Selective Harm Reporting in Inflammatory Sinonasal Trials: A Systematic Review of Registry-Publication Concordance.

Abstract

To assess the completeness, consistency, and transparency of adverse event (AE) reporting in clinical trials of inflammatory sinonasal diseases by comparing data from ClinicalTrials.gov and corresponding peer-reviewed publications. ClinicalTrials.gov registry and MEDLINE-indexed journal articles reporting results of interventional trials focused on inflammatory sinonasal disease. We identified trials with posted results between 2014 and 2024 and matched registry records with corresponding publications. Data extraction included structured and narrative AE fields. Outcomes assessed were serious adverse events (SAEs), other adverse events (OAEs), mortality, and AE-related discontinuations. Reporting concordance was analyzed using descriptive statistics, Bland-Altman plots, funnel plots, and linear regression to evaluate temporal trends and reporting predictors. Only direct numerical matches were accepted; we made no inferences from narrative text. Among 108 included trials, 57 (52.8%) met criteria for likely applicable clinical trials (ACTs). AE reporting was more complete on ClinicalTrials.gov than in publications. For example, 94.7% of ACTs reported SAEs in the registry compared to 80.7% in corresponding publications, and death reporting increased from 38% to 100% in registry data following the 2017 Final Rule. However, publication reporting did not show corresponding improvement. Funnel plots revealed dispersion in AE rates among smaller trials, while linear regression showed modest gains in registry reporting over time. Despite regulatory improvements, publication-based AE reporting remains incomplete and inconsistent. Clinical trial registries remain an essential, yet underutilized, resource for harm-related evidence in sinonasal disease research.