Bibliometrics indicators in the Caribbean islands: French West Indies authorship analyzed from 1989 to 2018

Bibliometrics are mathematical methods used in various scientific domains to measure scientific output based on the number of publications, the prestige of the journal, and the number of citations of the research after its publication. To the best of our knowledge, only a few bibliometric studies have been performed in the Caribbean or the French West Indies. The aim of this study was to quantify the collaborations and scientific output between the two university teaching hospitals of Martinique and Guadeloupe, at the regional, national and international level. A bibliometrics analysis was performed from the international databases Web of Science and PubMed, for the period from 1989 to 2018 inclusive (30 years). Three types of bibliometric indicators were used, namely quantitative indicators, performance indicators, and organisation-specific indicators. Affiliations of the first and last authors were identified from PubMed. Between 2014 and 2018, we recorded the affiliations of all authors (from the first to the last author) who collaborated on an article with the University Hospital of either Martinique of Guadeloupe Results

Several recent bibliometric studies have evaluated scientific repercussions or temporal trends in different areas of scientific and medical research (3). Other studies have focused on the impact of publication output of one or several institutions (4,5) or at a national level (6). To the best of our knowledge, only a few bibliometric studies have been performed in the Caribbean or the French West Indies (7,8).
For more than 30 years, national plans have been implemented in France to develop research within French hospitals, including in the regions of Guadeloupe and Martinique (French West Indies, FWI). These two regions are not geographically located in Europe, but rather, are situated in the Caribbean. University teaching 5 hospitals were set up in 1985 in Guadeloupe and in 1986 in Martinique. The implementation of successive programmes to boost attractiveness and to support emerging research teams has culminated in the existence of several research units that are certified by the French Ministry for Research and Higher Education (9), which are a rich breeding ground for research in these two regions. Indeed, the Caribbean zone presents several specific epidemiological features that represent major public health issues. Due to their geographical position, the FWI are exposed to severe climatic conditions that promote emerging vector-borne diseases (Chikungunya, dengue, Zika), as well as being exposed to meteorological phenomena such as earthquakes and hurricanes. Hurricanes have become more frequent, and increasingly severe in recent years. There is therefore a compelling need for close cooperation between Caribbean countries to face these threats.
Furthermore, the burden of chronic diseases (obesity, diabetes, cancer) is growing, and at the origin of substantial morbidity (10). Populating ageing is another major preoccupation for both islands, and is a priority topic for research, with the planned constitution of population cohorts in the coming years (11,12).
Taken together, these specificities of the FWI underscore the importance of medical and scientific cooperation between Caribbean countries, and the need for a To date, to the best of our knowledge, no bibliometric study has investigated scientific collaboration between the various academic institutions of the FWI and their scientific output. Given the specific epidemiological characteristics of 6 Martinique and Guadeloupe, it would be interesting to assess the development of research in these two regions. The aim of this study was therefore to quantify the collaborations and scientific output between the two university teaching hospitals of Martinique and Guadeloupe, at the regional, national and international level.

Methods
We performed a bibliometrics analysis from the international databases Web of Science and PubMed, for the period from 1989 to 2018 inclusive (30 years). All original research articles, reviews, letters, editorials and comments were selected; other publication types were excluded.

Research tools and analysis
To search for publications involving the University Hospital of Martinique and/or the University Hospital of Guadeloupe, we used InCites, an analytical tool developed and marketed by Clarivate Analytics (14) that aggregates summary measures of scientific output and citations at the level of countries, organisations and specific disciplines. An organization enhanced feature available in the Web of Science allows the user to search by an organization name that has been unified, to yield comprehensive and specific search results. More than 215 different search terms were thus used by InCites in search engines for these searches. and "Top 10%" indicators systematically provided by InCites were also used to identify the most highly cited articles at international level. The Journal Citation Reports were used to rank journal by quartiles of impact factor, with journals in Q1 (i.e. the top 25%) corresponding to the most widely cited journals in the field.
Organization-specific indicators, also known as structural indicators, were used to measure connections between publications, authors, and research fields. To estimate the structural indicators, the full set of data for each article identified was extracted from PubMed (15).
Affiliations of the first and last authors were identified from PubMed. Each affiliation Quantitative indicators are presented in Table 1    Detailed analysis of the bibliographic metrics is shown in Table 2. For both hospitals, the average number of authors per article was lower when an author from that hospital was first or last author. The average number of authors per article was highest when the first or last author was from mainland France. 14 ± 11 7 ± 5 8 ± 4 10 ± 11 16 ± 6 7 ± 4 15 ± 13 10 ± 5 10 ± 3 10 ± 7 Total is not just a question of initiating projects on research areas of interest, but they must also be brought to a successful conclusion, and achieve appropriate publication and visibility.
In view of the size of the population, with an estimated 770,000 inhabitants as of 1 January 2016 in Martinique and Guadeloupe together (23), the level of technical equipment also has to take account of the critical mass of patients necessary to maintain certain medical specialty services, or to render further investment and development viable. Specialties that are offered in these two regions therefore develop in synchrony, working in complementarity as regards the healthcare opportunities on offer in each island (specialists, number of positions, and technical equipment). As a result, which disciplines will produce scientific publications is often a question of the geographical location of the experts in that particular field.
This explains the predominance of certain research themes across the two regions. In terms of future perspectives, the dissemination of our findings among the university hospital research teams will be an essential step in understanding how to work towards more collaborative partnerships in the future. Among the possibilities for cooperation, participating in European partnership programmes such as ERDF or ETC (e-health CARES 2014-2020) gives improved visibility to our islands' ability to lead operational projects on topics such as the healthcare pathway, access to technological innovation, or e-health. The research areas targeted by these programmes are issues that are common to the whole region, or that aim to exploit the specificities of the region, and public health is one of these. It therefore behoves the research teams in both Martinique and Guadeloupe to demonstrate unequivocally their ability to rise to the strategic challenge of research and innovation.

Conclusions
To the best of our knowledge, this is the first study to perform a bibliometric Declarations Ethics approval and consent to participate : Not applicable.

Availability of data and materials : Not applicable
Competing interests : The authors declare that they have no competing interests. additionnalfile1.docx