Medical research productivity in the Arab countries: 2007-2016 bibliometric analysis

Background The aim of this study was to assess recent trends in medical research productivity in Arab countries. Methods We collected bibliometric data for the world countries, Arab countries, and Arab institutions for 2007-2016, using Essential Science Indicators, Journal Citation Reports, and Web of Science database. We collected the number of published papers overall and per year, citations per paper, and number of papers published in top quartile and top 10% journals. For the 10 most productive institutions, we additionally collected the number of papers with correspondence authors affiliated with the institution. Results The Arab world produced 189 papers per one million people, about a quarter of the value for other world countries. Four Arab countries (Qatar, Tunisia, Lebanon, and Kuwait) produced more than 695 papers per one million people, exceeding the world average. The average number of citations per paper was 9.2; it rose to more than 15 for papers with international collaboration. At the institutional level, the number of citations showed upward trends, with six institutions having an average citation per paper higher than that of all Arab countries. For the 10 most productive institutions in Arab countries, the percentage of papers involving international collaborations ranged from 42% to 79%; of these, 9% to 29% were led by authors from the same institution. For these 10 most productive institutions, the percentage of papers published in the top quartile journals and with a lead/corresponding author from the institution ranged from 7 to 32%; that percentage drops to 1% to 10% for papers published in top 10% journals. Conclusions Although medical research output in Arab countries at both the country and the institution levels has increased over the past 10 years, it is still lagging behind the rest of the world. The percentage of papers involving international collaborations was relatively high, but the majority of these papers were led by authors from outside the local institution, particularly when published in the top 10% journals.


Results
The Arab world produced 189 papers per one million people, about a quarter of the value for other world countries. Four Arab countries (Qatar, Tunisia, Lebanon, and Kuwait) produced more than 695 papers per one million people, exceeding the world average. The average number of citations per paper was 9.2; it rose to more than 15 for papers with international collaboration. At the institutional level, the number of citations showed upward trends, with six institutions having an average citation per paper higher than that of all Arab countries. For the 10 most productive institutions in Arab countries, the percentage of papers involving international collaborations ranged from 42% to 79%; of these, 9% to 29% were led by authors from the same institution. For these 10 most productive institutions, the percentage of papers published in the top quartile journals and with a lead/corresponding author from the institution ranged from 7 to 32%; that percentage drops to 1% to 10% for papers published in top 10% journals.
Conclusions Although medical research output in Arab countries at both the country and the institution levels has increased over the past 10 years, it is still lagging behind the rest of the world. The percentage of papers involving international collaborations was relatively high, but the majority of these papers were led by authors from outside the local institution, particularly when published in the top 10% journals.
journal of health global Research in the medical sciences plays an important role in a country' s economic growth, long-term sustainable development, and improvement in the standards of living and quality of life [1]. Governments across Europe as well as the United States have heeded these claims and have increased research spending in an effort to aide their ailing economies [1].
The situation is different in Arab countries however. Arab countries are lagging behind in the number of original research publications, number of publications in top journals and citation frequency [2], to name a few. The number of medical publications from Arab institutions in all 22 Arab countries between 1996 and 2012 was 76 417 reports, which is equivalent to almost half that of Turkey VIEWPOINTS PAPERS and equal to only 4% of medical publications from United States based institutions [3]. This lag holds true across a number of other publications that have also assessed medical research in Arab countries and found it to be lagging [2,[4][5][6].
It is therefore essential to quantify current research output as it constitutes the basis for strategic planning and decisions making [7]. Bibliometric analysis, now a widely accepted method for assessing research in many fields, is employed to quantify and assess research output and also depict its growth and spread [8]. Comparisons of bibliometric characteristics between countries and institutions can reveal differences in research orientations, capacities, and collaboration patterns [9]. Additionally, universities and academic institutions increasingly rely on scientific analyses for making decisions regarding hiring, promotion, tenure, funding, and salary increases [9][10][11][12][13].
Published studies assessing medical research in Arab countries have mainly used PubMed to quantify productivity [2,[4][5][6]. Since improving research productivity in Arab countries necessitates better understanding of its current status, we aim to assess more recent trends in medical research productivity in Arab countries, in terms of both quantity and quality using Essential Science Indicators, Journal Citation Reports, and Web of Science database.

Overall design
This study compared medical research output: 1) between Arab countries combined and the rest of the world; 2) across individual Arab countries; and 3) across institutions in the region (hereafter referred to as "institutions").

Eligibility criteria
We used the following eligibility criteria: • Countries: we included all 22 countries classified as Arab countries according to the Arab League [14]; • Institutions: we included the 10 most productive universities or research hospitals in Arab countries. We could have chosen a larger pool, for example, 15, 20,

