Publications

Retraction changes the status of a source, but claims built on it change only if someone reassesses them. We define correction reach as the proportion of identifiable downstream assertions that depend on a corrected source and were reassessed after the correction, with the decision documented. A documented no-change decision counts as reached. We specify a Correction Reach Protocol (trigger, trace, notify, reassess, decide, record) and apply it in a feasibility pilot to five retraction events. Of 85 candidate citing works, 43 had assessable citation contexts and yielded a fixed list of 89 assertions. Two independent raters coded this list plus five control assertions. They agreed on dependency for 93.6% of 94 assertions (Gwet’s AC1 0.93; Cohen’s κ 0.39, low because nearly all codes were dependent). None of the 81 dependent assertions in 38 papers had documented reassessment after 8.9 to 10.6 years (one-sided 95% upper bound 3.6%). In a known-positive control, one correction reached all five dependent assertions of a meta-analysis 132 days after the retraction. Correction reach can be coded on real cases; independent segmentation and citation-free lineage remain untested. It can show whether large language models widen or narrow the gap between producing and correcting claims.

The digital transformation of medical research has outpaced the ethical frameworks designed to govern it, creating systemic vulnerabilities that legacy governance models cannot adequately address. Drawing on the Peruvian experience, we examine the convergence of metric-driven research evaluation, regulatory gaps, and digital disruption, and we argue for hybrid governance as an operational necessity for preserving scientific integrity. Our analysis reveals three interconnected findings. First, regulatory codification alone is insufficient to prevent misconduct. Peru’s National Code and RENACyT system proved vulnerable to large-scale authorship trafficking, exemplified by the “La Granja de Científicos Bamba” scandal involving over USD 53 million in state incentives. Second, institutional risk, assessed via SCImago IRIS, is not linked to the performance of university publications. Most of Peru’s top 20 universities fall into medium or significant risk categories, though low-risk cases indicate that maintaining integrity is possible at the institutional level. Third, the digital era compounds these vulnerabilities through challenges that legacy frameworks cannot address. We propose a reorientation of medical research ethics along five axes: from privacy to data justice, from consent to continuous engagement, from individual to systemic accountability, from voluntary RRI to mandatory Ethics by Design, and from national codes to global hybrid governance. The Peruvian case offers distinctive lessons for the international community: rapidly expanding research output, a regulatory framework that appears robust yet is operationally weak, and organized fraud at scale. This combination exposes the universal vulnerability of metric-driven systems to gaming when oversight lags behind digital realities.

Scientific disinformation has become a major challenge for contemporary societies, affecting public trust in science, evidence-based decision-making, and the communication of scientific knowledge. (1) Methods: This study maps the communication-oriented academic research on science communication and disinformation, identifying its main publication trends, thematic areas, and methodological approaches. A Systematic Mapping Review was conducted using the Web of Science Core Collection, specifically restricted to Communication subject categories to bound the disciplinary scope. The search identified 127 research articles published between 2003 and 2026, which were analyzed through bibliometric, thematic, and methodological coding. (2) Results: Results show a marked expansion of research since 2020, with 91.34% of the analyzed articles published between 2020 and 2026. The United States constitutes the main production node, while research is distributed across 52 journals. Thematic analysis shows a predominance of the media ecosystem and communication actors, followed by reception, trust, and cognitive processing. Public health, social platforms, and trust in scientific authority are among the most recurrent specific topics. Qualitative approaches predominate (51.2%), while content analysis and surveys are the most frequent methodological strategies. Despite growing interest in prebunking, debunking, and media literacy, computational approaches remain marginal. (3) Conclusions: These findings reveal a rapidly consolidating but methodologically uneven field and identify priorities for future research, particularly computational methods, interdisciplinary collaboration, and intervention-oriented studies.