Bibliometric Study of Pain after Spinal Cord Injury

Background The prevalence of comorbid pain after spinal cord injury (SCI) is relatively high in clinical observations and has continued to increase over time. Neuropathic pain (70.14%) is the most popular subject in academic journals after SCI. However, studies that used the bibliometric method to analyze comorbid pain after SCI are still lacking. This study is aimed at combining and integrating acquired information to analyze the global trends of research on the comorbidity of pain after SCI in the last three decades (1990–2019). Methods Systematic works of literature published from 1990 to 2019 were obtained from the Web of Science Core Collection. CiteSpace software was used to analyze the relationship of publication year with the country, institution, journals, authors, references, and keywords. The regression analysis is used to evaluate the percentage of the category increase or decrease over time significantly. IBM SPSS Statistics was used in the statistical analysis. Results A total of 730 publications were included in the analysis. A remarkable increase in the number of publications was observed in the study period (P < 0.05). A total of 202 academic journals focused on the categories of clinical neurology, neurosciences, and rehabilitation, and the annual growth rate of articles in these three categories was statistically significant (P < 0.05). The USA (356, 48.77%) and the University of Miami (64, 8.77%) were the country and institution with the highest number of publications, respectively. Spinal Cord, which was the main journal for research on pain after SCI, had the most publications (88, 12.05%). Burst keywords showed that the individual, inflammation, and central sensitization with pain after SCI are the research development trends and focus in this research field. Conclusions Overall, this study provides the latest research direction for pain after SCI. This historical overview of research into pain after SCI will be a useful basis for further research into development trends, focus issues, cooperators, and cooperative institutions.


Introduction
Pain is a frequent complication of spinal cord injury (SCI), and approximately half to two-thirds of patients with SCI suffer from pain; pain after SCI is typically chronic and neuropathic [1][2][3]. The male-to-female ratio of SCI is high. Most cases are caused by traffic accidents, falls, sport activities, or violence [4]. SCI causes permanent disabilities and brings a heavy burden to people's quality of life, level of functioning, and chances of returning to work [5]. Pain after SCI manifests in many ways. More than 50% of patients with SCI suffered chronic pain within 1 year after SCI [6]. The presence of chronic pain seriously affects the patients' daily life and the social impact. Chronic pain in SCI is related to great emotional distress, and pain hinders the ability of SCI patients to participate in active rehabilitation programs [7]. Acute pain is accompanied by injury and recovery and subsides as tissue scars fade, whereas chronic pain occurs because of poor neuroplasticity [8,9]. Pain after SCI is difficult to treat because the underlying mechanisms are not yet fully understood. At present, the larger proportion of patients with SCI in most countries and regions is under 30 years old, and the highest incidence rate of 49 every 1 million was recorded in New Zealand [10]. Pain after SCI also brings a medical and economic burden to the government. People with pain after SCI in Canada spend more than US$2.67 billion annually [11].
In view of the high incidence of pain after SCI, a growing number of researchers have studied pain after SCI, and relevant articles have been published in academic journals. A number of randomized controlled trials have studied the treatment of pain after SCI [12][13][14][15]. Scientific research studies increasingly apply bibliometrics in quantitative analysis [16][17][18]. Bibliometric analysis revealed that research on the application of stem cells in SCI is a rapidly developing research field [19]. However, a quantitative analysis of pain associated with SCI has not yet been conducted.
To address the shortage of quantitative analysis of pain after SCI research, the purpose of this study is to provide a basis for the global scientific research on pain after SCI over the last 30 years (1990-2019). By being able to understand the types of pain after SCI in the past 30 years, current research hotspots provide a theoretical basis for follow-up research. Papers using CiteSpace are on the rise, especially in the medical field, where there have been many related studies [20][21][22]. CiteSpace 5.6.R5 (Drexel University, Philadelphia, USA) is a commonly used software application for bibliometric analysis. The global research trend on pain after SCI includes four aspects: the number of published papers; the distribution and cooperation between authors, institutions, and countries; a citation burst analysis of keywords; and the cocitation analysis of authors and references.

