Diagnostic performance and clinical utility of open oral food challenges using double-blind, placebo-controlled food challenges as the reference standard: A systematic review
In brief
Negative open food challenges matched blinded tests in up to 97% of cases
In this review of eight studies, open food challenges had a negative predictive value of 88% to 97% against blinded challenges, suggesting a negative result may help rule out food allergy. But study quality was often poor, results varied, and positive predictive value ranged from 0% to 89%; open challenges cannot yet replace blinded testing.
- Journal
- Allergology international : official journal of the Japanese Society of Allergology (Q1)
- Published
- 27 October 2026
- Study design
- Systematic review / meta-analysis of RCTs
- Evidence level
- Level 1, High (CEBM 1a)
- Authors
- Yu Kuwabara, Norio Kawamoto, Takeshi Yamamoto, Tomoyuki Arima, Kazutaka Nogami, Shuichi Suzuki, et al.
- PMID
- 42946619
- DOI
- 10.1016/j.alit.2026.09.006
Why clinicians should know about it
- Picked for Immunology and Allergy (paper of the day, 29 October 2026): Diagnostic accuracy review of oral food challenges
Abstract
BACKGROUND: Double-blind, placebo-controlled food challenges (DBPCFCs) are regarded as the reference standard for diagnosing IgE-mediated food allergy owing to their objectivity and ability to minimize bias. However, DBPCFCs are logistically complex, time-consuming, and not always feasible in routine clinical settings. In contrast, open oral food challenges (OFCs) are more accessible and widely used but potentially less reliable due to a lack of blinding. We evaluated the diagnostic performance of open OFCs and descriptively compared their clinical utility. METHODS: We conducted a systematic review of literature available in the Cochrane Central Register of Controlled Trials, PubMed, and Ichushi databases between January 1, 2000, and December 31, 2024, incorporating randomized controlled trials and observational studies in which both open OFCs and DBPCFCs were performed. Risk of bias (RoB) was assessed using QUADAS-2. RESULTS: Eight studies were included. Using DBPCFC as the reference standard, the diagnostic performance of open OFC was as follows: sensitivity, 0.91-0.92 (n = 2); specificity, 0.83-1.00 (n = 3); positive predictive value, 0.00-0.89 (n = 6); and negative predictive value, 0.88-0.97 (n = 2). RoB was high mainly owing to the patient selection and the flow and timing domains. CONCLUSIONS: Given the small number of studies examined, high RoB, and heterogeneity among studies, caution is needed in generalizing these results. Although a negative open OFC result may be reliable, the current evidence is insufficient to establish open OFCs as a complete substitute for DBPCFCs, and large-scale, standardized studies in more diverse populations are needed to clarify their diagnostic performance and clinical utility.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.