https://orcid.org/0009-0002-0425-7375
Abstract
Delayed graft function (DGF) is an early complication after kidney transplantation. The literature on DGF has experienced substantial growth. However, there is a lack of bibliometric analysis of DGF. This study aimed to analyze the scientific outputs of DGF and explore its hotspots from 2013 to 2023 by using CiteSpace and VOSviewer. The 2058 pieces of literature collected in the Web of Science Core Collection (WOSCC) from 1 January 2013 to 31 December 2023 were visually analyzed in terms of the annual number of publications, authors, countries, journals, literature co-citations, and keyword clustering by using CiteSpace and VOSviewer. We found that the number of papers published in the past ten years showed a trend of first increasing and then decreasing; 2021 was the year with the most posts. The largest number of papers was published by the University of California System, and the largest number of papers was published by the United States. The top five keyword frequency rankings are: ‘delayed graft function’, ‘kidney transplantation’, ‘renal transplantation’, ‘survival’, and ‘recipients’. These emerging trends include ‘brain death donors’, ‘blood absence re-injection injuries’, ‘tacrolimus’, ‘older donors and recipients’, and ‘artificial intelligence and DGF’. In summary, this study reveals the authors and institutions that could be cooperated with and discusses the research hotspots in the past ten years. It provides a reference and direction for future research and application of DGF.
Introduction
Delayed graft function (DGF) is one of the most frequent early complications following kidney transplantation [Citation1]. The term is defined as follows in the majority of recent research: Within seven days following surgery, dialysis is administered [Citation2]. According to pertinent studies [Citation3], the incidence of DGF following kidney transplantation with organ donation following a citizen’s death ranges from 20% to 50%, while the incidence following kidney transplantation with a living donor is between 4% and 10%. The risk of kidney loss after transplantation will rise if DGF occurs and treatment is delayed. To lower the incidence of DGF, numerous studies are continuously being conducted on the subject [Citation4].
VOSviewer and CiteSpace are resources for examining scientific literature. They are mostly employed to assess the body of work in particular domains in order to illustrate the direction and organization of a given discipline’s research [Citation5]. In the big data era, researchers may use these programs to visualize and analyze vast amounts of data, allowing them to demonstrate trends, hotspots, and knowledge field development over time [Citation6].
There isn’t a bibliometrics analysis on delayed renal function recovery at the moment, and it’s still unknown what the state of the field is at and where the research is now concentrated. In order to provide references for further research and application, this paper uses CiteSpace and VOSviewer software to visually analyze the annual number of publications, authors, publishing institutions, countries, journals, cited references, keywords, and other aspects of DGF over the previous five years.
Materials and methods
Search strategy
On 19 January 2024, online literature data were collected from the Science Citation Index Expanded database using a search strategy consisting of the following terms: TS= (kidney transplant OR kidney transplantation OR renal transplant) AND TS= (Delayed graft function OR DGF), and the period of the literature search is from 1 January 2013 to 31 December 2023.
Data analysis
The literature’s inclusion-exclusion process was carried out independently by two writers. The documentation required for the study, including the complete documentation and reference sources, is first exported in the ‘Endnote Desktop’ format, imported into Endnote 20, with two authors reading the article topics and abstracts independently, excluding the non-relevant documentation, and, if the authors differ, invited to the third author to participate in the discussion, then exported the filtered documentation in the pure text format, and imported to CiteSpace 1.6.18 and VOSviewer 6.2.R4 for analysis. Duplicate studies were removed, but the study did not restrict the kind of articles that were included. Based on the information gathered about the literature, a visualization bibliometric network was built using CiteSpace 6.2.R4 and VOSviewer 1.6.18. The analysis process is shown in .
Figure 1. Workflows of Bibliometric analysis: Outlines the step-by-step process of the bibliometric analysis, including data extraction, software utilization, and visualization techniques..
We did a visual analysis of some of the material in the literature, such as the annual number of publications, authors, countries, journals, literature co-citations, and keywords, and selected the corresponding options on the software operations page for analysis. In the text, using CiteSpace and VOSviewer, the two softwares created 17 graphs, of which was created with the use of Citespace. In the process of mapping, too many nodes made the image not intuitively clear enough, so adjusted the k value (k = 5) and properly changed the location of the nodes and the color of the graph to make it more beautiful; the rest of the picture has no parameter adjustment. Furthermore, were created using VOSviewer, and since it does not affect understanding, the parameters we use are the default values of the system.
Figure 2. Analysis of authors analysis by CiteSpace: Provides an in-depth exploration of the authorship patterns, highlighting prolific authors and potential collaborations.
Figure 3. Analysis of agencies Co-occurrence by CiteSpace (k = 5): visualizes the collaboration and frequency of engagements among different agencies.
Figure 4. Analysis of country Co-occurrence by CiteSpace: Shows the country-wise distribution and collaboration in research, as analyzed by citespace.
Figure 5. Burst analysis of the cited documentation by CiteSpace: Identifies significant surges in citations over the specified period.
Figure 6. Analysis of Keyword Co-occurrence by CiteSpace (k = 5): Maps the network of keywords, showing their frequency and relationships.
Figure 7. Keyword clustering analysis by CiteSpace (k = 5): groups keywords into clusters to identify thematic concentrations in the literature.
