Performance of Risk Stratification Tools for Cephalosporin Allergy Testing
Introduction
Cephalosporin allergy labels (CALs) are common and limit use of first-line antibiotics.1,2 Given that most patients with CALs are not allergic, formal verification of these labels is critical for antimicrobial stewardship.1-3 Many studies have evaluated risk stratification to facilitate penicillin allergy evaluation and use of direct challenges, but few studies have focused on CALs.4,5 Studies have validated CEPH-FAST, Urticaria 1-1-1, and Vanderbilt criteria for CAL risk stratification.6-8 However, to our knowledge, none of these tools have undergone subsequent external validation.6-8 We aimed to examine the use and external validity of CEPH-FAST, Urticaria 1-1-1, and Vanderbilt criteria for cephalosporin allergy risk stratification.
Methods
We performed a retrospective review of the medical records of patients with CALs evaluated at a Yale University–affiliated allergy and immunology clinic between December 1, 2014, to July 21, 2025. The Yale University Institutional Review Board deemed this cross-sectional study exempt from ethics review and waived the informed consent requirement because of secondary use of data. We followed the STROBE reporting guideline.
Under consecutive sampling, patients were eligible if they were 18 years or older and had undergone complete cephalosporin allergy evaluation. We defined complete evaluation as a consultative visit, followed by skin testing with drug challenge (if testing had a negative result) or direct cephalosporin challenge. Consistent with a guideline-recommended testing approach1 and the original CEPH-FAST study,6 we included only patients who had a cephalosporin challenge with the implicated cephalosporin or a cross-reactive cephalosporin. Excluded patients had a negative skin testing result but did not undergo a cross-reactive cephalosporin challenge to ascertain tolerance. A positive skin test result or an allergist-adjudicated immune-mediated allergic reaction was considered a verified cephalosporin allergy. We recorded whether each tool could be scored using allergists’ documentation.
Additionally, we calculated tool performance characteristics (eg, sensitivity, specificity, negative predictive value [NPV], positive predictive value [PPV], area under the receiver operating characteristics curve [AUROC]). Because intravenous (IV) cephalosporins are associated with higher risk for anaphylaxis, we performed sensitivity analyses excluding IV CALs.6,7,9
The eMethods, eTable, and eFigure in Supplement 1 provide more information. Data analysis was performed with Stata 18.0 (StataCorp LLC).
Results
Among 121 eligible patients (median [IQR] age, 49.0 [33.0-67.0] years; 104 females [86.0%]), 23 (19.0%) had verified cephalosporin allergy (20 with a positive skin testing result, 1 with a positive immediate challenge, and 2 with benign delayed morbilliform rash) (Table 1). CEPH-FAST and Vanderbilt criteria could be scored using allergists’ documentation in 120 (99.2%) and 106 (87.6%) patients, respectively, compared with only 9 patients (7.4%) for Urticaria 1-1-1.
