Analysis of Retracted Publications on Methotrexate

Abstract

Background

Methotrexate (MTX) is a frequently used antifolate agent in rheumatology, oncology, and dermatology. Retractions are crucial for preserving scientific integrity by fixing the literature when errors, ethical violations, or data falsification are detected. This study seeks to comprehensively examine retracted MTX publications, detecting trends, reasons, and potential ramifications.

Methods

A descriptive cross-sectional study was performed by searching PubMed for retracted MTX publications without temporal constraints. Bibliometric data were retrieved, encompassing publication details, journal indexing, citation metrics, and Altmetric Attention Scores (AAS). Reasons for retraction were categorized.

Results

Thirty-six retracted papers were identified. The predominant cause for retraction encompassed data concerns (n = 15), fraudulent activity (n = 11), and manipulation of the peer review process (n = 10). China recorded the most retractions (n = 14), followed by the USA (n = 6) and Egypt (n = 3). The median period until retraction was 712 (91–9,893) days. Most retracted articles were published in non-rheumatology journals. Citation analysis showed a median of 5 (0–105) citations. Additionally, the median AAS was 1 (0–14).

Conclusion

Data integrity concerns, fraudulent activities, and manipulation of peer review processes constitute major obstacles in MTX-related research. The considerable number of retractions in certain countries underscores the need to promote research ethics and measures to protect scientific integrity. Enhancing peer review, implementing more intense data transparency, and promoting post-publication evaluation are essential measures to protect scientific integrity and uphold confidence in clinical recommendations.

INTRODUCTION

Methotrexate (MTX) is an antifolate medication widely used in various medical fields, including rheumatology, oncology, and dermatology. MTX serves as a fundamental treatment in rheumatology, functioning as the disease-modifying antirheumatic medication for rheumatoid arthritis and several autoimmune disorders, owing to its capacity to regulate immune responses and diminish inflammation.1,2,3 In oncology, MTX is a notable chemotherapeutic drug for treating hematological malignancies, including leukemia and lymphoma. In addition, MTX is utilized in dermatology for psoriasis and gastroenterology for inflammatory bowel disease.4,5

In the digital technology age, retractions are essential protection for upholding scientific integrity. Retractions are provided when a published paper is discovered to contain significant errors, unethical breaches, data fabrication, plagiarism, or methodological flaws, thereby ensuring that unreliable material does not mislead scientists, physicians, or authorities.6 Retractions are crucial for correcting scientific documents, averting the dissemination of erroneous outcomes, and preserving public confidence in research. The rising utilization of artificial intelligence (AI) in scientific publishing raises worries about AI-generated fraud, data manipulation, and inappropriate authorship behaviors, which can result in an increase in retractions.7 Furthermore, editorial deficiencies, including fake peer reviews and insufficient oversight, have been recognized as frequent factors contributing to retractions, underscoring the necessity for enhanced quality assurance protocols.8,9 The manipulation of peer review structures, wherein authors provide fraudulent reviewer profiles or employ third-party firms to produce favorable critiques, compromises the integrity of scientific publishing.10

This study aimed to evaluate retracted MTX-related articles with the following objectives:

• • Examining the temporal distribution of retractions, covering peak years and the period between publication and retraction.

• • Determining which countries have the most retracted MTX-related publications.

• • Categorizing the main reasons for retraction and examining if particular types of misconduct are more frequent in MTX-related publications.

• • Examining the citation metrics of retracted documents and analyzing the altmetric data.

METHODS

Study design

This descriptive cross-sectional study systematically discovered and analyzed retracted documents regarding MTX without any time restrictions. The inquiry was performed in PubMed utilizing the following search phrase: "retracted publication"[publication type] [pt] and MTX. The dataset was last updated on February 10, 2025, guaranteeing the thorough inclusion of all relevant retracted publications up to that date. Only papers marked as 'retracted publications' were considered, whereas errata and corrections were not evaluated. This methodology guaranteed a targeted examination of retractions related to MTX in the scientific literature.