Figure 8. Keyword and cited document Co-occurrence Network by CiteSpace (k = 5): connects keywords with the most frequently cited documents, illustrating the thematic focus of the literature.
Figure 9. Detection of the Strongest citation Bursts from 2013 to 2023 by CiteSpace: Highlights the most impactful citations and their temporal distribution over the decade.
Figure 10. Detection of the Strongest citation Bursts from 2019 to 2023 by CiteSpace: Focuses on the most significant citation bursts in the more recent half-decade.
Figure 11. Citation-author Network by VOSviewer: Maps the relationship between authors and their citation impacts in the field.
Figure 12. Citation-organizations Network by VOSviewer: Depicts the interlinkages between citing organizations and the impact of their contributions.
Figure 13. Co-authorship organization Network by VOSviewer: Reveals the collaborative networks between organizations based on co-authorship data.
Figure 14. Analysis of country Co-occurrence by VOSviewer: Presents a similar analysis as in , but utilizing VOSviewer for a comparative perspective.
Figure 15. Citation sources Network by VOSviewer: Illustrates the sources that contribute significantly to the citations in the study.
Figure 16. Citation-author Network by VOSviewer: Maps the relationship between authors and their citation impacts in the field.
Results
Publication volume and trend
In all, 2058 literary works were chosen and acquired for this investigation. As illustrated in , the number of published papers has been on the rise in recent ten years, peaking in 2021 and continuing to decline thereafter.
Figure 17. The number of DGF documents from 2013 to 2023: Presents the annual distribution and trend of DGF documents over the specified period.
Authors analysis
There were 518 authors (N = 518) in the recent studies on the delayed recovery of kidney transplantation function, which is a considerable number. With 18 papers to date, or 0.875% of all authors, Dorry L. Segev was first in terms of total number of papers, followed by Sandesh. Parajuli with 16 papers. That amounts to 0.777%. A higher node radius in the co-occurrence graph denotes a higher number of publications, and the lines connecting nodes show a co-occurrence relationship. A closer association between nodes is shown by a thicker line [Citation7]. As seen in , the result displays E = 918, meaning that 918 authors collaborated with the other 518 authors. As illustrated in , the citation-author network highlights writers that have contributed significantly, while authors that share the same hue indicate close collaboration. In addition, in order to determine the H-index of the top 10 authors, we ranked their published articles from highest to lowest citation times. When a paper’s serial number exceeds its citation count, subtract one from the resulting sequence number to give the H-index. For more information, see . It is noteworthy that Dorry L. Segev continues to be the author with the highest H-index. The H-index still has some drawbacks, though. The H-index has comparative importance for researchers with nearly the same academic career time, for example, because the actual contributions of the authors are not taken into consideration and because the H-index primarily depends on the researchers’ academic career duration. Therefore, the H-index can only be used as a reference for evaluation authors.
Table 1. Top 10 authors with the highest number of publications by CiteSpace.
Analysis of agencies
University of California System was the largest agency with 2.770%, showing N = 109, E = 222 in the results of the agency summary chart, indicating a total of 109 agencies and 222 cooperating agencies, for details, see . In addition, the results shows University of California System (50 documents), Institution National de la Sante et de la Recherche Medicale (49 documents) and University Paris Cite (43 documents) as the most influential institutions in the field, see and . It is noteworthy that Yela University, University of Pennsylvania (university penn) and University Hospital have had significant collaboration with other institutions, see .
Table 2. Top 10 agencies with the highest number of publications by CiteSpace.
Analysis of country
The country with the highest number of submissions was the United States (580), accounting for 28.18%, followed by China (184), which accounted for 8.94% of the total volume, As illustrates. In the results of the country summary chart showed N = 83, E = 228, indicating that a total of 83 countries had published articles in the Web of Science database, and there are some links between countries, As and illustrates. And the United States, England, and Germany maintain close cooperative relationship with other nations.
Table 3. Top 10 countries with the highest number of publications by CiteSpace.
Analysis of journals and quoted literature
Between 2013 and 2023, a total of 2058 articles were published, containing eleven different types of literature, as shown in . It is noteworthy that the vast majority of research papers (77.891 per cent) were followed by review papers (11.127 per cent).
Table 4. Types of journal articles by CiteSpace.
shows the top 10 journals ranked by number of published and quoted, with an influence factor range of 0.784 to 74.699. The influence factor on academic journals is an indicator of how many articles the journal has recently published are quoted annually on average. It is often used to assess the impact of journals in their respective fields. In addition, visualization of Citation sources and Co-citation sources is done using VOSviewer, and . Nodes of the same color represent a cluster, representing a height of collaboration.
Table 5. High-cited journals ranked in the top 10 of the DGF by CiteSpace.
