The problems identified by the editors centre on the article’s reference list, the apparatus that connects any piece of scholarship to the wider body of knowledge it claims to build upon. According to the retraction note, multiple references were found to be incorrect or unverifiable, specifically those numbered 13, 7, 6, 4, 15 and 53. In plain terms, these citations either pointed to sources that did not match what the text claimed, or could not be traced to any verifiable publication at all. For a paper whose entire subject was the careful appraisal and utilisation of evidence in clinical settings, the irony is difficult to miss.

More serious still, the editors reported that several cited works could not be confirmed as existing publications at all. References 5, 12, 16, 47 and 48 fell into this category, raising the possibility that they were fabricated, a form of citation fraud that has become one of the most corrosive problems in modern scholarly publishing. In addition, the notice flags duplication: references 9 and 11 appear to be duplicates of references 49 and 53 respectively, an error pattern that suggests the bibliography was assembled without adequate verification or cross-checking.

Retractions of this kind follow a well-defined procedural pathway. When concerns about a published article surface, whether through reader complaints, post-publication scrutiny, or internal editorial review, the journal typically investigates with input from the authors and their institutions. If the concerns prove substantive and cannot be resolved, the editor may issue a retraction notice, which is permanently linked to the original article and indexed in bibliographic databases. The original paper remains online, but it is watermarked and cross-linked to the notice so that readers encountering it later understand its status. In this case, the retraction note is itself the version of record now attached to the study’s digital object identifier.

The author responses, or lack of them, form a notable part of the record. According to the notice, Ukpai Eze disagrees with the retraction, while David Agom did not explicitly state whether or not they agree with it. None of the remaining authors responded to correspondence from the publisher about the retraction. This pattern, partial dissent combined with widespread silence, is not unusual in retraction cases, but it complicates the narrative. Disagreement from one author does not overturn an editorial decision, yet it signals that the underlying facts may be contested within the authorship itself, and it leaves readers with an incomplete account of how the problematic references came to be in the manuscript.

The original study sat at the intersection of two consequential fields: nursing management and knowledge translation. Evidence-based practice, the discipline of grounding clinical decisions in the best available research rather than in tradition or habit, depends heavily on leaders who can find, appraise and deploy research findings within busy hospital environments. Nurse managers in acute care settings are often the practical gatekeepers of this process. They decide which guidelines gain traction on the ward, which training programmes receive time and resources, and whether research findings actually change bedside practice. A study describing their strategies for knowledge utilisation would, if reliable, be of genuine interest to hospital administrators, nursing educators and health policy researchers alike.

That is precisely why reference integrity matters so much in a paper of this kind. A qualitative or mixed-methods study of management practice does not stand alone; it interprets its findings through the lens of prior literature on organisational behaviour, knowledge management and clinical leadership. If the citations anchoring that literature are wrong, invented, or duplicated, the interpretive framework collapses. Readers cannot check the intellectual lineage of the claims. Systematic reviewers, who rely on accurate reference lists to map research landscapes, may be misled into counting phantom publications as real evidence. In fields that directly inform patient care, such distortions carry a cost beyond academia.

The case also illustrates the technical standards that journals now apply when auditing bibliographies. Verification typically involves checking each citation against publisher databases, digital object identifier registries and indexing services such as PubMed and Crossref. A citation that resolves to no matching title, author combination or journal is flagged as unverifiable. Duplicated references, where the same work appears under two different entries, are detected through title and author matching. Incorrect references, where the cited source exists but does not support the claim attributed to it, require closer reading. The granularity of the BMC Nursing notice, citing specific reference numbers by category, reflects this systematic audit and serves as a public accounting of what was found.

For the wider research community, the episode is a reminder that the credibility of a paper rests on layers of verification that most readers never see. Peer review, however rigorous, does not always include exhaustive reference checking, and the sheer volume of global publishing has strained the capacity of editorial offices to catch every anomaly. Automated screening tools, reference-matching software and post-publication scrutiny have all grown in importance as complements to traditional review. Retraction, once seen as a catastrophic rarity, is now understood as a routine corrective mechanism, a sign that the system’s error-detection is functioning rather than that science itself is failing.

What remains after the retraction is a cautionary document. The original article, published in March 2025, is now permanently linked to a notice that dismantles its evidentiary scaffolding, and the journal’s stated loss of confidence in its data and conclusions stands as the authoritative record. For nursing researchers, the lesson is procedural: every citation must be verified against the actual source, and authorship carries responsibility for the bibliography as much as for the findings. For clinicians and managers who might have drawn on the study to shape evidence-based practice initiatives, the retraction is a signal to look elsewhere for guidance. And for readers of science journalism, it is a demonstration that the machinery of correction, slow and unglamorous as it is, does eventually turn, even when the subject of the failed paper is the proper handling of evidence itself.

