A Comprehensive Bibliometric Analysis of Biomedical Research Waste: Current Situation, Development, and Trends

Abstract

Since the 20th century, the rapid increase in academic publications has turned research waste into a significant challenge in scientific research. However, there is currently no comprehensive bibliometric analysis available to evaluate the progress in this field. In this study, we retrieved all relevant articles published between January 1, 2004, and December 31, 2024, from the Web of Science (WoS) database, yielding a total of 876 articles. Various tools, including CiteSpace, VOSviewer, and R-bibliometrix, were employed for comprehensive analysis. The results revealed that these articles collectively received 39 768 citations, with an average of 45.4 citations per paper. Both the number of published articles and the growth rate saw a rapid increase after 2014. The United States and the England emerged as the leading countries in research output. A keyword analysis identified 3 main themes: (1) the types of trials where research waste is most prevalent, (2) the primary factors contributing to research waste in trials, and (3) strategies to mitigate research waste. Notably, the experimental design phase has been identified as the primary source of research waste. Additionally, research waste was most prevalent in internal medicine, followed by surgery and pediatrics. Through these analyses, we provide valuable insights into the characteristics of research waste over the past 2 decades, highlighting an increased focus on this issue in the future.

Figure 1

Schematic diagram of research waste analysis. This figure illustrates the overall workflow of the study, including literature screening, basic analysis, and detailed analysis.

Figure 2

(A) Annual literature publications. The yellow bars represent the number of articles published each year, while the blue curve denotes the fitted trend, illustrating the overall publication trajectory. (B) Annual citations of the literature and significant milestones. The horizontal axis depicts the number of citations, while the vertical axis represents the corresponding year. Key publications and major events related to research waste are labeled for each pivotal year.

Figure 3

(A) Number of national publications from 2004 to 2024. The length of each bar corresponds to the number of publications from each country or region, with the United States and England displaying significantly higher publication counts compared to other countries. (B) Global network of countries. The size and color intensity of each node represent the level of influence of each country in the field of research waste, while the connecting lines illustrate international collaboration.

Figure 4

Inter-agency collaboration. This figure, generated using VOSviewer, illustrates the collaborative network between research institutions and their research connections.

Figure 5

Collaborative relationship among authors. Each node is represented by a different color, corresponding to authors belonging to a specific group. The size of each node indicates the frequency of co-occurrence, and the connections between nodes indicate co-occurrence relationships between authors.

Figure 6

Double mapping overlay of journals. Individual nodes in the figure represent different journals, and the width of the lines indicates the strength of the citation relationship, visualizing the citation relationship between the citing journal on the left and the cited journal on the right. The left side shows the journals that cited other journals, and the right side shows the cited journals.

Figure 7

(A) Keywords heat map. The size of the keywords in the figure represents the frequency with which they appear in the research waste article, with larger words indicating higher frequency. (B) Keywords cluster analysis. This figure illustrates the keyword clustering analysis in the field of research waste. Each color represents a keyword group, which are clustered together by their relevance, and the keywords are numbered to show their importance in research waste.

Figure 8

Top 20 keywords with the most robust citation bursts. The figure shows the top 20 keywords with the strongest citation bursts, highlighting the sudden shift in information, represented by the red bursts on the timeline, indicating a rapid increase in citations. These major shifts are typically identified using Kleinberg-J’s burst detection method in CiteSpace.

Figure 9

(A) Changes in research waste types every 4 years. The horizontal axis of the figure represents the time period (every 4 years) and the vertical axis represents the number of articles published. Different colors represent different steps, mainly: registration (red), research design (blue), termination (green), results reporting (purple), publication (orange), citation (yellow), and those that are not explicitly categorized are labeled as other (brown). (B) Research design category diagram. This figure illustrates the main types of research waste in experimental design.

Figure 10

(A) The subject classification and proportion of publications from 2004 to 2024. The sectors in the figure show the distribution of different medical disciplines in the topic of research waste and the percentage for each field reflects the relative share of that field in the literature related to research waste. (B) The classification and number of internal medicine publications. The radar chart illustrates the number of publications in each subdiscipline of internal medicine within the context of research waste. (C) The classification and number of surgical publications. The radar chart illustrates the number of publications in each subdiscipline of surgery within the context of research waste.