The quoted frequencies are analyzed, as depicted in . ‘Siedlecki A, 2011, AM J TRANSPLANT. V11, P2279, DOI: 10.1111/j.1600-6143.2011. 03754. x ‘[Citation8] is the document that is most commonly mentioned. This review, which was published in the American Journal of Translation, examines the risk factors for DGF from donor identification through postsurgery and beyond. It also outlines the substantive mechanisms of ischemic and immune kidney injury and discusses preventative measures for DGF, with an emphasis on therapeutic targets that lessen immunological response and alleviate ischemic disease. We can have a more intuitive understanding of the mechanism and preventive measures of DGF, which is conducive to laying a theoretical foundation for the subsequent study of DGF. Furthermore, Yarlagadda SG [Citation9] is the author of the second most cited piece of literature. She carried out a systematic review and meta-analysis of the long-term relationship between DGF and graft survival, demonstrating that patients with DGF had higher serum creatinine and a higher risk of acute rejection following transplantation than patients without DGF, and that DGF was not associated with patient survival at five years of follow-up. Irish WD cites them as number three [Citation10]. They conducted multivariate logistic regression analysis on 24,337 dead donor kidney transplant recipients and proposed a model for predicting DGF following kidney transplantation. As a sort of readily available tool for predicting DGF, they created a nomographic chart, a description of the relative contribution of risk factors, and a new web-based calculator (http://www.transplantcalculator.com/DGF), also can be used for predicting DGF and transplantation of long-term results.
Keyword co-occurrence and clustering analysis
Key words are natural language words that express the characteristics of the topic of the thesis, with substantive meaning and unstandardized treatment, while high-frequency keywords can reflect the topics of focused research in the field of study [Citation11]. Keyword co-occurrence analysis results are shown in , the larger the node in the Figure indicates the volume of submissions [Citation12], the article according to the frequency of the use of keywords, extract the top 20 frequencies of the keyword, the top five key words are: delayed graft function, kidney transplantation, renal transplantation, survival, recipients, see for details. A cluster chart was obtained by reference to cluster labels in keywords, and a cluster of keyword charts in the relevant fields of DGF research, see , showing a total of 7 clusters, namely: showing a total of 11 clusters, namely: #0 Kidney transplantation, #1 Expression, #2 cold storage, #3 Delayed graft function, #4 Acute rejection, #5 clinical research. #6 outcome. Study showed that Q > 0.3 indicates remarkable division structure, S > 0.5 indicates rational clustering, and when S ≥ 0.7 indicates convincing cluster results [Citation13], the key word cluster chart of the study is significant, reasonable, and the result is convincing (Q = 0.5273, S = 0.7539). Furthermore, as the illustrates, we combined keywords with citation data for analysis. According to the results, Siedlecki A [Citation8] has been cited the most, and patients’ survival and delayed graft function are the keywords associated with this node. Debout A [Citation14] was the second most cited. Acute kidney injury, survival benefit, cold ischemia time, impact, risk, unite states, and transplant function are the main nodes associated with it. They discovered that for every hour of cold ischemia, there was a substantial increase in the probability of graft failure and death following kidney transplantation. The terms ‘graft survival’,’ ‘time’,’ ‘donation’,’ and ‘cardiac death’ were linked to Summer DM [Citation7], which came in third place in terms of citations. The findings of their investigation indicated that whereas kidney was more vulnerable to circulatory death. According to these findings, the pathophysiology and contributing factors of DGF are the primary research areas for DGF. Numerous academics are also continuously investigating approaches to lower the incidence of DGF.
Table 6. Top 20 high frequency keywords in DGF research by CiteSpace.
Keyword burst detection
As seen in , this study’s keyword burst detection was carried out in the previous ten and five years, respectively. The findings indicate that throughout the last 10 years, the most common research methodologies employed were randomized controlled trials (RCTs) in 2013–2014 and 2016–2017, risk factor analysis in 2014–2015, Meta-analysis in 2016–2019, clinical practice in 2018–2019, and its focus will be on clinical research from 2018 to 2021. Regarding perioperative medication use in kidney transplant surgery, the most commonly used medications were Mycophenolate mofetil in 2013–2014, Balliximab induction in 2013–2015, Antithymocyte globulin (ATG) in 2014–2017, and tacrolimus in 2021–2023. In addition, there were additional investigations on the pathophysiology and indicators of DGF from 2013 to 2016. Numerous academics started researching kidney transplant donation and selection criteria after 2018. Over the past 10 years, the primary outcome indicators have been the recipient’s long-term prognosis, creatinine, graft function, and survival benefit. It is important to note that the results of the analysis we performed five years ago did not significantly differ from those obtained ten years ago. Secondly, many studies have begun to use the method of predisposing scores to reduce mixed bias [Citation15], with the basic principle of predictive scores being to replace multiple synergistic variables with a single percentage, to balance the distribution of synergies between groups and control groups, and to treat mixed factors in non-randomized studies in a similarly randomized way, reducing selective bias [Citation16]. What’s more, Changes in glomerular filtration rate over one year were also used to assess graft function.
Discussion
Current status of DGF research for ten years
We analyzed 2058 articles from 1 January 2013 to 31 December 2023.With the advancement of science and technology, kidney transplantation technology is becoming more mature, and more and more researchers are starting to focus on post-renal transplant complications to increase the survival time and survival rates of transplanted kidneys. In almost 10 years of research, research in this area has been active, indicating that the DGF is still receiving widespread attention in recent years, but the volume of submissions has declined in 2021, possibly due to a decrease in the number of renal transplant surgeries caused by the outbreak of COVID-19. In the last 10 years, there have been more authors in the field of the study, and the direct links between them are closer. Dorry L. Segev is the most notable in the volume, and the most cited writer is Siedlecki A. The University of California System is the institution with the highest volume of submissions among the issuing agencies, and the University of Pennsylvania is the one with the closest link between the other institutions. Through analysis of authors and institutions, potential collaborators and institutional information can be provided to researchers willing to study the DGF in the future. In addition, the most quoted journal is TRANSPLANTATION, while the largest number of quoted literature is published in AMERICAN JOURNAL OF TRANPLANTION. It is worth noting that DGF’s research is more prominent in the United States, and its cooperation with other countries is the closest, the analysis may be related to the early start of kidney transplantation in the US, the rapid development of science and technology [Citation17]. A total of 68 countries have submitted their findings in the last 10 years, and cooperation between countries can be strengthened in future if conditions permit, thereby enhancing the national scientific research capacity and quality in this field.