Subject of Research: Retraction of a nursing management study on evidence-based practice due to unverifiable references

Article Title: Retraction Note: Leading evidence-based practice: nurse managers’ strategies for knowledge utilisation in acute care settings

Article References: Ominyi, J., Nwedu, A., Agom, D., & Eze, U. (2026). Retraction Note: Leading evidence-based practice: nurse managers’ strategies for knowledge utilisation in acute care settings. BMC Nursing, 25(1), Article 883. https://doi.org/10.1186/s12912-026-05399-w

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05399-w

Keywords: retraction, BMC Nursing, evidence-based practice, nurse managers, research integrity, citation fraud, reference verification, acute care, knowledge utilisation, nursing research, publishing ethics, Springer Nature

The retracted article was authored by Muhammad Asif and Zhou Gouqing, both affiliated with the College of Journalism and Communication at Hunan Normal University in Changsha, China. It was first published on 16 May 2024 in the journal’s fourth volume, appearing as article number 37. The paper was positioned as a critical review, a genre that synthesizes existing literature to evaluate the state of a field, in this case the application of artificial intelligence methods across multiple dimensions of communication research. Reviews of this kind carry particular weight because researchers rely on them to orient themselves in fast-moving areas, which makes their accuracy and provenance especially consequential.

According to the retraction notice, the publisher’s concerns centered on the article’s references. A number of citations in the paper appeared to be contextually incorrect, a pattern the publisher stated may suggest that generative AI was used in writing the manuscript. This specific failure mode has become one of the most recognizable fingerprints of machine-generated academic text. Large language models, when asked to produce scholarly prose, frequently generate citations that look plausible, with real author names, real journal titles, and realistic formatting, but that do not actually support, or in some cases do not correspond to, the claims they are attached to. In a critical review, where the entire argument rests on the accurate representation of prior work, corrupted references strike at the foundation of the paper’s contribution.

The publisher’s account of the process is notable for what it says about authorial responsibility. Springer Nature stated that the authors did not provide a satisfactory explanation when confronted with the concerns, and that the publisher therefore no longer has confidence in the reliability of the article’s contents. That sequence, from suspicion to inquiry to unsatisfactory response to retraction, mirrors the standard machinery of research integrity investigations, but the trigger in this case is distinctly modern. Rather than allegations of fabricated data or plagiarized passages, the core issue was transparency: the apparent use of a powerful writing technology without the disclosure that journal policies now typically require.

The authors’ responses to the retraction diverged in a way that integrity researchers will recognize. Muhammad Asif disagrees with the retraction, formally registering his objection in the published notice. Zhou Gouqing has not responded to correspondence from the publisher about the retraction at all. This split outcome, one author contesting the decision while the other remains silent, complicates the narrative and leaves open questions about how the manuscript was produced and who was responsible for its citation record. Retraction notices are deliberately spare documents, recording what the publisher has concluded and how authors have responded, without adjudicating every underlying dispute in public.

The case arrives amid a broader reckoning over generative AI in scholarly communication. Since large language models became widely accessible, journals and publishers have scrambled to establish rules for their use. The consensus position among major publishers, including Springer Nature, is that AI tools may assist authors in limited ways, such as improving language and readability, but that AI cannot be listed as an author, that authors remain fully responsible for all content including the accuracy of references, and that any use of generative AI must be transparently disclosed in the manuscript. The rationale is straightforward: peer review is built on the assumption that named authors stand behind every claim. When a model has silently shaped the text, that chain of accountability is broken, and reviewers who believed they were evaluating human scholarship may instead have been evaluating synthetic prose.

Detection, however, remains an imperfect science. Unlike image duplication or text recycling, which can be flagged by established software, undeclared generative AI leaves no definitive signature. The contextual incorrectness of references is one of the few reliable indicators, because fabricated or mismatched citations can be checked against the literature directly. Publishers also rely on reader reports, editorial vigilance, and post-publication scrutiny. The two-year gap between the original publication in May 2024 and the retraction in October 2026 illustrates how long these investigations can take, as publishers gather evidence, correspond with authors, and weigh the response before acting. Retraction is deliberately treated as a last resort, because it carries serious professional consequences, but publishers have increasingly signaled that confidence in a paper’s reliability, once lost, cannot be restored by partial corrections.

The irony at the heart of the case has not been lost on observers of the research integrity scene. A paper assessing the innovative application of artificial intelligence in communication research was withdrawn because of the suspected undeclared application of artificial intelligence in its own writing. The episode demonstrates that expertise in a subject does not immunize scholars against the temptations and pitfalls of the tools that subject describes. Communication scholars, of all researchers, might be expected to appreciate how generative systems reshape the production of text, and yet the retraction suggests that the pressures driving AI use, including the volume of writing expected of academics and the speed at which AI-related topics are moving, cut across disciplines.