Table 1. Demographic Characteristics, Medical History, and Testing Results of Patients View LargeDownload (opens in new tab)Go to Figure in ArticleCharacteristicPatients, No. (%) (N = 121)Age, median (IQR), y49.0 (33.0-67.0)Sex Female 104 (86.0) Male 17 (14.0)Race and ethnicitya Asian 3 (2.5) Black or African American 4 (3.3) Hispanic or Latinx4 (3.3) White 107 (88.4) Other or unknownb7 (5.8)Primary language not English1 (0.8)Medical and allergy historyc Charlson comorbidity index, median (IQR)2.0 (0-4) Pregnant at time of initial allergist visit24 (19.8) Pregnant at time of cephalosporin challenge6 (5.0) Asthma54 (44.6) Atopic dermatitis21 (17.3) Allergic rhinoconjunctivitis65 (53.7) Chronic urticaria15 (12.3) Food allergy30 (24.8) Multiple (≥2) drug allergy labels 100 (82.6)Implicated cephalosporin Cefaclor17 (14.0) Cefazolin19 (15.7) Cefprozil10 (8.3) Ceftriaxone4 (3.3) Cefuroxime14 (11.5) Cephalexin45 (37.2) Unspecified17 (14.0) Multiple cephalosporin allergy labels14 (11.5)Cephalosporin index reaction historyc Anaphylaxis22 (18.2) Perioperative anaphylaxis specifically16 (13.2) Angioedema7 (5.8) Nonurticarial rash43 (35.5) Urticarial rash38 (31.4) Remote: >5 y prior84 (69.4) Severe cutaneous adverse reaction2 (1.6) CEPH-FAST, Vanderbilt criteria, Urticaria 1-1-1 score documented in allergists’ clinical note0Risk stratification tool: documentation sufficient to scored CEPH-FAST120 (99.2) Urticaria 1-1-19 (7.4) Vanderbilt criteria106 (87.6)Testing characteristics and results Testing data completion date, median (IQR)January 18, 2023 (May 19, 2021, to March 4, 2024) Positive skin prick or intradermal test20 (16.5) Positive immediate challenge following negative skin test1 (0.8) Positive direct cephalosporin challenge0 Benign delayed morbilliform rash following challenge2 (1.6)
CEPH-FAST and Vanderbilt criteria showed specificities of 84.5% (95% CI, 75.8%-91.1%) and 88.2% (95% CI, 79.4%-94.2%) and NPVs of 94.3% (95% CI, 87.1%-98.1%) and 96.2% (95% CI, 89.2%-99.2%), respectively (Table 2). Both tools showed good discriminatory ability (AUROC, CEPH-FAST: 0.81 [95% CI, 0.72-0.91]; Vanderbilt criteria: 0.87 [95% CI, 0.79-0.95]). When excluding IV CALs, the NPV of CEPH-FAST increased slightly, but sensitivity decreased for both CEPH-FAST (78.3% [95% CI, 56.3%-92.5%] to 33.3% [95% CI, 4.3%-77.7%]) and Vanderbilt criteria (85.7% [95% CI, 63.7%-97.0%] to 25.0% [95% CI, 0.6%-80.6%]).
Table 2. External Validation of CEPH-FAST and Vanderbilt Criteria for Identifying Positive Test Results in Patients Undergoing Skin Testing and/or Cross-Reactive Cephalosporin Challenge View LargeDownload (opens in new tab)Go to Figure in ArticleCEPH-FAST (n = 120)CEPH-FAST excluding IV cephalosporin allergy labels (n = 94)Vanderbilt criteria (n = 106)Vanderbilt criteria excluding IV cephalosporin allergy labels (n = 81)Positive test result prevalence in cohort, No. (%)23 (19.2)6 (6.4)21 (19.8)4 (4.9)Patients with low risk, No. (%)87 (72.5)83 (88.3)78 (73.6)74 (91.3) Positive test result, No. (%)5 (5.7)4 (4.8)3 (3.8)3 (4.1)Patients with high risk, No. (%)33 (27.5)11 (11.7)28 (26.4)7 (8.6) Positive test result, No. (%)18 (54.5)2 (18.2)18 (64.3)1 (14.3)Sensitivity (95% CI), %78.3 (56.3-92.5)33.3 (4.3-77.7)85.7 (63.7-97.0)25.0 (0.6-80.6)Specificity (95% CI), %84.5 (75.8-91.1)89.8 (81.5-95.2)88.2 (79.4-94.2)92.2 (83.8-97.1)PPV (95% CI), %54.5 (36.4-71.9)18.2 (2.3-51.8)64.3 (44.1-81.4)14.3 (0.4-57.9)NPV (95% CI), %94.3 (87.1-98.1)95.2 (88.1-98.7)96.2 (89.2-99.2)95.9 (88.6-99.2)Positive LR (95% CI)5.06 (3.03-8.45)3.26 (0.90-11.84)7.29 (3.97- 13.38)3.21 (0.50- 20.68)Negative LR (95% CI)0.26 (0.12- 0.56)0.74 (0.42-1.31)0.16 (0.06- 0.46)0.81 (0.46-1.44)OR (95% CI)19.68 (6.50-59.13)4.39 (0.83-24.19)45.00 (11.75-168.82)3.94 (0-33.84)AUROC (95% CI)0.81 (0.72-0.91)0.62 (0.41-0.82)0.87 (0.79-0.95)0.59 (0.34-0.83)
Discussion
This study of patients evaluated for CALs showed that, among 3 risk stratification tools, CEPH-FAST and Vanderbilt criteria had similar performance characteristics compared with the original studies,6-8 with acceptable sensitivity and specificity and robust NPVs. However, we observed a substantial decrease in sensitivity when excluding patients with IV CALs. Compared with oral medications, IV medications are associated with higher risk for anaphylaxis, likely due to differences in bioavailability, allergic reaction physiological function, and sensitization.1,6,7,9 In contrast, in both children and adults, verified oral cephalosporin allergy is rare (<2.5% oral CALs).7,10 Our findings underline the importance of using these tools alongside a comprehensive allergy history, considering IV CALs’ higher risk (eg, when using CEPH-FAST), and recognizing that these tools are best at identifying patients with low-risk histories for direct challenge (but not cases of true allergy).
