Among 818 patients, 94.1% were tolerant to the graded PC for amoxicillin, 2.1% reacted immediately (within 1 hour), and 3.8% had nonimmediate reactions. Of those tolerating the PC for amoxicillin and requiring subsequent full treatment with amoxicillin, 89.1% tolerated the full treatment.
The 3 groups were comparable regarding most clinical characteristics of the suspected reactions. Error bars indicate standard error.
A child with a history of serum sickness–like reaction (A) developed a nonimmediate macular rash (B) after a graded oral provocation challenge for amoxicillin.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Mill C, Primeau M, Medoff E, et al. Assessing the Diagnostic Properties of a Graded Oral Provocation Challenge for the Diagnosis of Immediate and Nonimmediate Reactions to Amoxicillin in Children. JAMA Pediatr. 2016;170(6):e160033. doi:10.1001/jamapediatrics.2016.0033
Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
The diagnostic properties of a graded provocation challenge (PC) among children presenting with a rash in the course of amoxicillin treatment are currently unknown.
To assess the accuracy and the negative predictive value of the PC in a cohort of children referred with suspected allergy to amoxicillin.
Design, Setting, and Participants
A cohort study was conducted between March 1, 2012, and April 1, 2015, at the allergy clinic of the Montreal Children’s Hospital, Montreal, Quebec, Canada. All children referred with suspected allergy to amoxicillin were approached. In addition, 346 eligible children were followed up to assess reactions to subsequent use of amoxicillin at the time of illness in cases with negative PC results. Data were collected on clinical characteristics, suspected antibiotic exposure, personal and first-degree relatives’ comorbidities, and history of atopy and management of the reaction. Univariate and multivariate logistic regressions were compared to determine factors associated with immediate and nonimmediate reactions to the PC.
All children had a graded PC.
Main Outcomes and Measures
Reactions to the graded PC, the negative predictive value of the PC for nonimmediate reactions, and factors associated with immediate and nonimmediate reactions to the PC.
A total of 818 children were assessed (median age, 1.7 years [interquartile range, 1.0-3.9 years]; 441 [53.9%] male). Among all participants, 770 (94.1%) tolerated the PC, 17 (2.1%) developed mild immediate reactions, and 31 (3.8%) developed nonimmediate reactions. The graded PC had a specificity of 100.0% (95% CI, 90.9%-100.0%), a negative predictive value of 89.1% (95% CI, 77.1%-95.5%), and a positive predictive value of 100.0% (95% CI, 86.3%-100.0%). Among all 346 participants eligible for annual follow-up, 250 (72.3%; 95% CI, 67.2%-76.8%) responded, 55 of whom received subsequent full treatment with amoxicillin; 49 of these 55 participants (89.1%) reported tolerance to subsequent full treatment with amoxicillin, while 6 (10.9%) developed nonimmediate cutaneous reactions. History of a reaction occurring within 5 minutes of exposure was associated with immediate reactions to the PC (adjusted odds ratio = 9.6; 95% CI, 1.5-64.0), while a rash that lasted longer than 7 days (adjusted odds ratio = 4.8; 95% CI, 1.4-16.4) and parental history of drug allergy (adjusted odds ratio = 3.0; 95% CI, 1.3-6.8) were associated with nonimmediate reactions to the PC.
Conclusions and Relevance
Graded PCs provide an accurate and safe confirmatory test for skin-related reactions to amoxicillin. Further studies are required to assess factors associated with the PC outcome groups.
Respiratory tract infections, in particular acute otitis media (AOM), are some of the most frequent infections for which pediatric patients seek medical treatment (15.8 ± 1.5 million clinical encounters for AOM in the United States during 2006-2008).1 This is associated with significant financial and sociological burden in North America. During the past decade, guidelines have been developed that recommend observation in lieu of antibiotic use for patients with AOM under select circumstances. Despite this, recent studies in the United States indicate that 83.1% ± 2.2% of children with a diagnosis of AOM received an antibiotic prescription, mostly amoxicillin (40.9%).1 Up to 10% of children develop rashes while receiving antibiotics.2 Most are diagnosed without further evaluation as allergic to the implicated antibiotic,3 and most continue to avoid the suspect antibiotic in favor of alternatives, which may be less effective, more toxic, and more expensive.