Data extraction

The bibliographic information of the retracted documents was methodically acquired, exported, and captured in an Excel file for further assessment. The obtained data comprised noteworthy publication details involving:

• • Title: The precise title of the retracted paper.

• • DOI number: The distinctive digital identification assigned to the publication.

• • Authors: The researchers participating in the article.

• • Publication date: The date when the article was initially published.

• • Retraction date: The formally documented date of retraction.

• • Time to retraction: The interval (in days) between the publication and retraction dates.

• • Journal: The journal in which the article was published.

• • Article categorization: The category of the article (original article, review, case report, etc.).

• • Country of the corresponding author: The country associated with the corresponding author.

• • Reason for retraction: The justification given by the journal or publisher.

Furthermore, citation metrics for the retracted papers were obtained from the Web of Science, facilitating an evaluation of the influence of these publications.11 The journals that published the retracted publications were then classified according to their standing in Web of Science, particularly into:

• • Science Citation Index Expanded (SCIE)

• • Emerging Sources Citation Index (ESCI)

To elucidate the prominence and impact of the journals that published the retracted papers, each journal’s quartile ranking and impact factor (2023 value) were recorded. The journal with the highest impact factor was determined. Details of articles not indexed in the Web of Science were documented individually.

The journals ‘subject area and category’ were recorded via SCImago Journal & Country Rank.12

The Altmetric toolbar was employed to examine Altmetric data, offering insights into the digital interaction and distribution of retracted publications concerning MTX. Altmetric Attention Scores (AAS) were acquired to measure the attention garnered by this content across multiple online platforms. The AAS is a weighted indicator that indicates the real-time attention an article has garnered. This score is calculated using a proprietary algorithm that allocates distinct weights to different categories of web mentions. The algorithm is perpetually refined to maintain precision and pertinence.13

Categorization of retraction reasons

A structured categorization approach was implemented to classify the reasons for retraction comprehensively. The classification was founded on the underlying factors specified in the retraction notices, ensuring a thorough and standardized method for analyzing retracted papers. The classification was as follows8,14:

• • Error: Retractions in this category resulted from methodological deficiencies, including insufficient data gathering, inaccurate data presentation and reporting, misinterpretation of results, or defective study design. The integrity of the study can have been compromised by honest errors or unintentional oversights, which could result in errors.

• • Fraud: This category covered occurrences of data falsification, fabrication, or manipulation, encompassing intentional modifications of numbers, visuals, or experimental results. Fraudulent practices undermine research integrity and breach ethical standards in the scientific community.

• • Author disagreements and conflicts: This subgroup comprised incidents in which authorship concerns resulted in retraction. Examples include the unauthorized publication of research without the consent or approval of one or more researchers, including fictitious authors, or issues between authors and funding authorities regarding intellectual contributions or project ownership.

• • Duplication: This category comprised the publication of identical or considerably equivalent material in different journals. Such cases undermine the uniqueness and originality of scientific contributions and violate publication ethics.

• • Ethics issues: This class contained retractions that were the result of ethical misconduct, such as the failure to acquire informed consent from participants in the study, the absence of permission from an institutional ethics committee, or the infringement of research ethics guidelines.

• • Peer review issues: Retractions in this group were linked to fraudulent, biased, or manipulative peer review techniques. This encompassed cases in which the peer review process was undermined, such as undisclosed conflicts of interest among reviewers, fake reviewer identities, or author-influenced assessments.

• • Plagiarism: This category includes incidents of unauthorized utilization of another researcher's proprietary knowledge involving text, graphs, tables, research methodologies, or hypotheses. Occurrences of self-plagiarism, wherein researchers reissued their own material without appropriate reference or disclosure, were also categorized under this group.

• • Data concerns: Publications were retracted due to issues regarding the validity, trustworthiness, or integrity of the data utilized in the research. This encompassed concerns over inadequate data transparency, absence of reproducibility, and distorted analysis of stats.

• • Irrelevant citations: This subset encompasses instances where publications are retracted due to excessive, improper, or manipulated citations, frequently aimed at artificially enhancing citation measures or favoring certain researchers or journals.