DGF research hotspots
We used two software tools to analyze keywords. Through this process, we have identified several emerging research trends in the field, including ‘brain death donors’, ‘blood absence re-injection injuries’, ‘tacrolimus’, and ‘older donors and recipients’, ‘Artificial Intelligence and DGF’. These areas are summarized below.
Donation after Brain death
Depending on the donor’s organ origin, kidney transplants can be divided into a donor kidney after death and a living kidney. Donation after Brain death (DBD) and donation after circulatory death (DCD) are the main sources of kidney donations.
The concept of brain death was introduced in the late 1960s [Citation18], before which most of the organs used for transplants were donated to DCD. In 1968, the United States adopted the Harvard Standard on Brain Death, which legally guaranteed the legality of organ donation after brain death, effectively expanded the source of donors, and promoted the steady development of clinical surgical organ transplants. High incidence of DGF is caused by anxiety about surgery and inherent longer periods of thermal absence. Associated with worse long-term results, organ transplants for DCD have gradually decreased [Citation19], but over the past decade, many countries around the world have reintroduced DCD to reduce the waiting time for kidney, accounting for approximately 20% of organs transplanted in Switzerland and the United States, up to 40% in the United Kingdom [Citation20].
Many studies have demonstrated that there is no difference between the transplant survival rate of DCD and DBD, but the incidence of DGF is higher [Citation21,Citation22]. However, the current DGF studies on the impact of DBD and DCD two types of transplant survival rates are less relevant, de Kok MJ [Citation23]. investigated the effect of DGF on the survival rate of both DBD and DCD types, and found that DGF seriously affects the 10-year transplant rate of DBB transplant plants, but DG F does not affect the transplant life rate in DCD transplant plant, and the obvious difference in the effect on DBB and DCD transplant growth rates is significant, but it difficult to explain. The reason may be related to the activation of donor-type-specific recovery pathways in DCD movements, but there is little research in this area, and one possible explanation for this phenomenon is the presence of more serious transplant-related damage in DBD transplant plants. Another explanation from Histology and gene expression is that the different effects reflect the differences in ‘graft resilience’, that is the capacity of transplant plants to address negative environmental changes, and the DCD donor kidney has more ‘graft resilience’ than the DBD transplant. Saat TC [Citation24] detected levels of expression of inflammation, Using a mouse model, cell protection and damage genes at different times after removal of organs, and found that the gene expression spectrum of DBD kidneys had inflammatory and damage responses to brain death. In contrast, the DCD kidney only exhibits slight upgrading of inflammatory and damaged genes. De Kok MJ [Citation23] also conducted studies in the area of organology and gene expression and found that the different influence of DGF between DBD and DCD graft were associated with the activation of donor-type-specific recovery pathways in DCD. However, fewer such studies could lead to new perspectives on the management of different types of kidney transplantation and to innovative approaches to drug discovery.
Ischemia-reperfusion injury
The mechanism of occurrence of DGF involves many more complex pathological links as well as pathogenic factors, but to date is not quite clear. Studies have shown that a series of reactions activated by local renal hemorrhage and oxygen deficiency caused by ischemia-reperfusion injury (IRI) are important causes of DGF. The main mechanisms of the disease include: stress oxidation reaction [Citation25], the death of upper skin cells in kidney tubes [Citation26], immune response [Citation27] and so on. IRI has a significant impact on transplant plant damage. Unfortunately, despite decades of research and numerous successful interventions in pre-clinical settings, IRI remains a major problem in clinical practice [Citation28]. Renal ischemia injury in deceased donors was more frequent and more severe than in living donors. The quality and function of the donor kidney is crucial. Inadequate adaptation, acute rejection may cause poor long-term prognosis for DGF patients. Currently commonly used renal protection measures are: Hypothermic Machine Perfusion (0–8◦C) Citation1], Normothermic Machinery Perfusion (35–38◦C) [Citation29], Subnormothermic machine Perfusion (20–34◦C) [Citation30], and Controlled Oxygenated Rewarming [Citation31].
Ponticelli, C [Citation32] assessed many different prevention strategies for DGF and found that only prolonged dopamine infusions and low-temperature machine infusion proved useful. Several drugs with anti-oxidant potentials have been proposed for the treatment of IRIs, including targeting the nuclear factor erythroid 2–related factor 2 (Nrf2) [Citation33], hydrogen sulfide (H2S) [Citation34], and mitochondria-targeting antioxidants [Citation35], but these drugs are mostly based on animal experiments, with few clinical practical studies in kidney transplantation. Traditional antioxidants (such as glutathione, vitamin E, panfenol) are unsatisfactory in preventing oxidative damage to fibroblasts. To date, different antioxidants targeted at granulomas have been studied, the most outstanding of which are MitoQ [Citation25] and SS-3 1[Citation25]. The rapid development of machine infusion technology and the wide range of rehabilitation therapies being explored provide prospects. However, further expansive experimental evidence is required to transform these techniques into clinical practice. This demands an In-depth understanding of the I/R damage that occurs during kidney transplantation, as well as a reasonable pre-clinical model for evaluating treatment strategies. So far, the best strategy for protecting the transplanted kidney has not been determined.
