Introduction
Mankind has been trying to understand and treat spinal pathologies since the beginning
of human civilization. Earliest description of cervical spine fractures and its treatment
are found in the Babylonian papyrus (1550 BC).[1] The first recorded evidence of cervical spine surgeries being performed is in the
Hammurabi's code of law ((1955–1912 BC).[2] As understanding of the anatomy of cervical spine increased, there were attempts
of surgeries for cervical spine. Both the anterior and posterior approach were being
practiced. The first description of posterior cervical spine surgery was given by
Hadra in 1891.[3] Anterior approaches to the neck were developed for other conditions needing exposure
of oropharynx, buccal cavity, retropharyngeal space, etc.[4]
[5]
[6] There was an underlying fear of going anteriorly where many vital structures came
in the way. Hence, posterior approach was favored by many surgeons due to the familiar
anatomy and absence of any neuro vascular structure in the field. The initial results
were poor due to instability arising from lack of bone grafting techniques and weak
implants. Improved results were reported in the 1950s with use of bone grafting.[7] Only indication for cervical spine surgery in earlier days was treatment of complications
of Pott's disease but some contemporary surgeons extended the indications to cervical
spine fractures, instabilities and tumors as well.[4]
[8] Gradually, surgeons moved away from posterior approach surgeries to the anterior
approach surgeries in the late 1950s as it did not compromise the posterior stabilizing
structures of the cervical spine.[9]
[10]
[11] The interest in the posterior approach was renewed in the 1970s after the Japanese
introduced various forms of laminotomies to overcome the instabilities arising from
earlier posterior surgeries involving laminectomies.[12]
[13]
[14] There is greater acceptability of posterior instrumentation of cervical spine among
surgeons worldwide that was helped by the availability of strong implants like pedicle
screws, hooks, etc. and development of novel surgical techniques like lateral mass
fixation.[15]
[16] Concurrent developments in the field of microscopy and widespread availability of
computed tomography scan and magnetic resonance imaging further increased the accuracy
of surgeons performing posterior instrumentation in the cervical spine.[17]
[18] Recent advancement in the field of three-dimensional (3D) printing has made it possible
to safely plan and operate upon the pediatric cervical spine and fixation of atlanto-axial
vertebrae.[19]
[20]
There has been a lot of literature being published on posterior instrumentation of
cervical spine. The rate and volume of publication have increased rapidly in recent
times and it is not possible for individuals to be updated with the latest developments
in the field. Hence, there is a chance of missing a critical development or relevant
information. Literature reviews become important in this respect to effectively analyze
the data already published in scientific literature and provide an evidence-based
insight for the researchers to upgrade his professional judgement and expertise.[21] Among various qualitative and quantitative approaches available for performing literature
reviews, bibliometrics provides an objective and a reliable analysis. It does a statistical
measurement of the science, scientist, and scientific activity.[22] The overall process is complex and requires an extensive knowledge of programming
skills among the researchers to be able to employ the steps involved in data analysis
and visualization necessary for bibliometric analysis. It is made easy by the availability
of automated workflow software and tools.[23]
This study aims to perform a bibliometric analysis of all the articles published on
the topic of posterior instrumentation of cervical spine and to study the general
publishing trends over the years in this topic in terms of journals, authors, topics,
keywords, collaborating countries, etc.
Material and Methodology
Data Extraction
The published data on posterior instrumentation of cervical spine was extracted from
“Web of Science (WOS)” for the analysis.[24] We included articles published in English language only. Only the original and review
articles were included in the data extraction and analysis, while editorials, letter
to editors, conference presentations, and commentaries were excluded from the list.
The query was designed to look for all the available articles included the terms “cervical
spine,” “surgery,” “posterior instrumentation,” “pedicle screw fixation,” and “lateral
mass fixation” with appropriate Boolean operators as applicable ([Table 1]).