Study limitations include the retrospective design; single site setting; and a predominantly female, White sample with atopic comorbidities. Although skin testing is not a guideline-concordant approach to ruling out drug allergy (without a challenge),1 not including patients with negative results also introduces potential for bias. Because this study was conducted at a tertiary referral center and excluded patients directly delabeled, it included a sample enriched for true allergy: in a (general) population with lower prevalence, the tools’ NPV may be higher and PPV may be lower. As the verified allergy cases included patients with positive skin testing results (and the diagnostic validity of cephalosporin skin testing has not been well established), false-positive results were possible. Lastly, medication sensitization patterns vary geographically,9 which may affect generalizability.
This research builds on prior studies and lends valuable evidence supporting CEPH-FAST and Vanderbilt criteria in evaluating CALs, along with a comprehensive allergy history that considers administration route of the implicated cephalosporin. Future research should validate these tools in other populations and for IV cephalosporins specifically. Multisite trials and/or meta-analyses may also aid in modeling predictive values over a wide range of population characteristics and in creating tools maximizing sensitivity to identify true allergy cases instead of maximizing NPV to identify candidates for direct challenge.
Back to top
Article Information
Accepted for Publication: July 21, 2026.Published: September 16, 2026. doi:10.1001/jamanetworkopen.2026.34079Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2026 Belmont AP et al. JAMA Network Open.Corresponding Authors: Ami P. Belmont, MD, MHS, Department of Internal Medicine, Yale University School of Medicine, 333 Cedar Street, PO Box 208013, New Haven, CT 06520 (ami.belmont@yale.edu); Cosby A. Stone Jr, MD, MPH, Division of Allergy, Pulmonary, and Critical Care Medicine, Vanderbilt University Medical Center, 1161 21st Ave S, T-1218, MCN, Nashville, TN 37232-2650 (cosby.a.stone@vumc.org). Author Contributions: Dr Belmont had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.Concept and design: Belmont, Valente, Stone.Acquisition, analysis, or interpretation of data: All authors.Drafting of the manuscript: Belmont, Valente, Workineh, Stone.Critical review of the manuscript for important intellectual content: Belmont, Novotny, Kwah, Su, Wong, Phillips, Stone.Statistical analysis: Belmont, Workineh, Stone.Obtained funding: Belmont, Stone.Administrative, technical, or material support: Belmont, Valente, Novotny, Kwah, Su, Phillips, Stone.Supervision: Belmont, Phillips, Stone.Conflict of Interest Disclosures: Dr Belmont reported receiving a Yale Physician Scientist Development Award and Clinical and Translational Science Award (CTSA) from the National Center for Advancing Translational Science (NCATS) during the conduct of the study. Dr Su reported receiving grants from the American College of Allergy Asthma and Immunology Allergists’ Foundation Community Grant Program outside the submitted work. Dr Phillips reported receiving a grant from the Agency for Healthcare Research and Quality (AHRQ); grants to institution from the National Institutes of Health (NIH) and the National Health and Medical Research Council outside the submitted work; and personal fees from UpToDate, Elion, Verve, Mirador, Rapt, Johnson & Johnson, AstraZeneca, and Intellia outside the submitted work. Dr Stone reported receiving grants from AHRQ and the National Institute of Allergy and Infectious Diseases outside the submitted work and personal fees from Stallergenes-Greer, Larimar Therapeutics, Bracco Diagnostics, Astra-Zeneca, AllergyWatch, Novartis, and FARE outside the submitted work. No other disclosures were reported.Funding/Support: This publication was funded by a Yale Physician Scientist Development Award and CTSA grant UL1 TR001863 from NCATS (Dr Belmont).Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.Disclaimer: The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.Data Sharing Statement: See Supplement 2.Additional Information: Study data were collected and managed using REDCap (Research Electronic Data Capture) tools hosted at Yale University. REDCap is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for data integration and interoperability with external sources.
