Scientific publications from Arab world in leading journals of Integrative and Complementary Medicine: a bibliometric analysis

This study was restricted to 591 documents retrieved from ISI Web of Knowledge, bearing Arab countries affiliation and, thus, cannot be generalized to the ICM research activity obtained by other sources such as Google Scholar or Scopus. To our knowledge, this is the first report to analyze the quality and quantity of ICM-based research from the Arab region. bibliometric indicators demonstrated that research output in the field of ICM was low in most Arab countries. The lack of industry-academia partnership in applied health research (including government-academia partnerships), and a general weakness in scientific writing may lead to the lack of scientific research in most Arab regions [14, 29‐31]. The total number of ICM documents found in ISI Web of Knowledge between 1980 and 2013 demonstrated a yearly increase. The annual number of documents published in the period of publication (1980–2013) indicated that research production from Arab World was low in the earliest two decades but showed a noticeable increase in the recent decade. A possible explanation for this change in publication output could be attributed to an increase in the number of specialized ICM journals indexed in ISI [1]. On the other hand, ICM research output has pursued the general growth in scientific research output observed in the recent decade and mainly in recent years [1‐3, 5, 16, 17]. Furthermore, ICM research productivity has followed the biomedical research output in the Arab region especially in the recent decade [8, 11, 29, 30].

In the current study, the research output for each country was different in comparing to others. Our findings demonstrated that there were a few countries, such as KSA, Morocco and, Egypt where their total ICM research output was obviously superior to that in the other remaining Arab countries. Earlier studies showed that KSA and Egypt or Morocco had the highest research productivity among the Arab countries [23, 29, 32]. On the other hand, after adjusting for economic growth and population size, Somalia, Morocco, Palestine, Egypt, and Yemen demonstrated the highest research output. We didn’t find any report concerning ICM in Arab world, thus we are incapable of interpreting this result in the light of other results. On the other hand, some precious studies using the same bibliometric instrument for research assessment have demonstrated similar result [8, 14, 15, 29, 30, 33, 34]. Countries with fast growing economies such as KSA had higher number of ICM research output. Furthermore, population size is another factor related to enhance research output in the field of ICM such as in Egypt. Our results confirmed that research activity depended on economic growth, population size, or overall scientific research output of each country [15, 35].

In the current study, the average citation rate for ICM publications originating from Arab region was 16 citations per document. This result was slightly higher than the average citations of ICM journals [1‐3, 5, 16, 17]. Overall, ICM journals have higher citation numbers in comparing to other categories in ISI Web of Knowledge [23, 36]. This is likely related to several factors. First, even if it is reasonable to argue that ICM is a growing scientific field, it is a highly competitive area in terms of scientific legitimacy [2]. Second, a great majority of the publications in our study are original research articles, and this is supported by previous studies that found that the most cited documents are reviews and original studies [2, 3, 5]. Third, the establishment of ICM as a scientific field is also manifested in increasing numbers of scientific forums that may improve opportunities to get more citations or research funds [2]. In our study, research activity related to ICM started in the 1980s and obviously increased after 2000, while the bulk of the publications (34.2 %) were published from 2011 to 2013. The publications that were published before 2000 were most frequently cited articles and there was a relationship between the number of citations and started publications year [3]. A previous study showed that highly cited publications are usually involving international collaboration and often authored by a large number of scientists [37]. The citation is used as a key indicator of research quality in the study. Highly cited publications positively correlated to the h-index of the institution and country individual author, and to individual author [38‐41]. Furthermore, h-index for ICM publications from Arab countries was 47. This finding was slightly higher than other scientific disciplines [36, 42], and slightly lower once compared to another scientific discipline from Arab countries [23]. One probable explanation for variations in average total citation and h-index between ICM publications of Arab countries and those for other non- ICM publications is the IF of ICM journals in which authors published their works. It is believed that articles published in high IF journals have higher possibility of being cited, whereas this is not definite [15]. The h-index is a simple way of measuring performance, impact, visibility, and quality of research. Newly published documents have a clear disadvantage because of the short publication period and they should not be compared with documents with longer publication period. Another problem is how to deal with multi-authored papers. Furthermore, in citation analysis there is a problem of how to treat self-citations. Finally, one should also be aware that a country’s h-index depends on the citation database that is used [43].