For the research community, the practical lessons are concrete. Authors who use generative AI in preparing manuscripts should check every reference against the actual source, disclose the tool and the nature of its use in accordance with journal policy, and treat the model’s output as a draft to be verified rather than a finished product. Reviewers and editors, meanwhile, are being trained to scrutinize citation lists for the telltale signs of hallucinated scholarship: references that do not exist, papers that say something different from what is claimed, and bibliographies that blend real venues with implausible content. Institutions, too, have a role, since guidance on AI use in academia has often lagged behind the technology’s adoption in daily practice.

The retracted article remains visible online, as retracted papers typically do, but it now carries the retraction notice as a permanent marker, and the citation record has been updated to reflect the journal’s current volume and article numbering. The case, documented under the DOI 10.1007/s44163-026-02440-4 in Discover Artificial Intelligence, joins a growing public ledger of AI-related retractions that is helping the community calibrate its norms in real time. Whether the number of such cases rises or falls in the coming years will depend less on detection technology than on whether the norms of disclosure and verification take hold among authors before the next generation of writing tools arrives.

Subject of Research: Retraction of a journal review article over undeclared generative AI use and incorrect references

Article Title: Retraction Note: Innovative application of artificial intelligence in a multi-dimensional communication research analysis: a critical review

Article References: Asif, M., & Gouqing, Z. (2026). Retraction Note: Innovative application of artificial intelligence in a multi-dimensional communication research analysis: a critical review. Discover Artificial Intelligence, 6(1), Article 1337. https://doi.org/10.1007/s44163-026-02440-4

Image Credits: AI Generated

DOI: 10.1007/s44163-026-02440-4

Keywords: retraction, generative AI, research integrity, Springer Nature, academic publishing, citation fabrication, Discover Artificial Intelligence, peer review, AI disclosure, scholarly communication, large language models, publishing ethics

The retracted review concerned a class of pollutants with genuine toxicological importance. Polycyclic aromatic hydrocarbons, often abbreviated PAHs, are a large family of organic compounds composed of fused aromatic rings. They form whenever organic material is incompletely burned, which means that the dominant human exposure routes include tobacco smoke, charred and grilled food, wood and coal combustion, vehicle exhaust, and a wide range of occupational settings such as asphalt paving, coking, and aluminum smelting. Because many PAHs are lipophilic and are metabolized in the liver, the organ is both the principal site of their biotransformation and a plausible target for their toxic effects. Enzymes of the cytochrome P450 family convert parent PAHs into reactive metabolites, some of which form DNA adducts and protein adducts, generate oxidative stress, and can disturb hepatocellular function. Serum markers such as alanine aminotransferase, aspartate aminotransferase, and alkaline phosphatase are therefore commonly used in environmental epidemiology as indirect indicators of liver injury in populations exposed to air pollution.

Against that background, a systematic review of observational studies examining PAH exposure and liver function would have been a useful contribution, since observational evidence in this field is scattered across occupational cohorts, general-population surveys, and studies using biomarkers of internal dose such as urinary hydroxylated PAH metabolites. The retraction does not establish that the underlying hypothesis is wrong; rather, it removes this particular synthesis from the citable literature because the journal can no longer vouch for its integrity. That distinction matters for readers who encounter the paper through search engines or reference managers: the retraction notice is now attached to the article record, and the canonical digital object identifier for the notice itself is 10.1007/s11869-026-02101-6, while the original article was published under a different identifier in 2023.

The first ground for retraction, substantial overlap with a previously published article, points to one of the most closely watched problems in contemporary scholarly publishing. Textual and data overlap between papers can range from redundant publication, in which the same work is submitted to multiple journals, to outright plagiarism, in which text or findings are reused without attribution. In this case, the journal identified the overlapping publication as a paper that itself has since been retracted: a systematic review on the association between prenatal exposure to polycyclic aromatic hydrocarbons and childhood intelligence, published in Environmental Science and Pollution Research in 2023, which the notice explicitly labels as a retracted article. Two authors were common to both papers. The fact that the companion article was also retracted suggests a broader pattern in which a cluster of related manuscripts, produced by large and geographically dispersed author teams, came under editorial scrutiny at more than one journal.

The second ground, that multiple references appeared to be irrelevant, is a hallmark that research-integrity specialists associate with certain forms of coordinated manuscript production. Citation manipulation, whether intended to inflate the metrics of particular journals or authors, or simply the byproduct of template-driven writing, undermines the scholarly function of a reference list, which is supposed to document the evidentiary basis of a review’s claims. In a systematic review, where the entire contribution depends on the transparent identification, screening, and synthesis of primary studies, irrelevant citations are not a cosmetic flaw. They make it impossible for readers to verify which studies were actually considered, how inclusion and exclusion criteria were applied, and whether the pooled conclusions follow from the evidence. When such problems surface and authors decline to engage with the journal, editors are left with little alternative to retraction.