• • Unknown: In certain cases, the retraction notices failed to explain the reasons adequately. These instances were classified as “unknown” to recognize the absence of explicit rationale in the existing documentation.

In publications with several reasons for retraction, each reason was individually recorded and classified to provide a comprehensive picture of the relevant elements.

Retraction notices assessments and validation

To improve the accuracy and reliability of the categorization, two independent researchers thoroughly examined all retraction reasons. Each researcher independently evaluated and classified the reasons for retraction according to the available retraction notifications and additional details. After the separate assessments, the researchers juxtaposed their groupings. In instances of disagreements, a consensus-driven resolution method was utilized. Two researchers jointly examined the contradictory examples, deliberated on the discrepancies, and reached a consensus regarding the final classification. This stringent vetting approach secured objectivity and reduced the possibility of bias in categorizing retraction reasons.

Data processing

The retrieved data were systematically organized and analyzed with Microsoft Excel. Results were reported as numbers (n), percentages (%), and median (minimum–maximum) values.

Ethics statement

This study did not include human or animal participants, as it was exclusively based on examining publically accessible retracted papers. Approval from an institutional ethics board was unnecessary as the research utilized open-access bibliometric and scientometric data.

RESULTS

Study selection

A total of 36 documents were obtained from PubMed utilizing the outlined search criteria. All articles related to MTX. Moreover, due to the lack of discipline-specific inclusion criteria, all listed publications were incorporated into the analysis.

Risk of bias

Not evaluated.

Time of publication and retractions

The earliest retracted article was published in 1985, while the most recent was published in 2023 (date range 1985 to 2023). The first retraction occurred in 2004, while the last was conducted in 2024 (date range 2004 to 2024). Fig. 1 depicts the yearly distribution of retracted publications. The number of retracted documents peaked in 2023 (n = 8) and 2024 (n = 8). The median lag between publication and retraction was 712 (91–9,893) days.

Fig. 1. Distribution of retracted publications by year.

Country analysis

Fig. 2 illustrates the countries of the retracted publications. Articles were submitted from fifteen distinct nations. China (n = 14) was in the first position, the USA (n = 6) was in the second, and Egypt (n = 3) was in the third.

Fig. 2. Country-based distribution of retracted publications. One corresponding author has affiliations from two different countries.

Journals and impact factors

The journals with more than one retracted publication were Evid Based Complement Alternat Med (n = 4), Comput Math Methods Med (n = 2), and Biomed Res Int (n = 2). Journal analysis was performed based on subject area and category. Medicine (n = 15) ranked first, followed by Biochemistry (n = 13), Genetics and Molecular Biology (n = 13), Toxicology and Pharmaceutics (n = 5), and Pharmacology (n = 5).

A total of 29 papers were published in SCIE-indexed journals, and 4 were featured in the ESCI category. Furthermore, 3 articles were not available on the Web of Science. In the quarter-based analysis via Web of Science, 13 articles were published in Q1 journals, 8 in Q2 journals, and 12 in Q3 journals. Additionally, no articles were published in Q4 journals.

The median journal impact factor was 2.90 (0.70–56.70). Ann Oncol had the highest impact factor, with a value of 56.70.

Article types and citations

Of the total, 28 publications were original articles, 4 were reviews and 4 were case report-series. The median citation of retracted publications was 5 (0–105).

Retraction reasons

Retraction reasons were categorized as specified in the methodology. More than one reason was possible for an article. The reasons for retraction were as follows: data concerns (n = 15), fraud (n = 11), peer review issues (n = 10), irrelevant citations (n = 10), duplication (n = 9), plagiarism (n = 5), error (n = 5), ethics issues (n = 3), author disagreements (n = 2), and unknown (n = 1) (Fig. 3).

Fig. 3. Distribution of retraction reasons.

Altmetric analysis

The median AAS value was 1 (0–14).

Reporting biases

There was no reporting bias.

