The importance of smoking, a growing, multidisciplinary area, and the respiratory system has increased in recent years, with a clear upward trend in production, impact and co-authorship in both.1–4

Bibliometrics is concerned with the analysis of written communication, and is aimed at the treatment and study of quantitative data from the scientific literature, and its social structure.5

Bibliographical databases are the main source of information used in bibliometric studies. The Web of Science [(WoS) ISI, Thomson Reuters] is an international, multidisciplinary tool, available for access to literature on science, technology, biomedicine and other disciplines. Its best known products, the Science Citation Index-Expanded (SCI-E), Social Science Citation Index (SSCI), and Arts & Humanities Citation Index (A&HCI), contain a large number of source journals. Journal Citation Report (JCR, in its two editions, JCR Science Edition and JCR Social Science Edition) is constructed from these databases; it includes journals from the SCI-E and SSCI as citable journals, and is where the famous journal “impact factor” is published. JCR offers a systematic, objective means to critically evaluate the world's leading journals, based on citation data. Much of the statistical information available for individual journals is also available for subject specialties, according to pooled data within each specialty from 1997; this provides an overview of coverage, performance and lists of citations in a single specialty (http://ip-science.thomsonreuters.com/es/productos/jcr/).

The aim of this study was to analyze the distribution of the production, repercussion and worldwide co-authorship of articles on smoking in the “respiratory system” category through the SCI-E in the period 2001–2010 and their inclusion in the first quartile of the JCR.

The most important finding of our study was that the production on smoking in the respiratory category increased over the period analyzed, with 40% of this production in Q1 of the JCR in prestigious journals, i.e. those with greater impact and importance. There have been numerous bibliometric studies published in recent years on biomedicine,6 the respiratory system3,4,7 and smoking.1,2,8–11 We can summarize the conclusion of all of these in the major growth experienced by science, both overall and if we analyze different and specific subject areas. Smoking is included in almost all areas that have seen an increase in their production in biomedicine and health sciences.12 In the last decade, the regions with a greater contribution to worldwide scientific production have been Western Europe, the USA and Asia, which contribute more than double the medium production areas (Eastern Europe, Latin America, the Middle East and the Pacific Region); African regions contribute proportionally less to worldwide production. It is known that Asia increased its scientific production by 94% between 1997 and 2009 (27.51% of the total); this is greater than the North American production (23.28% of the total) and almost reaches that of Western Europe (28.63% of the total).6 As we have seen in our study, Spain published a total of 85 papers on smoking, in the respiratory category, with this production remaining more or less stable over the years, like those published in the first quartile. Spain, at global level, practically doubled scientific production between 2000 and 2009. The percentage of Spanish scientific production in relation to the production in Western Europe and likewise, the participation of Spanish publications with respect to the world total, have grown steadily in the last decade. It has been shown that the annual average growth in scientific production in Spain in the period 1996–2009 was 6.95%, surpassing the Western European average by 2.58 points and the North American average by 4.62 points.6 The future projection for Spain shows, with respect to overall production, a continuing upward trend until at least 2016, maintaining, at any rate, its current position in the world ranking. Other countries, such as France, Germany and Japan will tend to decrease their production. China was the country with greatest expectations for increasing in their worldwide participation, to the detriment of the USA in particular.6 As previously mentioned, and again demonstrated in this study in the respiratory category, production in smoking shows a clearly upward trend, having doubled the number of total and original articles in the last decade; thus it is not yet in a state of saturation, but rather is escalating in knowledge.1

As we know, the journals in the JCR are classified into what are called quartiles, which represents a measure of position that summarizes the statistical data into meaningful groups. In the respiratory category, we found a similar total number of articles on smoking in Q1 in the EU-27 and in the USA in the period studied, with the evolution in the years studied very similar and with no defined trend. The evolution over time of the percentage of publications in the leading journals (Q1) in the 30 countries with highest values for time series6 has been previously analyzed in worldwide scientific production; it was found that the same countries that have more than 60% of their scientific production in Q1 were in the top positions from 2003 to 2009 (Holland, Denmark, Sweden, Israel, Switzerland, USA and UK). Spain, in overall production in biomedicine, was ranked in 21st position, with 49% of the scientific production published in Q1 journals.6 In our study, the top positions in Q1 for articles on smoking published in the respiratory category were held by (in decreasing order) the USA, UK, Canada, Japan, Italy and Holland. Spain was ranked in 12th position.