The authorship pattern of the retracted paper is itself noteworthy. The article listed authors affiliated with institutions in Indonesia, Saudi Arabia, Russia, Iraq, India, Colombia, Poland, Yemen, Uzbekistan, and Iran, a span of countries that would be unusual for a review of environmental health data focused on any single population. Large, internationally scattered author lists with limited apparent connection to the study subject have become a recognized red flag in the wave of retractions that has swept through the scientific literature in recent years, particularly among papers produced by commercial paper mills. Publication ethics watchdogs and major publishers have documented tens of thousands of retractions in recent years, a substantial fraction of them linked to systematic manipulation of the publishing pipeline rather than to honest error. While the retraction notice in this case does not allege paper-mill involvement, the combination of overlapping text, irrelevant references, and non-response from a dozen or more authors fits a familiar profile.

The non-response of the authors is a significant procedural element. Journals typically follow a due-process protocol when integrity concerns arise: they contact the corresponding authors, describe the concerns, and invite an explanation or rebuttal, sometimes allowing weeks or months for a reply. When authors engage, the outcome may be a correction, an expression of concern, or, if the explanation is unsatisfactory, retraction. When authors do not respond at all, editors must weigh the harm of leaving a flawed paper in the literature against the fairness of acting without the authors’ side of the story. In this instance the Editor-in-Chief concluded that the evidence of overlap and citation problems, combined with the authors’ silence, justified withdrawal of the article and a public statement that confidence in the work could no longer be maintained. The notice also records that none of the authors responded to the publisher’s correspondence about the retraction itself.

For the environmental health community, the retraction carries practical consequences. Systematic reviews are frequently cited in risk assessments, in the design of subsequent studies, and in policy documents that summarize the state of the evidence on pollutant-health associations. A retracted review on PAHs and liver function cannot be used as a summary of that evidence, and researchers citing the field should rely on the primary observational studies and on reviews that have survived editorial scrutiny. At the same time, the underlying scientific question remains open and legitimate. Epidemiologists continue to investigate whether chronic low-level PAH exposure, as measured by urinary metabolites or by proxies such as traffic-related air pollution, is associated with altered liver enzymes, non-alcoholic fatty liver disease, or other hepatic outcomes. Well-conducted reviews that document their search strategies, register their protocols, and report their screening decisions transparently remain the right tool for answering that question.

The episode also illustrates the machinery of post-publication correction that major publishers now deploy. The retraction notice is linked to the article record through the CrossMark service, which allows readers checking a document to verify whether they are viewing the current, authoritative version. The notice identifies the journal, the volume and article number, the publication date, and the relationship to the original paper, and it cites the retracted companion article explicitly so that readers can trace the overlap. This apparatus does not erase the flawed paper from databases, since retracted articles remain online with their notices attached; that permanence is deliberate, preserving the scholarly record while warning readers away from relying on the content. It also creates a public trail that institutions can consult when evaluating the output of researchers involved in retractions.

Ultimately, the case is a reminder that the credibility of environmental health science depends as much on the integrity of synthesis as on the quality of the underlying measurements. Air pollution research has produced some of the most consequential public health findings of the past decades, linking fine particulate matter, nitrogen dioxide, and combustion byproducts to cardiovascular, respiratory, and metabolic disease, and PAHs remain a plausible contributor to hepatic toxicity given their metabolism in the liver and their capacity to form reactive adducts. But every meta-analytic claim inherits the weaknesses of the papers it aggregates, and a review built on irrelevant citations and duplicated material cannot support confident conclusions. The retraction of this systematic review removes a compromised summary from the literature, flags a cluster of related concerns across at least two journals, and underscores the growing vigilance of editors and integrity researchers toward the patterns of overlap, citation manipulation, and authorial silence that have come to define the current era of scholarly self-correction.

Subject of Research: Retraction of a systematic review on polycyclic aromatic hydrocarbon exposure and liver function due to article overlap and citation integrity concerns

Article Title: Retraction Note: Exposure to polycyclic aromatic hydrocarbons and liver function: a systematic review of observational studies

Article References: Retraction Note: Exposure to polycyclic aromatic hydrocarbons and liver function: a systematic review of observational studies. (n.d.). https://doi.org/10.1007/s11869-026-02101-6

Image Credits: AI Generated

DOI: 10.1007/s11869-026-02101-6

Keywords: retraction, polycyclic aromatic hydrocarbons, liver function, systematic review, research integrity, Air Quality Atmosphere & Health, citation manipulation, environmental health, hepatotoxicity, air pollution, Springer Nature, publishing ethics