Tacrolimus
Tacrolimus is a macrolide drug with powerful immunosuppressive properties, which can effectively prevent transplant rejection of kidney, heart, lung, intestine and bone marrow, and was approved in the United States in April 1994 to prevent organ rejection after liver transplantation [Citation36].
Studies have shown that the clearance rate of tacrolimus in kidney transplant recipients decreases with the extension of time after transplantation, while the clinical manifestations of some recipients are contrary [Citation37], which may have an impact on the dosage and administration time of tacrolimus. Sandesh Parajuli [Citation38] found that in DGF subjects, there was no difference in DGF recovery time or number of dialysis between the immediate release of tacrolimus (tacrolimus) and the delayed release of tacrolimus. The primary challenge after kidney transplantation is how to maintain the balance between the effectiveness and toxicity of the immunosuppressive drugs used, and due to the lack of newer immunosuppressive drugs, further understanding of existing drugs may be the key to improving efficacy. Tacrolimus is a representative calcium immunosuppressive drug, and its blood concentration is affected by many factors such as population biological factors, transplantation time and type of transplanted organ, gene polymorphism, combined drug use, food and drug dosage form, etc. Clinical consideration should be given to its influencing factors to achieve individual medication for transplant patients.
Older donors and recipients
Due to the aging of the population, more and more elderly people are undergoing kidney transplantation, and the number of kidney transplant donors who were over the age of 70 is also increasing significantly, reaching twenty percent of the kidney transplants in some European countries [Citation39]. Many scholars have gradually begun to pay attention to this aspect of research. However, there are some controversies regarding the research results on this issue. Jeong-Hoon Lim [Citation40] indicated in his study that the incidence of DGF, graft survival rate and acute rejection free survival rate in elderly recipients (>60 years old) were not different from those in young recipients. However, older recipients have an increased risk of mortality, especially from infection-related deaths. The increased mortality among older recipients is predominantly due to infection, although the overall infection rate is not higher than in younger recipients, many cases of severe/critical infection lead to death in older patients, suggesting that older recipients are vulnerable to immunosuppressive therapy and need tailored immunosuppression. The results of Alberto Artiles [Citation41] were different, and he found that kidney transplantation in elderly recipients (>70 years old) had poorer patient and graft survival in the long term compared to the younger population. Benoit Mesnard [Citation42] conducted a systematic review of transplant outcomes from elderly donors (>70 years old), which showed that elderly donors were reliable sources of grafts. However, this type of transplantation was related to a high ratio of delayed graft function without an effect on long-term graft survival. However, the current research on this aspect is not deep enough, and a large number of clinical studies are still needed to find a better way to improve the prognosis of elderly recipients, control the occurrence of postoperative infection and complications, and improve the survival rate and time of elderly recipients.
Artificial Intelligence and DGF
Machine learning (ML) has steadily gained clinical attention due to the quick growth of artificial intelligence technologies and Internet-based medical care [Citation43]. Therefore, we will also discuss it as one of the hotspots. The foundation of artificial intelligence is machine learning, which describes how computers analyze and learn from data sets with known characteristics and results to identify possible relationships in the data and express them mathematically [Citation44]. Numerous academics have utilized artificial intelligence and machine learning in kidney transplantation as a means of personalizing patient diagnosis and care through the use of these technologies as clinical decision support tools.
Four categories of machine learning models exist: semi-supervised, supervised, unsupervised, and reinforcement learning [Citation45]. While the linkage of input data with markers using mathematical functions is primarily employed for classification or regression, supervised learning refers to the manual labeling of specific markers during model learning. Konieczny, Andrzej conducted data mining and machine learning experiments in a single center study to construct a prediction model of DGF with good accuracy [Citation46].
Unsupervised learning is a machine learning technique that primarily addresses clustering and dimensionality reduction issues. It can also be used to extract structure and valuable information from unlabeled high-dimensional data. Unsupervised learning can solve issues like anomaly detection and class clustering because it doesn’t require a lot of tags, lowers the cost of data labeling, and automatically finds the structure and pattern of the data. Caroline C. Jadlowiec [Citation47] examined the connection between DGF and kidney transplant outcomes using an unsupervised machine learning technique, and they discovered four unique clusters: There was no difference in the graft one-year survival between groups amongst young patients with high PRA retransplantation, older diabetic patients, young Black individuals without diabetes, and middle-aged adults with high blood pressure or diabetes.
Semi-supervised learning is a kind of learning method which combines supervised learning and unsupervised learning, and it is also a key problem in the field of pattern recognition and machine learning [Citation48]. Reinforcement learning is a machine learning method used to train intelligent agents with the goal of making the agent learn to perform the correct actions in a given environment to maximize the expected cumulative reward [Citation49]. semi-supervised learning and intensive learning are not applied in the DGF at this time.
