A Delphi‐Based Item Set for Pre‐Test Asthma Probability Assessment in Children Undergoing Methacholine Challenge Test
Tahani Ibrahim, Alessia Mariani, Maxime Beydon, Juliette Dardenne, Marc Koskas, Alexandra Costea, Lucia Maingot, Jessica Taytard, Nicole BeydonABSTRACT
Objective
To select, through a Delphi consensus process, the clinical items most relevant to pre‐test probability of asthma (PTPA) in children undergoing methacholine challenge test (MCT) and to explore their association with expert‐estimated PTPA.
Working Hypothesis
Delphi‐selected items are associated with PTPA estimated independently by pediatric pulmonologists.
Study Design
Three‐round e‐Delphi, followed by a retrospective analysis of item association with expert‐estimated PTPA. Patient‐subject selection: A database of 92 children referred for MCT at a tertiary center.
Methodology
A steering committee proposed 11 candidate items, rated by 23 experts (consensus threshold ≥ 80%). Five pediatric pulmonologists, blinded to MCT results and not involved in the Delphi, independently estimated PTPA from clinical history. Agreement was assessed with intraclass correlation coefficients (ICC). Associations between retained items and mean PTPA were assessed using a multivariable linear regression.
Results
Ten items were retained (nine reached consensus; pneumonia/atelectasis was retained by steering‐committee judgment). Single‐rater ICC was 0.78 (95% CI 0.70–0.85); average‐measures ICC was 0.95 (95% CI 0.92–0.97). Multivariable linear regression retained all ten items; each positively associated with PTPA. The strongest associations were seen with wheezing after age 3 years (β 15, 95% CI 7.4–23), wheezing within the last 3 years (β 14, 95% CI 6.4–22) and complete response to anti‐asthma treatment with relapse after treatment cessation (β 34, 95% CI 27–41).
Conclusion
Delphi consensus identified ten clinically relevant items, all associated with expert‐estimated PTPA. As the outcome was expert judgement rather than confirmed asthma, diagnostic value remains unestablished; outcome‐based studies are needed.