Antibiotic allergies are an immunologically mediated response to a pharmaceutical in a sensitized person and are mediated by IgE, IgG, or T lymphocytes.4 Immediate reactions are usually induced by an IgE-mediated mechanism and occur within the first hour following drug administration.5,6 These reactions usually appear as urticaria, angioedema, rhinitis, bronchospasm, or anaphylaxis (ie, involvement of ≥2 organ systems and/or hypotension).7 Nonimmediate reactions develop at least 1 hour after drug administration and are often induced by a T-lymphocyte–dependent reaction. They typically manifest as a macular or papular rash.8-10 Less common but more severe entities of nonimmediate reactions include serum sickness, acute generalized exanthematous pustulosis, drug-induced hypersensitivity syndrome, Stevens-Johnson syndrome, and toxic epidermal necrolysis.11,12
The relative proportion of established immediate and nonimmediate reactions among children presenting with a rash in the course of amoxicillin treatment is currently unknown. The clinical characteristics and demographic predictors of these reactions are also unclear, and it is not established whether currently standardized skin tests (ie, PRE-PEN [benzylpenicilloyl polylysine injection]) can contribute to the diagnosis of amoxicillin allergy.
Currently, the diagnosis of drug allergy is based on the collaboration of clinical history and confirmatory tests including in vivo tests (skin tests), in vitro tests (specific IgE levels and basophil activation tests), and antigen challenge tests.13 For many antibiotics (including amoxicillin), such tests are not validated, have a high false-negative rate, or are simply not available.14,15 For nonimmediate reactions including macular or papular rashes and serum sickness–like reactions (SSLRs; defined when there is skin and joint involvement),16 skin tests are even less useful given the high false-negative or false-positive results.13,15 Confirmation or exclusion of the diagnosis of drug allergy often depends on drug challenge, which is considered the gold standard.15,17 However, studies suggest that some of these reactions may be attributed to an interaction between the antibiotic and a pathogen.8-10 To our knowledge, there are currently no studies in which all children presenting with a rash during amoxicillin treatment have been assessed through a provocation challenge (PC) and prospectively followed up to assess future treatment.
In 2012, our team initiated a registry aiming to assess the accuracy of a graded PC for the diagnosis of immediate and nonimmediate allergic reactions in all children referred with suspected amoxicillin allergy. In addition, we followed this cohort up to 3 years to assess future use of amoxicillin and the development of allergic reaction on subsequent use in those with a negative graded PC result.
Question How accurate is a graded provocation challenge in children referred with suspected allergy to amoxicillin?
Findings In this cohort study, the graded PC had a specificity of 100.0%, a negative predictive value of 89.1%, and a positive predictive value of 100.0%. History of a reaction occurring within 5 minutes of exposure was associated with immediate reactions to the PC, while a rash that lasted longer than 7 days and parental history of drug allergy were associated with nonimmediate reactions to the PC.
Meaning The graded PC provides an accurate and safe confirmatory test for skin-related reactions to amoxicillin.
All consecutive children referred to the allergy clinic of the Montreal Children’s Hospital, Montreal, Quebec, Canada, with suspected allergy to amoxicillin were approached. Exclusion criteria were any reactions compatible with either Stevens-Johnson syndrome or toxic epidermal necrolysis. Patients with a history compatible with anaphylaxis were not excluded.
This observational study features a retrospective arm and a prospective arm. We characterized the suspected reactions to antibiotics retrospectively and investigated PC outcomes and future use of amoxicillin prospectively. We used the Standards for Reporting of Diagnostic Accuracy reporting guidelines to assess the properties of the graded PC. The study was approved by the McGill Research Ethics Board. All participants provided written informed consent.
After parents consented, a standardized questionnaire (validated previously with a pilot of 40 patients and their parents) was completed by a trained member of our team and the family on the clinical characteristics, comorbidities (including history of atopy and use of medications regularly and during the suspected reaction), suspected antibiotic exposure, and management of the reaction. All children were offered a graded PC (10% of the therapeutic dose of amoxicillin, then 20 minutes later 90% of the therapeutic dose, ie, 50 mg/kg/dose to a maximum of 1.5 g). Children were observed for at least 1 hour after receiving their last dose.