Research on the assessment and use of DGF in machine learning is still scarce. There is no ideal DGF model; the majority of the models mentioned above are based on single-center data and are primarily at the experimental research stage. In order to offer a solid foundation for early clinical detection and screening of DGF, future research can concentrate on more rigorous study designs, integrate rich big data from multi-centers, construct more machine learning models of DGF, and carry out efficient verification.
Limitations
In this study, we visually analyzed WOSCC databases using CiteSpace and VOSviewer software, which might have overlooked published papers in other databases. In addition, we only considered English-language articles because we are unable to evaluate articles in other languages simultaneously owing to technological restrictions. Furthermore, none of these software tools can gather data on all the writers of a piece, merge the names of the same authors across formats, or assess the veracity of studies conducted in the area. In this study, we only analyzed the research trends and hotspots of DGF in the past 10 years, we will continue to conduct this study in the future after this report and track its changes over time.
Conclusion
This study uses CiteSpace and VOSviewer software based on WOSCC to analyze the current state and focal points of DGF research in the past ten years. The analysis provides an intuitive overview of DGF research during this period, revealing publication patterns across different countries, journal citation counts, potential collaborative authors and institutions, as well as discussing keywords. These emerging trends include ‘brain death donors’, ‘blood absence re-injection injuries’, ‘tacrolimus’, and ‘older donors and recipients’, and ‘Artificial Intelligence and DGF’. The findings aim to offer valuable insights and directions for future DGF research and application.
Acknowledgments
I want to express my sincere gratitude to all of my friends and esteemed teachers who have shown me kind support. Above all, I am especially indebted to my esteemed thesis advisor, Professor Zhen Li, for her kind counsel, perceptive recommendations, and unwavering support during the whole writing process. Without his motivational guidance and insightful recommendations, I would be struggling to finish this thesis on my own. Furthermore, I would like to thank Mingqian Kuang for helping me to finish the data gathering and preliminary data analysis. Each author reviewed and approved the submitted version of the work in addition to contributing to its revision. In addition, we thank the Yunnan Province Technical Innovation Talent Training Target Project #1 (grant numbers: 202305AD160007) and the Yunnan Provincial Department of Science and Technology Kunming Medical University Applied Basic Research Joint Special Fund Project #2 (grant numbers: 202201AY070001-080) for their support of this work.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
References
- Sweet AL, Connelly CR, Dewey EN, et al. The effect of perfusate temperature on delayed graft function in deceased donor renal transplantation. Prog Transplant. 2023; 33(4):1–19. Epub 2023 Nov 14. PMID: 37964564. doi: 10.1177/15269248231212920.
- Zhao S. Establishment of a prediction model for delayed recovery of renal transplantation function at AIL-2 level. Nanchang university, 2022. doi: 10.27232/d.cnki.gnchu.2022.001085.
- Li J. Meta-analysis of risk factors for delayed renal function recovery after DD donor kidney transplantation. University of South China, 2021. doi: 10.27234/d.cnki.gnhuu.2021.000971.
- Mezzolla V, Pontrelli P, Fiorentino M, et al. Emerging biomarkers of delayed graft function in kidney transplantation. Transplant Rev. 2021;35(4):100629. Epub 2021 May 26. PMID: 34118742. doi: 10.1016/j.trre.2021.100629.
- Yuan S, Jing L, Jiamiao H, et al. Visual analysis of research progress of photodynamic sterilization technology based on citespace. Sci Technol Food Industry. 2024:1–30. doi: 10.13386/j.issn1002-0306.2023010162.
- Yan F, Zhangkai WANG, Xuewei Y, et al. Bibliometric analysis of antibiotic resistance genes in soil media based on web of science. Ecotoxicol. 2023;18(6):302–313. http://kns.cnki.net/kcms/detail/11.5470.X.20230908.1646.002.html.
- Summers DM, Watson CJ, Pettigrew GJ, et al. Kidney donation after circulatory death (DCD): state of the art. Kidney Int. 2015;88(2):241–249. Epub 2015 Mar 18. PMID: 25786101. doi: 10.1038/ki.2015.88.
- Siedlecki A, Irish W, Brennan DC. Delayed graft function in the kidney transplant. Am J Transplant. 2011; 11(11):2279–2296. Epub 2011 Sep 19. PMID: 21929642; PMCID: PMC3280444. doi: 10.1111/j.1600-6143.2011.03754.x.
- Yarlagadda SG, Coca SG, Formica RN, Jr, et al. Association between delayed graft function and allograft and patient survival: a systematic review and meta-analysis. Nephrol Dial Transplant. 2009;24(3):1039–1047. Epub 2008 Dec 22. PMID: 19103734. doi: 10.1093/ndt/gfn667.
- Irish WD, Ilsley JN, Schnitzler MA, et al. A risk prediction model for delayed graft function in the current era of deceased donor renal transplantation. Am J Transplant. 2010;10(10):2279–2286. PMID: 20883559. doi: 10.1111/j.1600-6143.2010.03179.x.
- Man Z, Chunfeng CAI. Visual analysis of research hotspots related to nursing deficiencies based on CiteSpace. J PLA Nurs. 2021;38(08):46–49.
- Furong C, Xianghua X. Visualization analysis of hot spots and trends of domestic and foreign breathing services based on CiteSpace. Evid-Based Nurs. 2019;9(17):3131–3137.