Arab authors collaborated most with authors from Europe region, especially from France and Germany, followed with authors in the Asia-Pacific region, especially from India and Pakistan. This may be because the majority Arab researchers graduated from or trained in these countries. Furthermore, recently, many PhD students from the Arab world pursued their graduate ICM education in Europe and Asia-Pacific region, where the concept of ICM is being emphasized at the research and academic levels. Research collaboration is a significant way to improve quantity and quality of research at the university level [23]. Studies have found that there is a positive correlation between research output, and international or national collaboration at the researcher level [23, 44‐46]. Previous studies have confirmed the significance of international collaboration in the quality of the research, which has a positive effect on citation rates [47‐49]. A more recent study published in the Lancet to improve medical research in the Arab world recommended that an Arab medical research council-inspired from the US National Institutes of Health, the Medical Research Council in the United Kingdome, and French Institute of Health and Medical Research (INSERM) in France-is necessary to establish strategies that promote medical and health research in the Arab world in collaboration with international institutions [31].

Institutions of higher education in Arab region predominated in the most top 20 prolific institutions for research activity in ICM field; this indicates that institutions of higher education were actively researching and interesting in the ICM research. This may be attributed to universities encouragement for academics and researchers to publish their works in journals indexed in ISI database with impact factors [8, 20, 50]. Bibliometric measures is helpful in most countries to evaluate the quality and prestige of research centers and authors [51]. Research administrators use objective criteria for evaluating the performance of staff, departments and institutions-indeed entire countries. The alternative evaluation method is subjective depends on peer reviews by committees that undoubtedly has strong biases. Furthermore, using citation analysis for evaluation is not an ideal method, but after so many years it has got a level of standardization that permits one to obtain informed visions of their performance. Thus, the impact factors with the number of citations can achieve a better picture of performance. However, the journals’ impact factors become more important when evaluating a researcher’s most recent papers that have not yet been cited. In addition, the basic assumption is that if a document is published in a high impact journal, it most likely means something about the general quality of this paper. However, it is not a guarantee that it will be highly cited [51, 52].

As in any bibliometric study [8, 17, 22], the current study is not without limitations. First of all, we used Web of Science criteria for including ICM journals. Articles published in non-Web of Science-cited journals were not included; which might contribute to scientific research output. However, the 22 journals included in this study were the major international journals dedicated to the discipline of ICM. In addition, we searched only for journals included in the ICM category of JCR, however some articles in the ICM field may be published in other journals concerning ICM, with a wider field of interest, such as medicine and pharmacology. There are several databases that can be used to analyze scientific literature. Each database has its advantages and disadvantages [53‐56]. Such comprehensive comparison among Web of Science, PubMed, and Scopus was discussed by different researchers [53‐56]. For the purpose of this study, Web of Science was used because it offers several advantages over PubMed and Scopus with regard to the objective of the current study. First, in Web of Science various scientific disciplines are grouped into categories based on the scope of indexed journals. For example, in Web of Science, the category “Integrative & Complementary Medicine” allows researchers to retrieve documents published in ICM category. Second, Web of Science offers a powerful analysis of data from various aspects and all data can be easily transferred to Microsoft Excel for statistical analysis or graphics. A limitation of the method of ‘citation patterns’ is that older journals or articles are more likely to have been cited more, simply due to being around longer. Another limitation is that some publications from Palestine may be affiliated with Israel because SCI do not identify Palestine as a separate country. Therefore, some publications from Palestine might be not included in our analysis.