The dose range for the PC was 550 to 1500 mg of amoxicillin depending on the child’s weight. Families of children with negative PC results were given a study email address and telephone contact to report any adverse reaction in the following week or on subsequent treatment.
In cases of positive immediate reactions, skin prick tests and intradermal tests with benzylpenicillin (10 000 U/mL in skin prick test; 1000 U/mL in intradermal test; and 10 000 U/mL in parallel) and PRE-PEN (0.04 mg/mL) were conducted 2 to 3 months later according to previously published protocols.14,18 In addition, trained members of our team contacted families annually for 3 years to assess amoxicillin use and development of future reactions. Given the poor predictive values of skin tests for the diagnosis of nonimmediate reactions and the lack of standardized available intradermal tests for amoxicillin itself, neither was used in this study.
The diagnosis of immediate antibiotic allergy was made only in those with a positive PC result within 1 hour of the PC’s last dose; this was defined as the occurrence only of objective symptoms, including urticaria, angioedema, wheezing, rhinitis, vomiting, diarrhea, protracted abdominal pain, or shock.14,15 A single episode of vomiting was not considered a positive reaction, while severe and repetitive vomiting was classified as positive. Nonimmediate reactions were defined as parents’ report of nonimmediate objective symptoms (as defined earlier) as well as the presence of arthritis or arthralgia at least 1 hour after the PC and up to 1 week afterward. Reactions to the PC were classified as mild when symptoms were limited to the oral mucosa or the skin; severe reactions included cardiovascular or respiratory symptoms or involvement of any 4 systems; and all other reactions were classified as moderate.19 All participants with immediate and nonimmediate reactions were invited for reassessment in the allergy clinic, and a PC with a third-generation oral cephalosporin (cefixime) was offered (10% and then 90% of the therapeutic dose, ie, 8 mg/kg followed by 1 hour of observation).
Descriptive statistics for demographic and clinical characteristics of reactions were calculated using 95% confidence intervals. To assess the percentage and 95% confidence interval for the different groups defined by the graded PC, a multinomial distribution was used for inferences. Univariate and multivariate logistic regression analyses were compared to assess potential confounders and evaluate the effects of sociodemographic predictors (including age and sex), presence of comorbidities (eg, atopic diseases, use of other medications such as nonsteroidal anti-inflammatory drugs and acetaminophen during the reaction and on a daily basis), family history of drug allergy and atopy in first-degree relatives, and clinical characteristics of the reaction (time between initial suspected reaction and challenge, whether this was the first exposure, and clinical symptoms of the suspected reaction, ie, type of rash, presence of arthritis or arthralgia or of symptoms involving the respiratory, digestive, or cardiovascular system) on immediate and nonimmediate reactions to the PC. All statistical analyses were conducted using R version 2.12.0 statistical software (R Foundation).
Between March 1, 2012, and April 1, 2015, 818 participants who were referred with a suspected reaction to amoxicillin consented to an amoxicillin PC (1 patient refused). The median age was 1.7 years (interquartile range, 1.0-3.9 years) and 441 (53.9%) were male. We identified 3 groups of PC outcomes: (1) tolerant to the PC; (2) immediate reaction to the PC (<1 hour); and (3) nonimmediate reaction to the PC (>1 hour) (Figure 1). The 3 groups were comparable regarding sociodemographic characteristics (the majority were male) (Table 1) and most clinical characteristics of the alleged reactions and comorbidities (Table 1 and Figure 2). However, a rash lasting longer than 7 days and parental history of drug allergy were more prevalent in children with nonimmediate reactions, and a history of reaction occurring within 5 minutes was more common in children with immediate reactions to the PC (Table 1).
Among 818 children who underwent the PC, 770 (94.1%) tolerated the PC. Seventeen patients (2.1%) reacted immediately. Immediate reactions were all mild and consisted of hives only. All resolved within a few hours after treatment with second-generation antihistamines. Among the 17 patients with immediate reaction, 5 reacted within 20 minutes after the initial 10% of the dose was administered. Results of skin prick tests and intradermal tests with benzylpenicillin and PRE-PEN tests 2 to 3 months later were positive in only 1 child (5.9%) who had an immediate oral amoxicillin PC reaction. The third group consisted of 31 children (3.8%) who had nonimmediate reactions to the PC. Reactions were all mild and varied from a macular or papular rash with angioedema to an SSLR (Figure 3). The median time between the PC and development of nonimmediate reactions was 12 hours (interquartile range, 5.0-36.0 hours) and 9 of the 31 children with nonimmediate reactions (29.0%; 95% CI, 14.9%-48.2%) reacted more than 24 hours after the PC. Thus, the graded PC detected reactions developing hours to days after initiation of treatment. According to the reaction severity classification by Perry et al,19 all reactions to the PC were mild with hives or a macular or papular rash.
