Publication Trends and Research Hotspots of the Cardiorenal Syndrome: A Bibliometrics and Visual Analysis from 2003 to 2023

Cardiorenal syndrome encompasses a range of disorders involving both the heart and kidneys, wherein dysfunction in one organ may induce dysfunction in the other, either acutely or chronically [1]. This syndrome represents a convergence of interactions between the heart and kidneys across several interfaces. Patients burdened with concurrent heart and kidney diseases continue to experience high rates of hospitalisation, symptom severity and mortality [1‒4]. The scientific statement issued by the American Heart Association underscores the critical need for guidelines, best clinical practice models and research funding geared specifically toward improving outcomes in cardiorenal medicine across both specialities, with a focus on future therapeutic needs [1]. Therefore, gaining a comprehensive understanding of the developmental trajectory and research hotspots in the field of cardiorenal syndrome research is imperative.

Bibliometrics is the science of using mathematical and statistical methods to analyse the distribution and patterns of books, articles and other literature. Bibliometric analysis provides a quantitative approach for reviewing and investigating existing literature in a given field [5]. Visualised co-citation analysis in bibliometrics contributes to enhanced data, rendering results more comprehensive. Additionally, bibliometric analysis provides insights into the evolution of a field [6]. The CiteSpace and VOSviewer softwares have been commonly used as visualisation and analysis tools in bibliometrics. CiteSpace employs set theory-based data normalisation for similarity measurement among knowledge units, utilising similarity algorithms to create both time zone and timeline views that elucidate the evolution of knowledge and the historical context of literature clusters, thereby providing insight into the developmental trajectory and trends within the field [7]. VOSviewer adopts a probabilistic-based data standardisation method and offers diverse visualisation views encompassing keywords, co-organisations and co-authors, including network, overlay and density visualisations. Importantly, it is well-known for its ease of use and generation of visually appealing representations [8].

This study conducted an in-depth analysis and review of research over the past 2 decades in the field of cardiorenal syndrome using bibliometrics. Additionally, the BioBERT biomedical domain-specific language representation model [9] was used to extract and statistically analyse keywords related to genes and diseases found in research articles within the cardiorenal syndrome field. The full term of BioBERT is Bidirectional Encoder Representations from Transformers for Biomedical Text Mining which, which is a domain-specific language representation model that has been pre-trained on large-scale biomedical corpora. The objective of this study was to quantitatively illustrate the developmental trajectory, research hotspots and evolutionary trends in the field of cardiorenal syndrome research and comprehensively map and analyse its development through knowledge mapping.

Data was obtained from the Web of Science Core Collection, which covers a wide range of publications from various disciplines. In particular, the editions “Science Citation Index Expanded” and “Social Sciences Citation Index” were used. The search strategy involved the following search terms: (((TS=(cardiorenal syndrome)) OR TS=(cardio-renal syndrome)) OR TS=(renocardiac syndrome)) OR TS=(reno-cardiac syndrome). The search period from January 1, 2003, to September 8, 2023. Data sources and search strategies are detailed in Table 1.

A total of 2,405 literature studies were retrieved. Based on the inclusion and exclusion criteria, 538 publications categorised as non-articles and reviews, such as meeting abstracts, editorials, and letters, were excluded. Additionally, 84 non-English publications and two duplicate entries were excluded by reading the titles, abstracts, and full texts of the articles. Ultimately, 1,781 publications were included in the analysis (Figure 1).

Two independent authors conducted a comprehensive review of the entire text based on the inclusion and exclusion criteria. Any disagreements between them were resolved through discussion, and the selected documents were exported in plain text format. The primary factors analysed were countries/regions, authors, citations and references. Key analysis methods included co-authorship, co-citation and co-occurrence. Microsoft Excel 2019 was employed for generating flowcharts and statistical tables. The BioBERT biomedical domain-specific language representation model was used to extract and statistically analyse the gene or disease entity words in the included studies’ abstracts. VOSviewer (ver 1.6.19) was used for visual analyses of the number of publications, authors, keywords, gene clustering and disease clustering. CiteSpace (ver 6.2.R4) was used for co-occurrence analysis and keywords’ timeline generation. Finally, for the analysis and visualisation of country and regional contributions and collaborative efforts, SCImago Graphica (ver 1.0.34) was employed.