In addition to analyzing productivity, bibliometric analysis includes measurements of the relative quality of the results published. Albeit with recognized limitations, this variable has been evaluated until now by counting the citations received by papers published in other scientific documents, which is used to calibrate the impact or visibility of these in the international community.13 One of the most widely used indices for determining quality is the IF, an index which has numerous defenders and many critics who have pointed out its abusive and erroneous use.14,15 Thus, other bibliometric indicators have been proposed to measure the scientific quality of a researcher or group of researchers; one of these is called the citation index, which is defined as the number of citations that an article receives by other authors. Using these data, the mean number of citations per document and percentage of non-citation can be extracted.14 The scientific productivity of a researcher will be considered of better quality the higher their citation index, the higher the mean number of citations per document and the lower the percentage of non-citations, although the analysis period should be established and normalized.14 This normalization is necessary, as not all subjects are cited equally, so tools like the crown indicator have been proposed (crown indicator: number of citations per document without self-citations/mean rate of citations of all articles from a certain scientific field).16 In this study, we used the number of citations and the citations/document index as an indicator of quality. As can be observed in our study, it was the EU-27 that reached a higher citations/document index, both in total production and in the production of articles published in the first quartile. In the evolution over time, the oldest articles are those with a higher number of citations. This is a normal indicator trend that has already been reported previously,1,6,8 since the citations received by publications in a certain period increase over time and opportunities for them to be cited grow, although some differences have been shown between document databases, with citations in Scopus more recent than in the Web of Science.17 It has been shown that scientific collaboration among countries and among institutions increases the number of citations. Thus, a positive correlation has been observed between the number of countries involved in international collaborative articles and the number of citations that these articles received; papers published in English and with international collaboration were associated with a higher number of citations. It is recognized that the gradual growth of international collaboration in biomedicine and life sciences over recent years is an important factor in attracting citations.1,2,6,8

As we can observe in our study, there is a co-authorship network among countries in the respiratory system category for articles on smoking, which is active, well-established and cohesive. In this study, the most productive countries and with greater visibility in research on smoking are no different to those found in other areas of knowledge6: Europe and the USA are the regions of greatest importance, influence and impact, occupying a central position in the network, which indicates that they have collaboration relationships with most countries. On the contrary, countries that make up the co-authorship network of articles included in Q1 is lower, with the USA, UK and Holland remaining centralized, indicating their visibility and importance. It is clear that industrialized countries monopolize most impact journals.18 Various approaches have been used to analyze international collaboration in scientific articles, but none have been able to fully explain its rapid growth. Applying network analysis tools show that the growth of international collaboration may be explained based on the principle of preferential organization of archives (Barabasi-Albert19 model), which indicates that the nodes of a co-authorship network that already have many connections will be those chosen preferentially by new research groups to initiate a scientific collaboration,20,21 i.e. a node rich in connections (collaborations) will increase its influence more rapidly due to its role as leader in that field. It is recognized that more citations are generated the higher the number of authors and institutions in a paper and, therefore, those authors who collaborate more and better manage these relationships obtain better final results.22 It is a challenge therefore to identify researchers with similar interests to formalize contacts and thus extend already existing networks.23,24 Clearly, scientific collaboration improves and accelerates research, increasing the quality25,26 (positive collaboration), and therefore is fundamental for scientific progress and advance, as it enables resources to be shared, gaps filled, and to promote synergies to achieve the necessary knowledge.

The main limitation of this study is that we restricted the search to journals included in the respiratory system category of the JCR and, logically, this area does not include the entire scientific production on smoking or the respiratory system, so the data should be used to determine the situation of smoking in the respiratory system area of the JCR, but not to assess all the scientific activity in these areas. Study limitations can also be found in the choice of database, and those arising from its normalization, so thorough quality control of the data was performed; finally there are limitations related to the time period studied, which means that only articles from publications that were sourced in the SCI at that time are visible.

We can conclude that the production of articles on smoking in the respiratory category increased over the period analyzed, with 40% of this production in journals in the first quartile of the JCR. We found a high citation rate, with older articles having more citations, and the EU-27 having higher impact and repercussion. The co-authorship network between countries in the respiratory category for articles on smoking is active, well-established and cohesive.

The authors have not declared any conflict of interests.