- Fang Y, Outmani L, de Joode AAE, et al. Bariatric surgery before and after kidney transplant: a propensity score-matched analysis. Surg Obes Relat Dis. 2023;19(5):501–509. Epub 2022 Nov 21. PMID: 36572583. doi: 10.1016/j.soard.2022.11.010.
- Debout A, Foucher Y, Trébern-Launay K, et al. Each additional hour of cold ischemia time significantly increases the risk of graft failure and mortality following renal transplantation. Kidney Int. 2015;87(2):343–349. Epub 2014 Sep 17. PMID: 25229341. doi: 10.1038/ki.2014.304.
- Yumeng K, Xiaoxuan L, Pengying Z, et al. Visual analysis of hot spots in houttuynia research based on bibliometrics. Chin Herb Med. 2023;;54(20):6777–6790.
- Jiang WSH. Looking at the world from Asia: sustainable and healthy development of brain death organ donation from an Asian perspective. J Organ Transplant. 2019;13(06):697–710.
- Huaman MA, Vilchez V, Mei X, et al. Donor positive blood culture is associated with delayed graft function in kidney transplant recipients: a propensity score analysis of the UNOS database. Clin Transplant. 2016;30(4):415–420. Epub 2016 Feb 24. PMID: 26840885. doi: 10.1111/ctr.12703.
- Kowalski K, Marschollek J, Nowakowska-Kotas M, et al. The impact of clinical education on knowledge and attitudes towards brain death among polish medical students - a cross-sectional study. BMC Med Educ. 2023;23(1):669. PMID: 37710211; PMCID: PMC10503106. doi: 10.1186/s12909-023-04637-y.
- Li MT, Ramakrishnan A, Yu M, et al. Effects of delayed graft function on transplant outcomes: a meta-analysis. Transplant Direct. 2023;9(2):e1433. PMID: 36700066; PMCID: PMC9835896. doi: 10.1097/TXD.0000000000001433.
- Müller AK, Breuer E, Hübel K, et al. Long-term outcomes of transplant kidneys donated after circulatory death. Nephrol Dial Transplant. 2022;37(6):1181–1187. PMID: 34919732. doi: 10.1093/ndt/gfab358.
- Bell R, Farid S, Pandanaboyana S, et al. The evolution of donation after circulatory death renal transplantation: a decade of experience. Nephrol Dial Transplant. 2019;34(10):1788–1798. PMID: 29955846. doi: 10.1093/ndt/gfy160.
- Molina M, Guerrero-Ramos F, Fernández-Ruiz M, et al. Kidney transplant from uncontrolled donation after circulatory death donors maintained by nECMO has long-term outcomes comparable to standard criteria donation after brain death. Am J Transplant. 2019;19(2):434–447. Epub 2018 Aug 16. PMID: 29947163. doi: 10.1111/ajt.14991.
- de Kok MJ, McGuinness D, Shiels PG, et al. The neglectable impact of delayed graft function on long-term graft survival in kidneys donated after circulatory death associates with superior organ resilience. Ann Surg. 2019;270(5):877–883. PMID: 31567503. doi: 10.1097/SLA.0000000000003515.
- Saat TC, Susa D, Roest HP, et al. A comparison of inflammatory, cytoprotective and injury gene expression profiles in kidneys from brain death and cardiac death donors. Transplantation. 2014;98(1):15–21. PMID: 24901651. doi: 10.1097/TP.0000000000000136.
- Granata S, Votrico V, Spadaccino F, et al. Oxidative stress and ischemia/reperfusion injury in kidney transplantation: focus on ferroptosis, mitophagy and new antioxidants. Antioxidants. 2022;11(4):769. doi: 10.3390/antiox11040769.
- Jie X, Wang L, Yang Y, et al. Research progress of autophagy in renal ischemia-reperfusion injury. Medical Review. 2019;27(04):658–663.
- Yu H, Kalogeris T, Korthuis RJ. Reactive species-induced microvascular dysfunction in ischemia/reperfusion. Free Radic Biol Med. 2019;135:182–197. doi: 10.1016/j.freeradbiomed.2019.02.031.
- Akalay S, Hosgood SA. How to best protect kidneys for transplantation-mechanistic target. J Clin Med. 2023;12(5):1787. PMID: 36902572; PMCID: PMC10003664. doi: 10.3390/jcm12051787.
- Mazilescu LI, Urbanellis P, Kim SJ, et al. Normothermic ex vivo kidney perfusion for human kidney transplantation: first North American results. Transplantation. 2022;106(9):1852–1859. doi: 10.1097/TP.0000000000004098.
- Guo Z, Luo T, Mo R, et al. Ischemia-free organ transplantation - a review. Curr Opin Organ Transplant. 2022;27(4):300–304. Epub 2022 Jul 5. PMID: 36354255. doi: 10.1097/MOT.0000000000000998.
- Schopp I, Reissberg E, Lüer B, et al. Controlled rewarming after hypothermia: adding a new principle to renal preservation. Clin Transl Sci. 2015;8(5):475–478. Epub 2015 Jun 5. PMID: 26053383; PMCID: PMC4744687. doi: 10.1111/cts.12295.
- Ponticelli C, Reggiani F, Moroni G. Delayed graft function in kidney transplant: risk factors, consequences and prevention strategies. J Pers Med. 2022; 12(10):1557. PMID: 36294695; PMCID: PMC9605016. doi: 10.3390/jpm12101557.
