Dry Eye Analysis: A Citation Network Study

Introduction Dry eye is one of the most frequent eye problems with prevalence and incidence from 5% to 50%. Citation network analysis allows us to simplify information in a visual way and provides a better understanding of the research done in a specific field. The objective of this paper is to quantify and analyse the relationships among the scientific literature in this field using citation network analysis. Materials and Methods The program used to analyse the citations was CitNetExplorer®. Previously, papers published in the research field during a predefined period were found using the keywords defined in Web of ScienceTM (WOS). Results Using the keyword “dry eye,” during the period 2007 to 2018, the most cited paper is by Lemp, MA (2007), with a citation index score of 913 in our citation network containing 6,500 most relevant papers. Analysing clustering, we found 5 relevant groups that match the main areas of research in this field: definition and classification, treatment, retina, refractive surgery, and quality of vision. Core Publication is composed of 64% of the papers in the network, which is a high percentage. It indicates a clear focus on the research carried out in this field. Conclusions This citation network analysis shows definition and classification of dry eye to be the most researched area in this field, followed by treatment.


Introduction
Dry eye is one of the most frequent eye problems reported by adults in eye care practitioner consultations. e prevalence and incidence vary from 5% to 50%, depending on the diagnosis criteria or on the type of dry eye. Women and older people are the most affected. Dry eye prevalence is rising, due to the gradual increase of the population age and to a higher incidence of some risk factors [1].
Dry eye was defined as a disease just 30 years ago. Since then, there have been many advances in the knowledge and definition of this pathology. In 1995, it was defined as a "disorder of the tear film due to tear deficiency or excessive evaporation, which causes damage to the interpalpebral ocular surface and is associated with symptoms of ocular discomfort". In 2017, a new definition was published in the DEWS II report [2].
Nowadays, according to the abovementioned report, dry eye is defined as "a multifactorial disease of the ocular surface characterized by a loss of homeostasis of the tear film and accompanied by ocular symptoms, in which tear film instability and hyperosmolarity, ocular surface inflammation and damage, and neurosensory abnormalities play etiological roles" [3]. is new definition admits the multifactorial origin of dry eye and points out the etiological factors implied in this disease.
Regarding dry eye classification, traditionally, dry eye was classified into aqueous tear-deficient and evaporative. Now, we know that both exist as a continuum and, as dry eye progresses, it is used to find characteristics of both subtypes [2]. e aqueous tear-deficient dry eye implies a failure of lacrimal tear secretion, while the evaporative dry eye is related to eyelids and ocular surface. e great number of patients suffering from dry eye and the impact of this disease on patients both financially and in terms of quality of life justify the analysis of all the research studies published about this public health problem.
Citation network analysis allows us to simplify information in a visual way and provides a better understanding of the research done in a specific field. It also lets us quantify the most cited papers and create groups based on connections between papers and citation frequencies [4]. Citation network analysis has been a very useful tool since the citation index concept appeared. ese citation networks appear as different fields related through connections. Authors, journals, and papers are an essential part of these networks [5].
Because research on dry eye pathology is so extensive, use of this citation analysis methodology will allow us to identify the most relevant authors, publications, and journals, as far as their citation is concerned, the most important years in terms of their publications, and the different clusters of study within dry eye pathology. e aim is to identify not only the most relevant research but also the different areas of study and thus focus future research on dry eye. e objective of this paper is to quantify and analyse the relationships among the scientific literature in this field using citation network analysis.

Materials and Methods
CitNetExplorer ® was used to analyse the citation networks of individual publications. is software is a tool for visualizing the most important publications in a field and showing the citation relations between these publications [6].
First, the researchers define the keywords and look for the publications according to these words in Web of ScienceTM (WOS). e file from WOS shows papers in the research field published during the predefined period. e final file is the selection of 6,500 publications of the WOS list sorted by relevance, because the citation network program does not allow more publications to be introduced. en, once the file is imported to CitNetExplorer ® , the software produces the first graph about the most cited publications, with a maximum of 40 papers for a clearer understanding. Quantitative analysis shows values of publications, citation links (total number of citations in the network), and time period. is analysis shows the most cited papers in order from the highest to the lowest, according to their citation index score. e clustering function allows us to identify groups according to the level of association among papers. In this way, subnetworks are obtained depending on the citations among them [7].
Finally, central publications are analysed through the Core Publication function, revealing the main papers of the field. For this analysis, only those that have 4 citations or more are selected.