Data collected
We collected for each country the following data: • Population size -using World Bank DataBank [15]; • Number of published papers per year; • Number of papers with international collaboration for Arab countries; that is, papers that include at least two authors with respective affiliations from two different countries. We sub-categorized VIEWPOINTS PAPERS collaborations according to whether they are with: 1) Arab countries; 2) the European Union (including the United Kingdom); and 3) Canada and the United States; • Citations per paper.
We also identified the top 25 medical research journals according to the number of papers published in those journals with at least one author affiliated with an institution from Arab countries, and specified their most recent impact factor scores.
We collected for the 10 most productive institutions in Arab countries the following data: • Number of published papers per year; • Citations per paper; • Number of published papers involving international collaboration; • Number of papers published in top quartile (Q1) and top 10% journals in their respective fields, as classified by the Journal Citation Reports. These were identified as follows: We first identified the top 25% and top 10% journals in each of Journal Citation Reports' 200 plus subject areas and then matched the titles against the 4530 journals classified as medical research by Essential Science Indicators to come up with a list of top 25% and top 10% journals in medicine; • Number of papers with correspondence authors affiliated with institutions in the Arab states. For Arab countries alone, we calculated:

Data analysis
• The percentage of papers produced through international collaborations with other Arab countries, countries from the European Union, and North America for each country and for the Arab countries combined; • Citation rate stratified by whether paper involved an international collaboration or not.
The second set of analyses compared the research output of institutions in Arab countries. We calculated for each institution: • We did not explore the distribution of our data due to its huge size and the limitations in extraction; extracted as a lump rather than separate variables. We therefore calculated averages where applicable.

Arab counties vs world
We identified a total of 76 978 papers published by research groups in Arab countries between 2007 and 2016.  Arab countries amongst each other Table 2 shows the medical research output of researchers in individual Arab countries. Research groups from Qatar, Tunisia, Lebanon, and Kuwait (in descending order) had a total number of papers per one million people above the world average of 695 papers per million. The number of citations per paper was higher than the world average of 15.1 in Sudan alone. Arab countries that had publications with a citation rate higher than that of all Arab countries combined were in descending order Sudan, Lebanon, United Arab Emirates, Qatar, Oman, Saudi Arabia, Jordan and Iraq.

VIEWPOINTS PAPERS
top quartile and only five have an impact factor above 3.0 ( Table 4). Also, only 24% of Arab countries' papers are published in top quartile journals in comparison to 38% for the rest of the world.

Institutions in Arab countries
The ten most productive institutions in the Arab countries are listed in Table 5. These institutions were from Egypt (n = 4), Saudi Arabia (n = 3), Lebanon (n = 1), Kuwait (n = 1), and Tunisia (n = 1). Research groups from the Kuwait University in Kuwait and the American University of Beirut in Lebanon had a greater than 50% share of research productivity contribution per country. Papers published by six of the 10 most productive institutions had an average citation per paper higher than that of Arab countries as a whole; none had an average citation per paper higher than world average. Figure 2 depicts the produc-   Table 6).
The percentage of papers with lead/ correspondence author from the   in Arab countries (green band), with percentage with international collaboration (blue band), and percentage with both international collaboration and lead authorship from the institution (purple band). Surface area is not proportional to percentage and institutions have not been ranked in any specific order. The percentage of papers with international collaboration and lead/correspondence author from institution was calculated from the total number of papers per institution.
same institution was higher than 50% in eight Arab institutions (Figure 3). Only two of these institutions' researchers had more than 20% of their papers published in top quartile journals. Researchers from one institution only had more than 10% of their papers published in the top 10% journals. Figure 4 illustrates the international collaborations between research groups for the ten most productive institutions, which ranged from 42% to 79% across the 10 institutions. The percentage of papers involving international collaborations and led by authors from the same institution ranged between 9% and 29% across the 10 institutions.

DISCUSSION
In summary, this study used bibliometric indicators to describe medical research productivity in 22 Arab countries during the last decade. Over the 2007-2016 time period, the number of medical research publications per one million people for researchers in Arab countries was almost about a fourth of that of the world. However, medical research productivity has been increasing in these countries at a faster rate, relative to the rest of the world during the same time period.
The study has a number of strengths. First, and to our knowledge, this is the first article to analyze the quality and quantity of medical research in Arab countries using bibliographic resources such as Essential Science Indicators, Journal Citation Reports, and Web of Science database. Previous studies on research productivity in Arab countries have mainly used the Medline database [2][3][4][5][6]. Nonetheless, our finding of paucity in medical research productivity is consistent with the findings of those studies [2][3][4][5][6]. Second, we adjusted the raw counts of publications by population number to avoid misleading results. As such, our study showed that productivity is higher in small countries, such as Qatar, Tunisia, Kuwait and Lebanon than in Saudi Arabia and Egypt. Other investigators reported similar findings for small countries when adjusting for population size [4,5].
The study also has a number of limitations. Although we included papers from a sizeable number of medical research journals (as classified by Essential Science Indicators), there is probably a number of local medical journals that were not captured by the used databases. However, we believe that including more journals would increase the numbers but in a proportional way that would not affect the interpretation of the results. Finally, the inclusion of "case reports" in this study (given the Web of Science classifies them as articles) could have lowered the number of citations per papers, particularly since major journals do not accept to publish case reports.
One of our major findings is the paucity in medical research productivity in Arab countries. Reasons stipulated to explain this paucity include political instability in these countries, regional conflicts [5], lack of proper research infrastructure and equipment [3], lack of freedom and democracy, brain drain, lack of funding, and difficulty of publishing in high impact journals [16,17].
At the institutional level, almost all institutions showed upward trends, although to different degrees, with six of them having average citations per paper higher than that of Arab countries as a whole. The ranking of these institutions changes when focusing on quality as opposed to quantity indicators.