- Tong F, Zhou X. The Nrf2/HO-1 pathway mediates the antagonist effect of L-Arginine on renal ischemia/reperfusion injury in rats. Kidney Blood Press Res. 2017;42(3):519–529. Epub 2017 Aug 30. PMID: 28854440. doi: 10.1159/000480362.
- Corsello T, Komaravelli N, Casola A. Role of hydrogen sulfide in NRF2- and sirtuin-dependent maintenance of cellular redox balance. Antioxidants. 2018; 7(10):129. PMID: 30274149; PMCID: PMC6210431. doi: 10.3390/antiox7100129.
- Wyss JC, Kumar R, Mikulic J, et al. Differential effects of the mitochondria-active tetrapeptide SS-31 (D-Arg-dimethylTyr-Lys-Phe-NH2) and its peptidase-targeted prodrugs in experimental acute kidney injury. Front Pharmacol. 2019;10:1209. PMID: 31780923; PMCID: PMC6857474. doi: 10.3389/fphar.2019.01209.
- Bentata Y. Tacrolimus: 20 years of use in adult kidney transplantation. What we should know about its nephrotoxicity. Artif Organs. 2020;44(2):140–152. Epub 2019 Sep 4. PMID: 31386765. doi: 10.1111/aor.13551.
- Zhu W, Xue L, Peng H, et al. Tacrolimus population pharmacokinetic models according to CYP3A5/CYP3A4/POR genotypes in Chinese Han renal transplant patients. Pharmacogenomics. 2018;19(13):1013–1025. Epub 2018 Jul 24. PMID: 30040022. doi: 10.2217/pgs-2017-0139.
- Parajuli S, Muth B, Bloom M, et al. A randomized controlled trial of envarsus versus immediate release tacrolimus in kidney transplant recipients with delayed graft function. Transplant Proc. 2023;55(7):1568–1574. Epub 2023 Jul 1. PMID: 37394382. doi: 10.1016/j.transproceed.2023.05.025.
- Agence de la biomédecine. Le rapport médical et scientifique de l’Agence de la biomédecine 2017 [online]. 2017 Cited the 30 June 2022]. Available on: https://www.agence-biomedecine.fr/annexes/bilan2017/donnees/organes/06-rein/synthese.htm.
- Lim JH, Lee GY, Jeon Y, et al. Elderly kidney transplant recipients have favorable outcomes but increased infection-related mortality. Kidney Res Clin Pract. 2022;41(3):372–383. Epub 2022 Feb 23. PMID: 35286795; PMCID: PMC9184840. doi: 10.23876/j.krcp.21.207.
- Artiles A, Domínguez A, Subiela JD,., et al. Kidney transplant outcomes in elderly population: a systematic review and meta-analysis. Eur Urol Open Sci. 2023;51:13–25. PMID: 37006961; PMCID: PMC10064232. doi: 10.1016/j.euros.2023.02.011.
- Mesnard B, Territo A, Campi R,., et al. Kidney transplantation from elderly donors (> 70 years): a systematic review. World J Urol. 2023;41(3):695–707. Epub 2023 Mar 13. PMID: 36907943 doi: 10.1007/s00345-023-04311-4.
- Kampaktsis PN, Tzani A, Doulamis IP, et al. State-of-the-art machine learning algorithms for the prediction of outcomes after contemporary heart transplantation: results from the UNOS database. Clin Transplant. 2021;35(8):e14388. Epub 2021 Jun 29. PMID: 34155697. doi: 10.1111/ctr.14388.
- Hosseini MP, Hosseini A, Ahi K. A review on machine learning for EEG signal processing in bioengineering. IEEE Rev Biomed Eng. 2021;14:204–218. Epub 2021 Jan 22. PMID: 32011262. doi: 10.1109/RBME.2020.2969915.
- Ben Ali W, Pesaranghader A, Avram R, et al. Implementing machine learning in interventional cardiology: the benefits are worth the trouble. Front Cardiovasc Med. 2021; 8:711401. PMID: 34957230; PMCID: PMC8692711. doi: 10.3389/fcvm.2021.711401.
- Konieczny A, Stojanowski J, Rydzyńska K, et al. Artificial intelligence-a tool for risk assessment of delayed-graft function in kidney transplant. J Clin Med. 2021;10(22):5244. PMID: 34830526; PMCID: PMC8618905. doi: 10.3390/jcm10225244.
- Jadlowiec CC, Thongprayoon C, Leeaphorn N, et al. Use of machine learning consensus clustering to identify distinct subtypes of kidney transplant recipients with DGF and associated outcomes. Transpl Int. 2022;35:10810. PMID: 36568137; PMCID: PMC9773391. doi: 10.3389/ti.2022.10810.
- Eckardt JN, Bornhäuser M, Wendt K, et al. Semi-supervised learning in cancer diagnostics. Front Oncol. 2022;12:960984. PMID: 35912249; PMCID: PMC9329803. doi: 10.3389/fonc.2022.960984.
- Gupta R, Srivastava D, Sahu M, et al. Artificial intelligence to deep learning: machine intelligence approach for drug discovery. Mol Divers. 2021;25(3):1315–1360. Epub 2021 Apr 12. PMID: 33844136; PMCID: PMC8040371. doi: 10.1007/s11030-021-10217-3.