Results
e keywords used for the search were "dry eye". e period chosen was from 2007 to 2018. e year 2007 was chosen as the starting point because of the change that the Dry Eye Workshop (DEWS) report represented in the definition and treatment of dry eye. 9,359 papers were found using the previous criteria.
We made a citation network with 6,500 most relevant papers, obtaining 44,942 citations across the network. Table 1 shows the 20 most cited papers in this network. e paper by Lemp et al., published in 2007, is the most cited, with a citation index score of 913. Figure 1 shows the graph of this network.
With the clustering function, we obtained 13 groups or clusters, 5 of them having a relevant number of papers, while the other 8 did not reach 1%. Figures 2-6 show the citation network of each group, and Figure 7 shows that there are no citations among different groups.
In group 1, we had 4,014 papers, almost 61% of the network. Lemp's paper, published in 2007 in Ocular Surface Journal [2], was the most cited in this group.
In group 2, we found 360 publications. Yoon's paper, published in 2007 in American Journal of Ophthalmology [26], was the most cited.
In group 3, we found 198 publications. Lim's paper, published in 2012 in e Lancet [27], was the most cited.
In group 4, we found 179 publications. Ambrosio's paper, published in 2008 in Journal of Refractive Surgery [28], was the most cited.
Finally, in group 5, we found 143 publications. Kaido's paper, published in 2007 in Cornea [29], was the most cited.
When we analysed relationships among clusters, we could not find any connections.
When we analysed the Core Publication, we found a total of 4,161 papers that cited, or were cited by, at least 4 papers. is 4,161 represented 64% of the papers, and the citation network across this group is 42,791. Figure 8 shows the graph of this network. Regarding clustering, we found 5 relevant groups that match the main areas of research in this field: definition and classification, treatment, retina, refractive surgery, and quality of vision.

Discussion
is clustering also shows a difference among the journals that published each of the clusters. It clearly indicates the editorial lines of each scientific journal, no matter how cross the subject was, as in the case of dry eye. e biggest cluster is the one related to definition and classification of dry eye, with more than 60% of the papers of this network. DEWS report heads this cluster due to the broad consensus on the new definition and classification of Chotikavanich et al. [25] Production and activity of matrix metalloproteinase-9 on the ocular surface increase in dysfunctional tear syndrome.

Investigative Opthalmology & Visual
Science. 2009 Jul; 50 (7): 3203-9. 2009 120 dry eye established in this report. is was used until DEWS II was published in July 2017, which included a redefinition of dry eye.
e second cluster we found is related to the treatment. It shows that dry eye is a chronic pathology and how treatment is one of the main challenges for vision Finding retina as a third cluster could seem surprising because retina has no bearing on the relationship between ocular surface and dry eye. However, it could be explained by the relationship between age and dry eye and between age and pathology of the retina. In other words, studies with ey reviewed the scientific literature and summarized the experience of the authors to propose methods for decreasing dry eye after surgery.
Finally, the fifth cluster relates to quality of vision. is is one of the concerns for patients suffering from dry eye. e absence of relationships among the clusters' most cited articles is striking. We assumed we would find some connections between clusters, but in fact, none of the five articles cites any of the others. erefore, we see five clearly different topics in five differentiated clusters.
Core Publication accounts for 64% of the papers in the network, which is a high percentage. is means there is a clear focus on the research carried out in this field. Definition and classification made up most of this Core Publication. Treatment and refractive surgery are also represented in this core, albeit in less quantity.

Conclusions
Dry eye is a very important field for researchers, with a very high number of publications and many connections among articles.
is citation network analysis shows that definition and classification of dry eye is still very important, most of the articles being related to this cluster. It is followed in relevance by treatment.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.