References
1.Khan
DA, Banerji
A, Blumenthal
KG,
et al; Chief Editor(s); Workgroup Contributors; Joint Task Force on Practice Parameters Reviewers. Drug allergy: a 2022 practice parameter update. J Allergy Clin Immunol. 2022;150(6):1333-1393. doi:10.1016/j.jaci.2022.08.028PubMedGoogle ScholarCrossref2.MacFadden
DR, LaDelfa
A, Leen
J,
et al. Impact of reported beta-lactam allergy on inpatient outcomes: a multicenter prospective cohort study. Clin Infect Dis. 2016;63(7):904-910. doi:10.1093/cid/ciw462PubMedGoogle ScholarCrossref3.Moraczewski
J, Lee
R, Koo
G,
et al. Direct oral challenges safely reduce the burden of low-risk cephalosporin allergy labels. J Allergy Clin Immunol Pract. 2025;13(9):2522-2525.e1. doi:10.1016/j.jaip.2025.05.054PubMedGoogle ScholarCrossref4.Trubiano
JA, Vogrin
S, Chua
KYL,
et al. Development and validation of a penicillin allergy clinical decision rule. JAMA Intern Med. 2020;180(5):745-752. doi:10.1001/jamainternmed.2020.0403ArticlePubMedGoogle ScholarCrossref5.Novotny
S, Elmansy
L, Kwah
J,
et al. High-risk penicillin reaction flags in the medical record. JAMA Netw Open. 2025;8(12):e2549081. doi:10.1001/jamanetworkopen.2025.49081ArticlePubMedGoogle ScholarCrossref6.Cox
F, Vogrin
S, Sullivan
RP,
et al. Development and validation of a cephalosporin allergy clinical decision rule. J Infect. 2025;90(6):106495. doi:10.1016/j.jinf.2025.106495PubMedGoogle ScholarCrossref7.Koo
G, Yu
R, Phillips
EJ, Stone
CA
Jr. Retrospective stratification of cephalosporin allergy label risk using validated penicillin allergy frameworks. J Allergy Clin Immunol Pract. 2022;10(9):2472-2475.e1. doi:10.1016/j.jaip.2022.05.032Google ScholarCrossref8.Sabato
V, Gaeta
F, Valluzzi
RL, Van Gasse
A, Ebo
DG, Romano
A. Urticaria: the 1-1-1 criterion for optimized risk stratification in β-lactam allergy delabeling. J Allergy Clin Immunol Pract. 2021;9(10):3697-3704. doi:10.1016/j.jaip.2021.05.037PubMedGoogle ScholarCrossref9.Shah
D, Rukasin
C, Wang
C, Phillips
E, Stone
C. Unraveling cephalosporin-associated anaphylaxis: a comprehensive analysis using FDA Adverse Event Reporting System data. J Allergy Clin Immunol Pract. 2024;12(5):1359-1361. doi:10.1016/j.jaip.2024.03.013PubMedGoogle ScholarCrossref10.Rukasin
CRF, Chow
TG, Van Stechelman
P,
et al. Direct challenges are safe and effective in children with low-risk allergies to oral cephalosporins. J Allergy Clin Immunol Pract. 2026;14(1):286-288.e1. doi:10.1016/j.jaip.2025.10.008PubMedGoogle ScholarCrossref