DISCUSSION

This article thoroughly examined retracted papers regarding MTX, elucidating retraction trends, triggering factors, and bibliometric features. The principal findings indicate that retractions have risen with time, reaching their peak in 2023 and 2024, with a median interval of 712 days between publication and retraction. The predominant reasons for retraction comprised data concerns (n = 15), fraud (n = 11), and peer review issues (n = 10). Particularly, the majority of retractions originated from China (n = 14), followed by the United States (n = 6) and Egypt (n = 3). Moreover, most retracted publications were published in non-rheumatology journals, with a relatively low median citation count (median = 5) and a restricted Altmetric effect, with a median value of AAS was 1.

The rising number of retractions in recent years indicates heightened scrutiny of published studies. The documented peak in 2023 and 2024 may indicate strengthened retraction mechanisms, more substantial journal supervision, and increased awareness of scientific misconduct. This corresponds with overarching trends in scientific publishing, where apprehensions regarding research integrity have resulted in a rise in retractions across various disciplines.15,16 The median time to retraction, which is 712 days, highlights the delayed identification and rectification of flawed publications, permitting inaccurate findings to be disseminated for prolonged durations.17 At this stage, the goal should be to minimize the lag period while hindering the spread of incorrect or deceptive data.

China exhibited the highest number of retracted MTX-related articles, a trend that aligns with broader retraction analyses in several biomedical disciplines.18,19 This might be attributed to elevated research output, publication pressure, and the development of research integrity policies in specific regions. The swift expansion of scientific research in China, alongside academic standards linking career progression to publication statistics, may result in expedited or substandard research, resulting in retractions.

Notably, most retractions transpired in journals listed in the SCIE rather than ESCI, indicating that higher-level journals are not immune from scientific misconduct. Moreover, the frequency of retractions in non-rheumatology journals signifies that research on MTX is extensively circulated across several disciplines, including pharmacology, oncology, and toxicology. This raises concerns regarding the potential for erroneous findings to impact various fields beyond rheumatology.

Retracted MTX-related papers primarily consisted of original articles, highlighting that methodological deficits, ethical issues, or data integrity difficulties most impacted fundamental scientific contributions. Citation analysis indicates that certain retracted studies have impacted future studies.20 The comparatively low AAS signifies restricted online interaction, implying that these publications failed to elicit much public discussion.

The high frequency of retractions due to fraud, data issues, and peer review manipulation highlights persisting flaws in the scientific publication system.21 These challenges underscore the need for increased monitoring, especially as technology improvements, such as AI-assisted article preparation, make falsifying or distorting research data easier.22 Furthermore, the abuse of peer review mechanisms, whether through false reviewer identities, manipulated assessments, or compromised editorial oversight, jeopardizes the integrity of academic communication. To solve these issues, journals could impose stricter data openness standards, requiring authors to provide raw datasets and reliable documentation. The use of open peer review, in which reviewer identities and opinions are made public, could improve accountability and lessen the risk of biased or fraudulent evaluations. Furthermore, post-publication verification measures, such as independent data audits and reanalysis campaigns, could serve as an additional barrier against misinformation, reducing the frequency of retractions while maintaining the scientific record's reliability.23,24,25

This study involves several limitations requiring addressing. The limited sample size of retracted MTX-related documents may restrict the generalisability of the findings. The results offer significant insights but may not comprehensively reflect wider trends in retractions across various fields of medicine. The analysis relied on PubMed, potentially omitting retracted publications largely indexed in alternative databases. The formatting of retraction notes was inconsistent among the journals.

The present article thoroughly analyzes retracted MTX publications, emphasizing that fraud, data concerns, and peer review manipulation are the predominant causes of retraction. A significant observation is the high number of retractions from China, potentially indicative of systemic pressures in academic publishing, including publication incentives, research misconduct, or difficulties in editorial control. Furthermore, most retracted articles were published in non-rheumatology journals, indicating that these disciplines may exhibit differing quality control standards. Despite the relatively low amount of MTX-related retractions, preserving the integrity of published research is essential, especially for drugs such as MTX, which underpin rheumatology treatment protocols. Enhancing peer review processes, implementing more rigorous data disclosure regulations, and promoting post-publication evaluation could bolster the integrity of scientific literature and guarantee that clinical guidelines are founded on high-quality evidence.