We found no significant predictors in those who had immediate reactions apart from reaction occurring within 5 minutes of exposure (adjusted odds ratio = 9.6; 95% CI, 1.5-64.0). However, the wide confidence interval precludes definitive conclusion on this effect. Our analysis revealed higher odds for nonimmediate PC reaction in those who reported a rash that lasted longer than 7 days (adjusted odds ratio = 4.8; 95% CI, 1.4-16.4) when controlling for age, sex, personal and first-degree relatives’ comorbidities, and history of atopy. In addition, parental history of drug allergy was associated with increased odds of nonimmediate reaction to the PC (adjusted odds ratio = 3.0; 95% CI, 1.3-6.8).
To address the potential for misclassification bias in our sample, 346 patients were eligible for annual follow-up via telephone call. Among them, 250 (72.3%; 95% CI, 67.2%-76.8%) responded, 55 of whom received subsequent full treatment with amoxicillin; 49 of these 55 patients (89.1%) reported that they tolerated subsequent full treatment with amoxicillin and 6 (10.9%) developed delayed reaction localized to the skin, identical to initial reactions (Table 2).
The graded PC had a specificity of 100.0% (95% CI, 90.9%-100.0%), a negative predictive value of 89.1% (95% CI, 77.1%-95.5%), and a positive predictive value of 100.0% (95% CI, 86.3%-100.0%).
All patients with immediate and nonimmediate reactions to the PC with amoxicillin tolerated a subsequent PC with cefixime.
To our knowledge, this is the largest study to assess the use of a graded oral PC in all children presenting with rash due to suspected amoxicillin allergy. We demonstrate that in this population, history and currently standardized skin tests are less useful in predicting amoxicillin allergy. We show that a PC is a useful and safe way to assess these children, including those presenting with SSLRs, and that cefixime may serve as a safe alternative for those who do have reactions.
Our unique study design incorporated an annual follow-up of our cohort and hence we were able to show that in addition to the 3.8% who reacted nonimmediately to the PC, 10.9% among those who have used amoxicillin for a full course again had nonimmediate rashes. Given that all those who reacted to the PC are considered true reactors and that in follow-up only 6 children had reactions among 55 who were tolerant of the PC and subsequently administered full treatment for a suspected bacterial infection (mainly AOM), our findings establish a specificity of 100.0% (95% CI, 90.9%-100.0%), a negative predictive value of 89.1% (95% CI, 77.1%-95.5%), and a positive predictive value of 100.0% (95% CI, 86.3%-100.0%), clearly demonstrating the usefulness of the graded PC for nonimmediate reactions. Given that our cohort data were limited to children who actually required full treatment when sick, we were not able at this point to determine the percentage of false-negative results for our sample and hence we were not able to estimate sensitivity. Although almost a third of nonimmediate reactions occurred more than 24 hours after the PC, it is possible that had we used a PC over a longer interval (5-14 days) for all cases we would have been able to identify an additional percentage of reactors. However, it is also likely that reactors would have experienced more substantial reactions. Further, it is possible that some of the reactions, as previously suggested in the literature,1,2 require an interaction between an infectious agent and the drug and thus would not be captured even with more prolonged PCs.
Amoxicillin allergy is overdiagnosed in children and can only rarely be reproduced by skin testing and/or PC. Our results confirm that the vast majority of children suspected to have amoxicillin allergy are tolerant to amoxicillin. Although others have reported that amoxicillin allergy is overdiagnosed, no studies to our knowledge have assessed all children with a graded PC. Our results are consistent with previously published reports20-22 suggesting that fewer than 10% of rashes occurring while an individual is receiving β-lactam treatment are allergic in nature. Zambonino et al22 found that 1.15% of children reacted immediately and 6.77% reacted late; however, not all of these were determined by a PC. Macy and Ngor23 found that fewer than 5% of those assessed for a suspected allergy to penicillin by an oral amoxicillin PC featuring a therapeutic dose had IgE-mediated reactions. Similar to previous studies assessing PC in those with amoxicillin allergy,20 none of the reactions to the PC were more severe than the initial suspected reaction.
