Scientific publishing plays a vital role in disseminating medical knowledge and supporting education and research. However, not all publications meet the required standards of validity, rigor, and ethical standards. Study validity may be compromised by misconduct or unintentional errors such as data fabrication, wrong methodology, plagiarism, and fraud. Because scientific evidence guides health policy and clinical practice, ensuring research accuracy is critical. Retraction serves as an important mechanism to protect research integrity and prevent the spread of misleading information. However, retractions remain relatively uncommon as their frequency has increased in recent years. Hence, understanding discipline-specific retraction patterns is essential for strengthening the editorial and publication practices. A recent study examined retraction trends and causes in the field of rheumatology.
In this study, the Retraction Watch Database (RWD) was reviewed on January 4, 2025. This database provides a comprehensive and regularly updated record of retracted scientific rheumatology-related articles. Relevant papers were detected by a two-step approach. First, the authors created a list of 65 rheumatology-specific keywords using a large language model. This list was subsequently reviewed and validated by a rheumatologist. Second, the selected articles were manually screened to confirm their relevance. Retracted medical articles outside rheumatology served as comparators. Retraction reasons were classified as figure/data errors, other causes, and misconduct. Journal impact factors were obtained from citation reports. Data were analyzed using appropriate statistical methods as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
A total of 381 retracted articles related to rheumatology published from 1989 to 2024 were identified. Most were original research articles (n = 303). Authors were predominantly from Asia (n = 261 [68.5%]), with China contributing the largest proportion (n = 193; 50.7%), followed by Japan (n = 30; 7.9%). The median journal impact factor was 3.2 (IQR: 2.5-4.7), and the median number of authors per paper was found to be 5 (interquartile range [IQR]: 3-7). Retraction occurred across 211 journals, with 12 journals reporting more than 5 retracted articles. Â
Most papers had multiple reasons for retraction (median 3 [IQR: 1-4]). Scientific misconduct was observed in 287 (75.3%), whereas 57 papers (14.9%) were retracted due to data errors, and 29 papers (7.6%) for other reasons. Data error retractions appeared in higher impact journals (median = 3.3) compared to misconduct-related (median = 2.9) or other retractions (median = 1.9), with p <0.001.
The median time to retraction was 18 months (IQR: 9-46). Retractions increased significantly over time, from 3 during 1990-1999 to 207 during 2020-2023, with p <0.001. Compared with other medical fields, rheumatology papers showed similar retraction patterns, except for a higher proportion of articles originating from Asia, particularly China.
This study’s limitations include uncertain retraction rates, potential detection bias, reliance on current impact factors, and possible missing relevant articles. Strengths include its originality and a systematic analysis of retraction causes, geographic distribution, and temporal trends.
In conclusion, the number of retracted rheumatology papers has increased substantially, mainly due to questionable research practices. However, retractions represent only a fraction of research misconduct, as many cases remain undetected or unpublished. These findings highlight the need for enhanced research integrity awareness, improved peer review processes, and preventive strategies to promote transparency and accountability in scientific publishing.
Reference: Vettori AM, Iudici M. Retractions in rheumatology: trends, causes, and implications for research integrity. Arthritis Care Res (Hoboken). 2026. doi:10.1002/acr.80005