The analysis of publications related to cardiorenal syndrome in the Web of Science Core Collection revealed a consistent upward trend over time (Fig. 2). From 2008, there was a significant increase in the number of annual publications in this field, with a notable surge from 2009 to 2012, witnessing a year-on-year increase from 29 articles to 111 articles. However, the growth rate of annual publications fluctuated between 2013 and 2019, with some years experiencing a decline. Nevertheless, the number of annual publications resumed its upward trajectory after 2019, peaking at 180 annual publications in 2022.

The geographical distribution of studies related to cardiorenal syndrome is illustrated in Figure 3. Figure 3a shows that the research activity in this field is concentrated in East Asia, North America, and Western Europe. Notably, the top five countries, ranked by publication volume, are the USA with 473 articles, followed by China (213), Italy (199), Japan (140), and Germany (80). Figure 3b illustrates the extensive international collaboration of the USA (centrality of 0.40) with various countries, including Italy (centrality of 0.18), Germany, and the UK.

The top five authors who have contributed significantly to the field of cardiorenal syndrome in the last 20 years are shown in Table 2. Claudio Ronco ranks first with 122 publications and also boasts the highest citation number (2,355) and impact factor (H-index = 36). The second and third highest numbers of publications belong to Peter A. McCullough (38) and James R. Sowers (34), with H-indexes of 20 and 24, respectively. Additionally, the second, third and fourth most cited authors are Mikko Haapio (1,069 citations), Rinaldo Bellomo (1,032 citations) and Andrew A. House (1,012 citations), respectively. Figure 4 visualises the distribution of authors and their collaborations, revealing that authors related to cardiorenal syndrome can be divided into nine clusters, with Claudio Ronco’s authors being the most interconnected.

Figure 5 provides insights into the co-citation frequency and citation impact of literature. The article by Rangaswami et al. [1] holds the highest citation frequency (67.07) and has been cited by a total of 479. This article, which is a scientific statement issued by the American Heart Association, focuses on the classification, pathophysiology, diagnosis and treatment strategies of cardiorenal syndrome. Following closely, the review article by Ronco et al. [10] holds the next highest citation frequency (51.11), with a total of 1,389 citations. Although the review article by Chan et al. [11] had the highest co-citation frequency, its citation frequency in the field of cardiorenal syndrome was not significant.

Table 3 highlights the top 10 keywords with the highest frequency in research papers related to cardiorenal syndrome. Notably, keywords such as cardiorenal syndrome, chronic kidney disease, heart failure, and mortality appeared in ≥5% of the papers. Co-occurrence analysis of keywords revealed a shift in research trend from topics like chronic renal failure, erythropoietin, risk factor, subcutaneous erythropoietin, and congestive heart failure to heart failure, injury, expression, and pathophysiology (Fig. 6a). Furthermore, Figure 6b presents a timeline analysis after clustering the keywords associated with cardiorenal syndrome, revealing five clusters, namely metabolic syndrome, acute heart failure, machine learning, anaemia, and heart failure. Notably, the clusters of metabolic syndrome and anaemia have persisted over the last 20 years, while clusters such as acute heart failure and machine learning emerged more recently. Furthermore, the research interest in the heart failure cluster shows signs of declining after 2021.

The BioBERT biomedical domain-specific language representation model was used to identify 1,831 genes from 1,781 articles, with a minimum occurrence threshold for each gene set at 10. Co-occurrence clustering analysis was performed using VOSviewer to generate a visualisation map (Fig. 7a). The results revealed that cardiorenal syndrome-related genes were clustered within domains such as inflammation and immune (with IL6 being the most prominent), blood pressure regulation, and electrolyte balance (with REN as the predominant gene) and synthesis and metabolism (with INS as the most prevalent gene).

Additionally, a total of 1,248 diseases were obtained from the 1,781 articles, employing a minimum occurrence threshold of 12 for each disease. Co-occurrence clustering analysis using VOSviewer revealed that cardiorenal syndrome-related genes were clustered in the domains of organ dysfunction (with cardiorenal syndrome being the most prevalent), acute organ injury (with acute kidney injury as the dominant disease), organ remodelling (with cardiomegaly as the most prevalent condition), and chronic organ injury (with chronic kidney failure as the dominant disease) (Fig. 7b).