Table 1
Inclusion and exclusion criteria for data extraction
|
Inclusion criteria
|
Exclusion criteria
|
|
Original articles & review articles with key words “cervical,” “surgery,” “posterior
instrumentation,” “pedicle screw fixation,” “lateral mass fixation”
|
Letter to editors, notes, conference papers
|
|
Language—English
|
|
|
Year—1991 to 2022
|
|
Data Analysis
The available data from WOS was extracted in plaintext format and was analyzed using
the Bibliometrix R package (University of Naples Federico II, Naples, Italy).[25] The data thus generated had information regarding authors, document type, journals,
number of cited references, titles of the studies, total number of citations, year
of publication, affiliation of the authors, corresponding authors, keywords, etc.
Bibliometric analysis was done using performance mapping and science mapping.[26]
Results
Generic Attributes of the Dataset
After applying the inclusion and exclusion criteria, a total of 1,953 documents were
found on the topic in WOS, published between 1991 and 2023. The annual growth rate
for the publications was 3.99%. The average age of the documents was 10.7 years. There
were a total of 1,782 articles and 171 reviews published in 198 journals. A total
of 3,421 author's keywords and 23,133 references were used. The average citations
per document came to 22.62 ([Fig. 1]).
Fig. 1 Main information.
Analysis of Authors
A total of 6,725 authors wrote all the included studies. Only 34 of them had written
one or more articles independently and the rest collaborated either locally or internationally
for writing articles. The average number of co-authors per document was 5.63. The
international co-authorship was found in 13.11% of the articles ([Fig. 1]). Sonntag VKH and Xu RM were the most relevant author with the maximum number of
articles published (n = 28) followed by Riew KD (n = 27; [Fig. 2]). While Sonntag VKH published his first article in 1991 and has last publication
in 2019 & Xu RM had his first publication in 1995 and his last publication came in
2018, Riew KD started publishing in 1998 and has a recent publication even in 2023
([Fig. 3]). Abumi K (h index= 17, g index= 24, m index = 0.68), Sonntag VKH (h index= 17,
g index= 28 and m index= 0.515), and Vaccaro AR (h index= 17, g index= 24, m index = 0.629)
had the highest impact in terms of “h-index” followed by Riew KD (h index= 15, g index=
26, m index = 0.6) ([Fig. 4]). In terms of local citations Abumi K (n = 717), Ito M (n = 470) and Xu RM (n = 465) were the top three authors ([Fig. 5]). The graph based on “Lotka's law” shows that majority of the authors (n = 4972, 73.9%) have contributed to a single article only, while the number of authors
who have contributed to five or more articles is very low (n = 320, 4.75%) ([Fig. 6]). The authors collaboration network showed a total of 11 clusters. In terms of number
of collaborating authors, cluster 3 (maximum collaboration between Vaccaro AR and
Hillbrand AS and Xu RM and Ebrahiem NA), cluster 5 (maximum collaboration between
Imagama S and Fujibayashi S), cluster 6 (maximum collaboration between Riew KD and
Lee SH), and cluster 10 (maximum collaboration between Ames CP and Theodore N) are
the largest ([Fig. 7]). The “3 plot graph” shows that most of the top 10 authors were from United States
and majority of them were associated with Barrow neurological institute, Arizona,
United States ([Fig. 8]).
Fig. 2 Most relevant authors.
Fig. 3 Authors production over time.
Fig. 4 Author local impact.
Fig. 5 Most local cited author.
Fig. 6 Lotka's law.
Fig. 7 Author collaboration network.
Fig. 8 Three plot graph—country, author, affiliation.