The larger proportion of nonimmediate reactions noted in our study may be due to the difference in catchment population (adults vs pediatrics) or because we administered the PC to all participants regardless of history. In addition, studies suggest that amoxicillin rash may develop due to the interaction with certain viral infections.1,2 Hence, our study that follows future use of amoxicillin during subsequent infection has likely captured amoxicillin-related hypersensitivity reactions that would have been missed in other circumstances.
Similar to other studies,24,25 our findings support the safe use of oral PCs in cases of SSLR. However, in contrast to previous studies, ours is the first study, to our knowledge, to exemplify the usefulness of the graded PC in these cases. Serum sickness–like reactions are characterized by fever, rash, and joint involvement. The pathogenesis is not known but is not associated with circulating immune complexes, hypocomplementemia, or vasculitis.24,26 In contrast with true serum sickness, renal and hepatic involvement is rare. The most frequent finding is of erythema and urticarial lesions that are often migratory. In the series reported by Hebert et al,27 many of the urticarial wheals had dusky to purple centers, which were morphologically suggestive of erythema multiforme. The other primary clinical feature is joint involvement including pain and swelling. A hallmark of SSLR is the benign outcome.27 The most common cause of SSLR was reported to be cefaclor.28 However, with decreased use of cefaclor, recent reports suggest amoxicillin as a principal trigger.29 Our results clearly demonstrate that a graded PC is an appropriate diagnostic test in children with SSLR to amoxicillin.
To our knowledge, there are no currently published positive predictive values for skin tests in the diagnosis of immediate and nonimmediate reactions to amoxicillin. Our results show that currently used standardized tests (PRE-PEN) are not useful for the diagnosis of immediate reactors in PC. In fact, this test was negative in 94.1% (95% CI, 69.2%-99.7%) of immediate reactions. Further, a graded PC was safe in both immediate and nonimmediate reactors. Given the high percentage of children tolerating the PC, it is likely that most immediate or nonimmediate rash reactions in the context of amoxicillin treatment are related to viral infections.20,30,31 The rash may be caused by a direct viral effect or through interaction between the virus and the immune system that triggers a reaction to the drug, as our results in the subgroup followed up and receiving full treatment suggest.11,32,33 Although allergic drug reactions have been considered to be immediate (IgE mediated) or nonimmediate (T-lymphocyte effector mechanisms), accelerated reactions have also been defined (within 2-6 hours after exposure).34 Studies suggest that T lymphocytes are implicated not only in nonimmediate reactions but also in immediate and accelerated reactions.35 The latter may account for the poor sensitivity of skin tests based on IgE-mediated mechanisms in the subgroup with immediate reaction to the PC.
In addition to the clinical significance, the direct health economic benefit of assessing true amoxicillin allergy by PC alone is clear. Among 17 participants in our sample who had an immediate reaction, just 1 (5.9%) had a positive skin test result. We estimated the direct health care system cost of a single skin test to be CaD $170 (US $126; including the cost of physician and nurse services as well as the PRE-PEN ampule).36,37 We approximated the cost of a single PC to be CaD $220 (US $164; including the allergist and nursing services as well as the amoxicillin).36,37 In our sample, 16 of 17 participants with a negative skin test result were eligible for a PC. The cost of a skin test and a PC for those 17 participants is approximately CaD $6420 (US $4776; including 1 patient with a skin test only and 16 with skin tests and PC), as compared with CaD $3740 (US $2782) if all 17 had a PC alone.
Further, it is estimated that there are 1 544 703 cases/year of AOM in Quebec in children aged 0 to 4 years.38 Among those, 95% will receive antibiotic treatment and 41% (601 662 children) will be treated with amoxicillin.1,39 Given that 13% of those receiving amoxicillin will report a reaction,40 78 216 children aged 0 to 4 years in Quebec will have a suspected reaction to amoxicillin in any given year. According to our estimates, 1564 (2.0%) of those will have an immediate reaction. Therefore, for immediate reactions the cost of a skin test for the 94 patients who will have a positive result and the 1470 who will have a negative result and require a PC (assuming 6% of children will have a positive skin test result as observed in our sample) is approximately CaD $589 280 per year (ie, CaD $15 980 + CaD $573 300 [US $438 338, or US $11 887 + US $426 451]), as compared with CaD $344 080 (US $255 945) if all had a PC alone.
