Systematic Review Finds Widespread Safety Reporting Gaps in Glaucoma Trials

Transparent reporting of adverse events remains essential as clinicians evaluate emerging glaucoma therapies. This graph from the study shows the discordance between published adverse events and those found in clinical trial registries. The ARD represents the absolute difference in reporting proportions (publication minus registry). Photo: Elghzali A, et al. Sci Rep. 2025;15(1):40563. Click image to enlarge. A new analysis is raising red flags about how consistently, and transparently, adverse events are reported in glaucoma clinical trials. For practicing optometrists who routinely counsel glaucoma patients on treatment risks and benefits, the findings highlight a critical need for greater vigilance when interpreting published data.In a systematic review and cross-sectional comparison published in Scientific Reports, investigators examined randomized controlled trials for glaucoma registered on ClinicalTrials.gov between 2009 and 2024. Each eligible trial entry was paired with its corresponding peer-reviewed publication, identified through PubMed and Google Scholar searches. The team then compared reporting of serious adverse events (SAEs), other adverse events (OAEs), mortality and participant withdrawals between registry records and published articles.Among the 57 eligible glaucoma trials, the authors found substantial and quantifiable gaps in safety reporting between ClinicalTrials.gov records and corresponding publications. Nearly one-third of trials (31.6%) showed discrepancies in SAE reporting, with mismatches affecting both participant counts (18 trials, 31.6%) and the number of serious adverse events themselves (27 trials, 47.4%). Discrepancies were even more pronounced for OAEs; more than three-quarters of trials (77.2%) showed mismatches in participant counts, and 89.5% (51 trials) differed in event counts. In many instances, key safety data were omitted entirely from publications, the researchers reported. Mortality reporting was similarly inconsistent, with 61.4% of trials listing mortality outcomes on ClinicalTrials.gov, compared with only 42.1% in published manuscripts. Almost half (47.4%) of trials exhibited discrepancies in mortality data. Withdrawals due to adverse events also differed in 19 trials (33.3%).“Across glaucoma randomized trials, we found frequent differences between journal articles and ClinicalTrials.gov in how they reported serious and OAEs, deaths and withdrawals,” the study authors wrote in Scientific Reports. “These appear to be systematic problems due to uneven definitions, timing gaps between registry updates and publication, and too few routine cross-checks.” The researchers admonished authors who, whether intentionally or unintentionally, emphasize their study’s primary outcomes at the expense of full disclosure of all findings. “While it may be exciting to publish positive results of a project the authors worked hard on, it is still imperative that authors publish all aspects of the data regarding their projects,” they wrote. “Not being transparent can lead to bias that not only distorts the safety profile of treatments but also undermines the integrity of evidence synthesis, potentially affecting clinical decision-making.” A practical path forward, they propose, is to standardize harms reporting (e.g., following the principles laid out in the CONSORT Harms 2022 guidance) with structured safety tables, align registry posting with publication workflows and adopt clear, auditable processes for resolving discrepancies and issuing updates. “These steps would reduce omissions, improve comparability and provide clinicians and patients with more reliable safety information.”Click here for the journal source. Elghzali A, Chaudhry M, Khan S, et al. Systematic review with cross sectional comparative analysis identifies adverse event reporting discrepancies between ClinicalTrials.gov and published glaucoma randomized controlled trials. Sci Rep. 2025;15(1):40563. This article was developed by the editorial staff in conjunction with experts in the field. In the process, AI may have been among the editorial tools used to meet the goals of human editors, who approved all content.