Letter: Meta-Analysis Critique: Refining the Association Between Helicobacter pylori and Eosinophilic Oesophagitis. Authors' Reply

Abstract

We appreciate the insightful comments from Dr. Rahman [1] regarding our meta-analysis [2] and welcome this opportunity for further discussion.

Helicobacter pylori (H. pylori) prevalence is influenced by socioeconomic and environmental factors such as diet, healthcare access and living conditions, all of which could independently affect the risk of developing eosinophilic esophagitis (EoE) [3, 4]. Failing to adjust for these confounding factors may result in an overestimation or misinterpretation of the association between H. pylori and EoE. A multivariate adjustment for these potential confounders could provide a clearer understanding of how H. pylori may contribute to the risk of EoE, offering more accurate insights into this potential connection.

We are aware that the setting in which participants are recruited could also play a significant role in introducing bias [5]. A sensitivity analysis, comparing symptomatic hospital patients with individuals with milder disease, would be a valuable approach to better understand how this influences results, providing a more accurate assessment of the relationship between H. pylori and the outcome of interest. We were not able to make this analysis due to the unavailability of pertinent data in included studies.

Longitudinal data are essential to clarify whether H. pylori infection contributes to the development of EoE or whether EoE treatments affect the ability to detect H. pylori. Long-term studies would help to establish whether H. pylori infection precedes EoE onset or whether the relationship is more complex, influenced by the treatments for EoE [6]. This would refine our understanding of the role H. pylori plays in the development of EoE.

Finally, relying solely on visual assessments of funnel plots may not always provide a clear picture of bias, and quantitative methods such as Egger's test could provide more rigorous insights [7]. Incorporating unpublished data is helpful to reduce this bias and improve the overall robustness of the study's findings. Future work should certainly focus on addressing these gaps.

We thank Dr. Rahman [1] for the opportunity to clarify these points.

Irene Spinelli: conceptualization, investigation, writing – original draft, methodology, writing – review and editing, formal analysis, project administration, data curation, resources. Gianluca Ianiro: validation, visualization, supervision, conceptualization, writing – review and editing, formal analysis, resources, data curation.

This article is linked to Spinelli et al papers. To view these articles, visit https://doi.org/10.1111/apt.70042 and https://doi.org/10.1111/apt.70122.