Cardiorenal syndrome represents a complex clinical scenario where dysfunction in either the heart or the kidney primarily impacts and subsequently influences the other organ [12]. Bibliometrics allows for the analysis of authors, institutions, countries (regions), and references in a literature database, facilitating a comprehensive understanding of a research area through tools such as CiteSpace and VOSviewer. This methodology offers a more comprehensive analysis of the literature and presents results in a visually intuitive manner, surpassing the scope of a traditional systematic review. In this study, bibliometrics was employed to explore the development of the cardiorenal syndrome field from 2003 to 2023 and to project potential future research trends.

The nuanced and highly interdependent relationship between the kidney and the heart was first described by Robert Bright [13] in 1836, who outlined the significant cardiac structural changes in patients with advanced kidney disease. Over time, significant progress has been made in elucidating the cardiorenal connection, encompassing haemodynamic phenotypes, pathophysiological mechanisms, therapeutic interventions, and clinical outcomes [1]. In the present study, the annual publication trend analysis confirms a consistent annual increase in the volume of literature related to cardiorenal syndrome over the past 2 decades. This trend suggests that the area of research has significant growth potential and is emerging as a research hotspot.

Keywords serve as concise summaries of a study’s aim, objective, and methodology. Thus, the analysis of keywords allows for the identification of thematic evolutions and research hotspots in a particular field of study over time. In this study, keyword clustering and timeline analysis revealed that current research in cardiorenal syndrome is focused on heart failure, injury, expression, and pathophysiology. The timeline analysis, following keyword clustering, identified five distinct clusters of keywords: metabolic syndrome, acute heart failure, machine learning, anaemia, and heart failure. Among them, the anaemia cluster has maintained relevance over the last 20 years and is likely to remain a key research area in the field of cardiorenal syndrome. The occurrence of this phenomenon is likely due to the pervasive presence of anaemia among patients with cardiorenal syndrome, and its substantial influence on the progression of the disease and the development of treatment strategies. Anaemia is a significant complication of chronic kidney disease and a common co-occurrence in patients with heart failure. It substantially raises the risk of hospitalisation and mortality in heart failure patients and contributes to the progression of kidney disease, increasing the likelihood of renal replacement therapy. Consequently, the effective management of anaemia is a crucial component of the treatment for cardiorenal syndrome.

Furthermore, the research fervour within the heart failure cluster declined after 2021, while the acute heart failure cluster gradually gained prominence, suggesting a shift toward granular research directions. These findings demonstrate the complex and multi-faceted nature of the overlap between cardiovascular and kidney diseases, spanning several interfaces. These include the haemodynamic interactions of the heart and kidney in heart failure [4, 14], the biochemical perturbations across the anaemia-inflammation-bone mineral axis in chronic kidney disease [12, 15], and structural changes in the heart unique to kidney disease progression [16].

The persistence of the metabolic syndrome cluster over the last 20 years aligns with reviews by Sowers et al. [17]. They explored the potential mechanisms through which obesity and other metabolic abnormalities contribute to heart and progressive kidney disease. Cardiovascular-kidney-metabolic syndrome reflects the interplay between metabolic risk factors, chronic kidney disease, and the cardiovascular system. The present study infers that the study of metabolic syndrome in cardiorenal syndrome has become a hot topic because it is not only closely related to the pathogenesis of cardiac and renal diseases but also involves the development of disease prevention and therapeutic strategies, which is important for improving the therapeutic efficacy of cardiorenal syndrome and improving the prognosis of patients.

The relatively recent emergence of the machine learning cluster suggests that, in the future, the application of methods like big data mining may become a prominent research hotspot within the cardiorenal syndrome field. Machine learning is a burgeoning field of medicine, particularly as the scale and complexity of biological data continue to grow, allowing for the development of informative and predictive models for underlying biological processes [18‒20]. Notable examples include studies by Urban et al. [21], who used a machine-learning approach to investigate worsening renal function in acute heart failure, and Tasic et al. [22], who used utility classifiers of machine learning to recognise cardiorenal syndrome and guide treatment planning. Machine learning algorithms are capable of analysing large amounts of complex medical data, including patients’ clinical characteristics, biomarkers, gene expression, etc., to predict disease risks and trends. This predictive capability helps in early identification of patients with cardiorenal syndrome, leading to timely interventions. At the same time, by analysing patient-specific data, machine learning models can help doctors develop more personalised treatment plans to improve the relevance and effectiveness of treatment.

Cluster analysis of associated genes revealed IL6, REN, and INS, which are protein-coding genes, as current research hotspots in the field of cardiorenal syndrome. Additionally, cluster analysis of associated diseases revealed that acute/chronic organ damage, structural remodelling and dysfunction are the current research hotspots in the field of cardiorenal syndrome. IL6 encodes a cytokine involved in inflammation and B cell maturation. Moreover, studies have demonstrated that inflammation, pro-inflammatory factors, oxidative stress and immune mechanisms can induce apoptosis in cardiorenal syndrome, resulting in organ damage or structural remodelling [23‒27]. REN encodes renin, an aspartic protease belonging to the renin-angiotensin-aldosterone system, involved in regulating blood pressure and electrolyte balance. INS encodes insulin, a peptide hormone with a vital role in carbohydrate and lipid metabolism. Insulin exerts numerous effects on the human body, including the cardiac tissue [28]. Moreover, insulin resistance is a common feature in conditions such as cardiorenal syndrome, hypertension, obesity, and diabetes [29, 30].

Between centrality refers to the role of a node in mediating the transfer of information between other nodes. In CiteSpace, nodes are considered as research subjects, representing authors, disciplines, and institutions. The higher the centrality of a research subject’s intermediation, the stronger the link with other research subjects and the greater the collaborative opportunities. The H-index, proposed by Hirsch in 2005, is a hybrid quantitative metric that can be used to assess the amount of scholarly output versus the level of scholarly output of a researcher [31]. In this study, centrality analysis of countries (regions) revealed that the USA is the highest contributor to research on cardiorenal syndrome, with the highest number of publications (473 publications) and extensive collaborations with other countries such as Italy, Germany and the UK (centrality of 0.40). Moreover, the co-citation analysis of literature identified a scientific statement [1] from the American Heart Association as having the highest citation frequency, highlighting the US’s significant academic influence in the field of cardiorenal syndrome. Italy also demonstrated a significant academic impact, ranking third in terms of the number of publication volumes (199) and possessing a centrality of 0.18. Notably, the analysis of study authors identified Claudio Ronco from Italy as the most influential author in this field, with the highest number of publications and citations (H-index = 36).

This research has several inevitable limitations. Firstly, this study only counted papers in the Web of Science Core Collection. Nevertheless, discrepancies between the papers included in different databases and the data associated with each paper may also contribute to the distortion of the results. Secondly, the deadline for this research search was September 8, 2023. So recently published important studies that have not been cited enough may have also been missed. Despite these limitations, this research presents some of the authors who have made significant contributions to the field of cardiorenal syndrome and summarises the development process of cardiorenal syndrome research. Meanwhile, this study used the BioBERT model for the first time to analyse the gene or disease entity in the field of cardiorenal syndrome. This study provides scholars who wish to join the field of cardiorenal syndrome with the direction to retrieve important literature and try to identify possible future research directions of cardiorenal syndrome.

Over the past 2 decades, the volume of literature related to cardiorenal syndrome has shown consistent growth. Currently, the most prominent research areas in this field are heart failure, injury, expression, and pathophysiology. Additionally, emerging trends for future research include anaemia, acute heart failure, metabolic syndrome, and machine learning. Moreover, the interplay of inflammation, oxidative stress and immune mechanisms in cardiorenal syndrome, affecting not only the heart and kidneys but also other organs, warrants further investigation. Currently, in terms of academic influence, the USA and Italy occupy prominent positions. As the global scientific community continues to undergo collaborative efforts, it is anticipated that an increasing number of significant studies will be conducted by multinational researchers working collectively.

BioBERT biomedical speech representation model was completed by Citexs Big Data Analysis Platform (https://www.citexs.com).

Ethical approval and consent were not required as this study was based on publicly available data.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

Yibo Shi and Zean Fu: concept/design, drafting article, data collection, and revision. Shixiong Wu: drafting article and data collection. Xinyi Yu: concept/design, drafting article, data collection, revision and guidance, and approval of article.

The data that support the findings of this study are available from the corresponding author upon reasonable request.