Analysis of Journals
The “Spine” journal (n = 335; h index = 64, g index= 103, m index= 1.93, TC = 14580) has been the leader
in terms of h-index, number of articles published, and total citations. The “European
spine journal” (n = 188; h index = 38, g index= 46, m index= 1.52, TC = 4862) is the distant second
in the list ([Fig. 9]). The “Spine” started publishing on cervical spine in 1991 and has been the leading
journal since then in terms of number of articles published. “World Neurosurgery”
is the latest in the top 10 list of journals and started publishing in 2009 and has
shown a rapid increase in the number of articles published ([Fig. 10]). The “Spine” (n = 16494), “European Spine Journal,” (n = 3576) and “Journal of Neurosurgery” (n = 3451) are the top three journals in terms of citations by the articles included
for the bibliometric analysis ([Fig. 11]). Three journals namely “Spine,” “European Spine Journal,” and “Journal of Neurosurgery—Spine”
were identified as the major contributors of articles on posterior instrumentation
of the cervical spine as per “Bradford's Law” ([Fig. 12]).
Fig. 9 Source local impact.
Fig. 10 Source production over time.
Fig. 11 Most local cited source.
Fig. 12 Bradford law.
Analysis of Relevant Countries
United States (n = 1720), China (n = 708), and Japan (n = 500) are the top three countries in term of number of publications ([Fig. 13]). In terms of international collaboration among authors, United States (n = 692, single-country publication [SCP] = 597, multi-country publication [MCP] = 95),
China (n = 361, SCP= 331, MCP = 30) and Japan (n = 211, SCP= 204, MCP= 7) are the top three countries. Canada (MCP ratio= 0.348),
Switzerland (MCP ratio= 0.238), Germany (MCP ratio= 0.214), and Korea (MCP ratio=
0.172) ranked highest, among the top 10 countries, in terms of fraction of total publication
involving authors from multiple countries ([Fig. 14]). The authors from United States collaborated mainly with authors from China, Korea,
Germany, and Canada. Similarly, authors from China collaborated with authors from
Korea, Denmark, Germany, Greece, and India in most of their articles. Indian authors
have very little collaboration with international authors compared with authors from
other countries. They mainly collaborate with authors from United Kingdom, Indonesia,
and Iraq ([Fig. 15]). United States (n = 19573, average citation per document = 28.3), Japan (n = 5222, average citation per document = 24.7), and China (n = 3975, average citation per document = 11) are the top three countries in term of
citations of their articles ([Fig. 16]).
Fig. 13 Country's scientific production.
Fig. 14 Top 10 countries based on multicountry publication (MCP) and single-country publication
(SCP).
Fig. 15 Country's collaboration network.
Fig. 16 Most cited country.
Analysis of Affiliations
The University of California, San Francisco (n = 52) followed by Barrow Neurological Institute, Arizona, United States (n = 47) and Catholic University of Korea, Seoul (n = 41) were the top three universities based on the number of publications on the
relevant topics ([Fig. 17]). Barrow Neurological Institute, Arizona, United States, and University of Iowa,
United States started research and publication on the topic much before 1991. University
of California, San Francisco, started their journey in the field in 2003 but has quickly
surpassed all the institutes in terms of number of articles published ([Fig. 18]).
Fig. 17 Most relevant affiliation.
Fig. 18 Affiliation production over time.
Analysis of Documents
A total of 1953 articles were included in the study. The most commonly used author's
keywords were “cervical spine” (n = 440), “pedicle screw” (n = 152), and “biomechanics” (n = 119) ([Fig. 19]). The co-word analysis showed three cluster of descriptors, cluster 1 was dominated
by keyword like “fixation,” “spine,” and “complication,” while cluster 2 was dominated
by the keywords “anterior,” “decompression,” “plate fixation,” and “follow up” and
the main keywords in cluster 3 were “fusion,” “cervical spine,” “stabilization.” and
“instrumentation” ([Fig. 20]). The article by J Harms et al (DOI—10.1097/00007632-200111150-00014, Spine, 2001)
had the highest number of citations by authors worldwide (total citations= 956, total
citations per year = 41.56) followed by article by ID Gelalis et al (DOI—10.1007/s00586-0110201103,
European Spine Journal, 2012) (total citations= 395, total citations per year = 32.91)
and K Abumi et al (DOI—10.1097/00007632-200004150-00011, Spine, 2000) (total citations=
341, total citations per year = 14.2) ([Fig. 21]). In terms of citation within the articles being analyzed, J Harms et al (DOI—10.1097/00007632-200111150-00014,
Spine, 2001) (n = 272) and EL Jones et al (DOI—10.1097/00007632-199705010-00009, Spine, 1997) (n = 240) topped the table ([Fig. 22]). The analysis of co-citation network of the contributing authors shows two clusters.
