Text

Background: The evidence on the primary prevention of allergic diseases is distributed across numerous systematic reviews and is difficult to navigate because of heterogeneous populations, interventions, outcomes, and evolving research approaches. Structured and continuously updated evidence resources are needed to support research, guideline development, healthcare, and public health.

Objective: To describe the conceptual and methodological development of the Allergy Prevention Living Overview of Evidence (L·OVE), a freely accessible, continuously updatable evidence repository, and to present selected initial findings.

Methods: A systematic search for systematic reviews was conducted in the Epistemonikos database. Reviews synthesizing evidence on primary preventive interventions aiming to prevent allergic diseases or allergic sensitization were eligible. A hierarchical taxonomy covering population, intervention, contextual dimensions, and outcomes was developed iteratively. Relevant records were identified using an automated, platform-supported classification and identification workflow validated through a relative recall approach and compared with a Boolean search strategy. At least two independent reviewers screened records and classified included reviews taxonomically; disagreements were resolved by consensus or consultation with a third reviewer.

Results: As of 18 November 2025, 8,248 records had been screened by title and abstract, 289 full texts assessed, and 245 systematic reviews included. Nutrition-related interventions dominated the evidence base (n=218), followed by pharmacological (n=29), skin-related (n=11), and nature-based interventions (n=5); multiple classifications were possible. The automated workflow achieved a sensitivity of 96.49%, compared with 92.98% for the Boolean strategy, and reduced the manual screening workload by 25.5% (8,248 vs 11,072 records).

Conclusion: The Allergy Prevention L·OVE provides a structured and continuously updatable resource for locating, classifying, and contextualizing systematic reviews. Combining automated evidence identification, a structured taxonomy, and expert curation can support more transparent use of the evidence base and facilitate the identification of research gaps. The repository supports evidence retrieval and mapping rather than evidence synthesis itself.

Competing interests: None