Our findings suggest that a rash lasting longer than 7 days as well as reported parental drug allergy might be independently associated with increased odds of a nonimmediate reaction to the PC. Given that most common viral rashes in children last up to 7 days,41 it is possible that longer duration of cutaneous symptoms may represent a true amoxicillin allergy. The association between nonimmediate reactions and reported family history of drug allergy is supported by previous reports on the association between specific genetic loci and other nonimmediate drug allergies, suggesting a familial effect.42,43 It is noteworthy that neither paternal nor maternal drug allergy solely was associated with nonimmediate reactions.
Our study has some potential limitations. Our catchment population recruited from a single pediatric center did not include cases of anaphylaxis and did not include adults. Hence, our results could not be generalized to all cases of suspected amoxicillin allergy but rather might only be generalized to pediatric cases presenting with cutaneous, nonanaphylactic reactions. In addition, it is possible that the reactions that occurred in children tolerating the PC but not a full subsequent course of amoxicillin are related to a viral reaction rather than to a true nonimmediate reaction. The latter might be more likely given that parents reported that subsequent reactions were identical to initial reactions. It is also possible that with more prolonged PC protocols, we would have been able to identify an additional number of nonimmediate reactions.3 However, even such a prolonged PC may not be able to detect cases resulting from an interaction between amoxicillin and a viral infection. Moreover, our PC reached high doses and provoked delayed reactions, confirming its utility for these types of reactions. Questionnaire answers rely on parental report and hence may result in recall and information bias. However, given that all participants were referred with suspected amoxicillin allergy and that the PC was administered to all, it is likely that if such a bias does exist it will be nondifferential and will not affect the observed associations. Given the lack of commercially available minor determinants (penicilloate and penilloate) other than penicillin G, we were not able to conduct these tests. However, recent studies suggest that testing with PRE-PEN and penicillin G vs PRE-PEN and a minor determinant mixture would not substantially increase the number of false-negative cases and that penicillin G can be safely used as an alternative to a minor determinant mixture in diagnosing penicillin allergy.44,45 Finally, given that not all participants required full subsequent treatment with amoxicillin, we were able to determine the percentage of false-negative results for only a subgroup of participants rather than all 818 participants. Hence, we were not able to provide sensitivity estimates.
Our study is the first, to our knowledge, to determine the percentage of immediate and nonimmediate amoxicillin allergic reactions through a graded PC in all children presenting with a suspected amoxicillin-induced rash. Our findings suggest that the presentation may be very similar in those with hypersensitivity and those who are tolerant of amoxicillin. Further, we showed that skin tests are less useful for the diagnosis of immediate reactions to amoxicillin. Finally, our study highlights the usefulness of the graded PC with an appropriate follow-up to diagnose non–life-threatening reactions to amoxicillin. Future studies are required to assess factors associated with specific PC outcomes and in particular should investigate specific associations with genetic markers to accurately determine future risk for antibiotic allergic reactions.
Corresponding Author: Moshe Ben-Shoshan, MD, MSc, Montreal Children’s Hospital, 1001 Boulevard Décarie, Montréal, QC H4A 3J1, Canada (email@example.com).
Accepted for Publication: December 30, 2015.
Published Online: April 4, 2016. doi:10.1001/jamapediatrics.2016.0033.
Author Contributions: Mr Mill and Dr Ben-Shoshan had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Primeau, Ben-Shoshan.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Mill, Dery, Ben-Shoshan.
Critical revision of the manuscript for important intellectual content: Mill, Primeau, Medoff, Lejtenyi, O’Keefe, Netchiporouk, Ben-Shoshan.
Statistical analysis: Mill, Ben-Shoshan.
Obtained funding: Ben-Shoshan.
Administrative, technical, or material support: Mill, Medoff, Lejtenyi, O’Keefe, Dery, Ben-Shoshan.
Study supervision: Primeau, Netchiporouk, Ben-Shoshan.
Conflict of Interest Disclosures: None reported.
Create a personal account or sign in to: