Database unit processing is used in bibliometric studies to visualize scientific activity in a given knowledge domain via scientific or bibliometric maps. These maps are constructed from the analysis of co-occurrence relations, ie, relationships between 2 co-occurring units of analysis.1 Bibliometric indicators of relatedness have been used to build bibliometric networks based on cocitation analysis (to measure intellectual structure) coauthorship analysis (to measure social structure), and coword analysis (to measure knowledge or conceptual structure). The conceptual structure of the field of cardiovascular research has been previously analyzed using predictors of knowledge domains.2
To visualize the conceptual structure and development over time of cardiovascular research in Spain, we analyzed articles published in Revista Española de Cardiología between 1997 and 2006 and 2007 and 2016. The articles were extracted from Science Citation Index Expanded, a scientific and medical index that is available through the Web of Science platform.3 In total, 2197 documents were retrieved. From these, we selected 202 keywords with a frequency of occurrence of 7 or more: 111 from the period 1997 to 2006 and 94 from the period 2007 to 2016. The method employed consisted of the following stages4: selection of keywords used by the authors of the articles, calculation of keyword co-occurrence frequency, clustering analysis, and visualization of results in density or heat maps. The maps were constructed using the VOSviewer software tool.5 In bibliometric density maps, areas with the highest density of co-occurrence relations between keywords are shown by a color close to red, while those with a lower density are shown by a color closer to yellow or green.
The conceptual structure of cardiovascular research between 1997 and 2006 is reflected in the bibliometric map shown in Figure 1. The keywords with the highest impact were related to 3 knowledge domains: cardiomyopathies/ischemic heart disease, heart disease/heart failure and echocardiography, and interventional cardiology. The domains related to the keywords with the lowest impact were electrophysiology/arrhythmias and epidemiology/risk factors and preventive cardiology. There were 2 very dense areas in the center of the map: a) an area related to the knowledge domain of cardiomyopathies (labeled with the keywords myocardial infarction, prognosis, and unstable angina), which was closely connected to the domain of interventional cardiology (labeled with coronary angioplasty, stent, and restenosis) and b) an area related to the domains of heart disease/heart failure and echocardiography (labeled with the keywords heart failure, echocardiography, and surgery).
The bibliometric map corresponding to the second period, 2007 to 2016, is shown in Figure 2. The knowledge domains with the highest keyword impact were heart disease/heart failure and echocardiography, cardiomyopathies/ischemic heart disease, and epidemiology/risk factors and preventive cardiology. The domains with the lowest keyword impact were electrophysiology/arrhythmias and interventional cardiology. Again, there were 2 high-density areas in the map containing the predominant keywords used during this period. The first contained 2 central areas: one related to heart failure (labeled with the keywords heart failure and prognosis) and another related to cardiomyopathies (labeled with myocardial infarction and acute coronary syndrome). The second high-density area contained interconnected keywords associated with cardiovascular risk factors, namely, risk factors, diabetes mellitus, epidemiology, cardiovascular disease, obesity, cardiovascular risk, and metabolic syndrome.
Three distinct blocks reflecting changes in the field of cardiovascular research between the periods 1997 to 2006 and 2007 to 2016 were observed. The first was a block showing a trend toward consolidation featuring keywords indicative of a relatively stable connection between the 2 periods, such as acute myocardial infarction, heart failure, coronary disease, and acute coronary syndrome: echocardiography, electrophysiology, arrhythmias, and atrial fibrillation. These terms were related to the knowledge domains of cardiomyopathies/ischemic heart disease, heart disease/heart failure and echocardiography, and electrophysiology/arrhythmias. The second block showed a decreasing trend for the domain of interventional cardiology, which, based on the keywords used by authors, was more prominent between 1997 and 2006. The third block showed a greater impact for the domain of epidemiology/risk factors and preventive cardiology in the second period. One of the most significant findings shown by the map was the increase in the frequency of occurrence and interconnections between keywords associated with cardiovascular risk factors, such as diabetes mellitus, obesity, and metabolic syndrome. This rise in frequency provides evidence that epidemiology/risk factors and preventive cardiology is an emerging knowledge domain in the field of cardiovascular research.
The main limitation of this study is that our results are based on data from a single journal, Revista Española de Cardiología. Nonetheless, we believe that the large number of articles analyzed provides a sufficiently representative picture of the field of cardiovascular research in Spain.
.