Inclusion Criteria and Exclusion Criteria.
Publications related to pain and SCI, such as articles, reviews, letters, and editorial materials, published in different academic journals were included. 872 articles were identified from the Web of Science Core Collection. Conference presentations, meeting abstracts, book reviews, news items, and corrections were excluded. After excluding 136 articles, 736 articles were included. No other specific limitation was imposed except that the chosen language was English. Six non-English language papers were excluded. Finally, 730 articles were included.

Data Extraction.
We followed a previous search strategy to search through the WoS database and then imported the gathered information to CiteSpace for analysis [22][23][24][25]. We obtained bibliometric indicators by calculating the number of publications, journals, references, citations, extracted Hindex, and keywords. The H-index is a mixed quantitative index that can be used to evaluate the amount and level of the academic output of researchers. The H-index means that an academic journal or researcher has at least H published papers that have at least H citation times per paper. For instance, an H-index of 20 indicates that academic journals or researchers had at least published 20 papers, and the citation frequency is at least 20. Besides, we calculated the number of single-author and multiauthor publications, the frequency of WoS subject categories, and types of pain category ranking percentage scores annually (the number of publications every year divided by the total number of publications in each category). The regression analysis is used to evaluate the percentage of the category increase or decrease over time significantly (the category as the dependent variable and the year as the independent variable). SPSS Statistics 22.0 software (Chicago, USA) was used for statistical analysis. P < 0:05 was considered statistically significant.  Figure 1 (b) shows that the number of article citations had a significant increase over time (t = 17:066, P < 0:001). We divided the study period of 30 years into six groups (1990-1994, 1995-1999, 2000-2004, 2005-2009   per paper (116.36). In accordance with the journal IF quartile of WoS, 35% of the 20 journals were in the first quartile (Q1), and 45% of the journals were in the second quartile (Q2). Figure 3 indicates the dual map of the journal. The map on the left represents the citing journals, and the map on the right represents the cited journals. In the dual-map overlay, the labels are marked according to the disciplines of the subject. A line connects the citing journal on the left side to the cited journal on the right side. The dual map indicates that most of the journals were from the molecular, biology, and immunology fields. Simultaneously, most journals were cited from the molecular, biology, and genetics fields.   Table 2). Figure  3.6. Distribution by Authors. The 730 articles were contributed by 1000 authors. The top 10 authors and cocited authors were ranked based on the number of journals published ( Table 2). Finnerup NB, who published 34 articles, ranked first, followed by Cardenas DD (29 publications) and Siddall PJ (27 publications). Siddall PJ was cocited 403 times, followed by Finnerup NB (269 times) and Widerstromnoga EG (164 times). Figure 9 indicates the cooperation between authors. Amongst the authors, Cardenas DD not only has many cooperative relations with Jensen MP but also has close cooperation with Turner JA. There is also close cooperation between Finnerup NB, which has the highest number of publications, and Siddall PJ, who ranks third in the number of publications. The proportion of single authors and multiple authors (authors ≥ 2) every 5 years is shown in Figure 10. The linear regression results showed that the percentage of papers with a single author significantly decreased (t = −3:557, P < 0:05) over time.     Figure 11. Cite-Space automatically generated the top 17 clusters. The modular Q value shows the significance of the community structure. The modularity Q score was 0.8557 (higher than 0.5), which indicated that the network was reasonably distributed to loosely coupled clusters. The largest cluster was labeled "neurofeedback," the second-largest clusters were "multidimensional" (#1) and "hyperexcitability" (#2), and the third-largest cluster (#3) was "quisqualic acid."

Distribution by
3.8. Analysis of Keywords. The top keyword with the strongest citation burst since 1991 was chronic pain, followed by central pain since 1992 ( Figure 12). These keywords with high citation bursts reflect the topic frontier. The current keywords with the strongest citation bursts included        Figure 7(b) shows the expansive network map of the cooperation of the countries by CiteSpace V with 97 nodes and 164 links. The link between the two nodes represents the frequency of cociting articles published by the two nodes, which implies the closeness of the connection between the two nodes. And we can easily acknowledge from Figure 7

