlled settings, for nonsevere beta-lactam allergy in pediatric patients, were shown to be safe in these studies. Nonetheless, they emphasized the critical significance of the assessment of risk before deciding to forego skin testing and accurate assessment of the initial reaction by a medical specialist with experience in allergy was deemed essential. Of note, all the studies were undertaken in allergy centers and further studies, by experienced clinicians who are nonallergists, are needed to move toward assessment within the community setting. The avoidance of skin tests in children is cost-effective and recommended in the pediatric population where fear of such unpleasant and invasive procedures is reported in up to 20% of patients.13,16,17 Four additional studies investigated this aspect of avoiding skin testing by directly conducting graded oral provocation challenges in pediatric cohorts. All of them indicated that over 90% of children were successfully delabeled, with no severe reactions observed among those with confirmed allergy.18–21 Where an allergy is confirmed, current guidelines recommend that an action plan is provided and a medication allergy alert be worn. Adrenaline autoinjectors are not routinely prescribed for drug allergy in Australia, however, this may differ internationally.28 Guidelines on the need for an extended course following the initial challenge are inconsistent, however, there is evidence that up to 15% of patients who have confirmed beta-lactam hypersensitivity are diagnosed during the extended course20 and this could improve the effectiveness of the assessment. Within pediatrics, 2 studies report that an extended course of at least 5 days with the culprit drug should be considered as they found that children within their study reacted with symptoms developing from day 2 to 6. All reactions were mild to moderate with no cases of anaphylaxis.20,21 These findings highlight the safety and effectiveness of the extended course in the diagnosis of pediatric beta-lactam hypersensitivity and the need for current guidelines to be reviewed to reflect this in the assessment process. While assessment and management of AALs is developing and drug challenges now effectively removing allergy labels, a new emerging concern is "relabeling," a term used when labels unfoundedly persist or are renewed after negative allergy testing. This has been discussed and considered most likely due to charting errors and clinician and patient understanding of antibiotic allergy assessment processes. However, the development of strategies addressing this issue is needed to ensure adherence to recommendations post delabeling of patients.29 CONCLUSIONS There is substantial evidence within adult literature indicating that antibiotic allergies are a significant health concern and an economic burden on our healthcare systems. While data in pediatric populations are limited, studies are now emerging that align with these findings. Evidence indicates a rise in reported antibiotic allergies among children, increasing with age and carried into adulthood, yet they are frequently overreported and insufficiently diagnosed, resulting in inappropriate prescriptions and prolonged hospital stays. The current practice of assessing antibiotic allergies in children is heterogeneous. There is an increasing necessity to reassess and refine diagnostic protocols to guarantee safety, precision and cost-effectiveness. Recent literature has progressed from the original gold standard of skin testing plus oral provocation challenge, with skin testing now being reported to be a poor predictor of penicillin allergy, especially within pediatrics, and direct oral provocation challenges considered safe for those with a history of a mild reaction. Although the algorithm in Figure 1 is described as safe and effective in tertiary hospitals and allergy centers, larger pediatric studies are required to evaluate the impact of this newer approach on patient outcomes and to analyze the cost-effectiveness of delabeling. Updating and streamlining current guidelines will potentially encourage a shift toward community assessment, where current studies are sparse. In addition, an emerging concern regarding relabeling in the literature also needs to be addressed29 and this could be mitigated through improved clinician education, accurate documentation and providing comprehensive information to parents following challenges. Optimally, successful identification and evaluation of antibiotic allergies during childhood could alleviate the burden on adult healthcare systems in the future, as they currently bear the greatest impact.