Research Focuses on the Comorbidity of Pain and Spine
Core Injury Research. As shown in Figure 4, clinical neurology was the most prolific research field on pain after SCI according to the subject categories of WoS (355), followed by neurosciences (279), rehabilitation (210), and anesthesiology (111). The top 10 subject categories were rehabilitation, clinical neurology, neurosciences, anesthesiology, sport sciences, critical care medicine, pharmacology, orthopedics, surgery, and experimental medicine research. According to the synthetic analysis of the number of publications and citations, the number of citations per paper, and theH-index, we could acknowledge that the proportions of the top three subject categories of WoS (clinical neurology, neurosciences, and rehabilitation) were all above 20% and the number of citations per paper was all above 30, implying that the top three subject categories had superior quality and were recognized as the mainstream subject on the pain after SCI research. Based on the types of pain, the majority of the included articles involved neuropathic pain and treating pain after SCI. Amongst the top 10 types of  pain (neuropathic pain, treating pain, nociceptive pain,  animal models of pain, spontaneous pain, persistent pain,  shoulder pain, musculoskeletal pain, low back pain, and  upper extremity pain), neuropathic pain had the largest number of publications (506), citations (16,045), and open-access papers (266) and the highest H-index value (77), indicating that neuropathic pain has attracted wide attention from researchers, and it is also an urgent problem for patients after spinal cord injury.   In the cocitation map of references, "neurofeedback" was labeled as the largest cluster (#0), the second-largest clusters were "multidimensional pain inventory" (#1) and "hyperexcitability" (#2), and the third was "quisqualic acid" (#3). Based on the analysis of keywords, "chronic pain" had the strongest citation bursts since 1991. Because of the multiple causes of neuropathic pain, this research area is very broad. However, these publications were mainly focused on neurofeedback and pain management. Patients after spinal cord injury inevitably suffer from different degrees and types of pain, so pain management is becoming more and more important. In addition, emerging interventions to manage pain are diverse, including exercise therapy that is not limited by venue and time. At present, many clinical studies [36][37][38][39] have shown that patients with spinal cord injury have good compliance with exercise therapy to relieve pain. Sumizono et al. [36] and Ditor et al. [38] have confirmed aerobic exercise can significantly alleviate neuropathic pain in patients with spinal cord injury. And the finding of Sumizono et al. [36] indicated that aerobic exercise alleviated neuropathic pain through the regulation of glial cell activation and expression of BDNF in the ipsilateral spinal dorsal horn and the endogenous opioid system. In addition, the result of Detloff et al.'s [40] study also suggested that there is a critical therapeutic window when exercise therapy may be effective at treating SCI-induced allodynia and that there are postinjury periods when exercise can be deleterious.

Strengths and Limitations.
This study was the first to combine and integrate acquired information for the bibliometric analyses of the focus issues, direction, and development trend of research studies about pain after SCI over the last 30 years. These publications were retrieved from the SCI-Expanded WoS. The publication selects a variety of journals to ensure the integrity and diversity of the data. Our study included 730 articles on pain after SCI that were published in academic journals, such as Pain, Neurology, and Journal of Neuroscience. Subject categories, the number of publications and citations, the H-index in WoS, the collaborative analyses of journals and countries or institutions, and the cocitation analyses of references or authors and keywords were included.
This study has several limitations. This study only selected the SCI-Expanded WoS database for the retrieval of articles, and non-English papers were not included. Therefore, these factors may cause publication bias. Some influential papers may not be highly cited, whereas others are frequently cited, and the results are widely known.

Conclusions
This study provides the latest research direction for pain after SCI. This analysis can enable research teams to collaborate and promote the clinical management of pain after SCI. The initial three publications substantially increased to 47 publications from 1990 to 2019. The USA contributed the greatest number of published articles, and Neurology was the most influential journal on pain after SCI. Although this study has some limitations, it showed the common types of pain after SCI, especially neuropathic pain. According to the type of pain, 512 papers focused on neuropathic pain; thus, neuropathic pain is the most common type of pain after SCI. This historical overview of research into pain after SCI will be a useful basis for further research into development trends, focus issues, cooperators, and cooperative institutions.

Data Availability
All research data used to support the findings of this study are included within the article and the supplementary information file.  Table 1: raw data on journal sources of pain after spinal cord injury. Supplementary Table 2: raw data on countries/territories involved in pain after spinal cord injury. Supplementary Table 3: raw data on institutions involved in pain after spinal cord injury.