Cluster 1 was formed with the article by Abumi K et al (DOI: 10.1097/00007632-199907150-00007,
Spine, 1999) as the reference article and the cluster 2 was formed with the article
by Wright NM et al (DOI—10.3171/foc.1998.4.2.2, Journal of Neurosurgery, 1998) as
the reference article ([Fig. 23]). The historical analysis shows two cluster of articles being published. Cluster
1 consists of articles published earlier and deal with pedicle screw fixation and
lateral mass fixation for cervical spine, while cluster 2 consists of articles published
in recent time and they deal with more complicated procedures like atlanto axial fixation
([Fig. 24]).
Fig. 19 Author's keywords.
Fig. 20 Co-word network.
Fig. 21 Most global cited document.
Fig. 22 Most local cited document.
Fig. 23 Co-citation network.
Fig. 24 Historiograph.
Analysis of Emerging Trends
The thematic map shows that the basic themes of research over the years have been
focusing on anatomy and lesions of cervical spine, decompression of the cervical spine,
and cervical spine fixation and fusion techniques, but the research in recent times
is focused more on the accuracy of insertion and placement of screws in the lateral
mass, pedicles or interbody fusion, use of CT scan, and atlanto axial fixation ([Fig. 25]). The analysis of the trend topics using author's keywords and the parameter set
to a minimum word frequency = 5 and number of words per year = 1, we find the keywords
“nerve root injury” (1998), “spinal stabilization” (1999), transpedicular screw (2000)
and animal model (2006) appear in early period denoting the focus of research on development
of techniques of pedicle screw fixation for cervical spine stabilization. The author's
keywords found in recent times are atlantoaxial dislocation (2016), subaxial cervical
spine (2018), 3D printing (2019), and paravertebral foramen screw (2022) indicating
a shift in focus of research more on accessing the upper cervical spine as the techniques
for middle and lower level cervical surgery have been perfected by the research over
the years ([Fig. 26]).
Fig. 25 Thematic map.
Fig. 26 Trend topics.
Conceptual Structure
The factorial analysis of the author's keywords using multivariate correspondence
analysis shows the keywords “spine,” “cervical,” “cervicothoracic junction,” “lateral
mass screw,” “spinal instrumentation,” “spinal fusion,” and “surgery” contribute maximally
to the articles selected for the analysis, while the keywords like “atlantoaxial instability”
and “vertebral artery” contribute the least ([Fig. 27]).
Fig. 27 Factorial analysis.
Discussion
Posterior instrumentation of cervical spine has been favored by spine surgeons across
the globe since early days. There has been a rapid advancement in the field with improvement
in the quality of implants and development of novel tools and techniques for fixation.
The development of high-resolution imaging modalities and availability of intraoperative
imaging and navigation tools for assistance during surgery have increased our accuracy
in terms of implant placement and have improved surgical outcomes in terms of final
reduction achieved and decreased complication rates. There is a large amount of literature
available on the topic (n = 1953 for WOS, n = 2105 for PubMed and n = 2928 for Scopus) that needs the help of software for detailed data analysis and
obtaining qualitative and quantitative.
We have used WOS for obtaining relevant data because it is one of the earliest and
largest citation index platform and the metadata provided in this platform is complete
in terms of authors, type of documents, journals, citations, keywords, and references
sourced across multiple databases. The data available from PubMed and Scopus is not
complete ([Fig. 28]).
Fig. 28 Missing data table.
It is evident that the annual scientific production in terms of articles published
has been rising steadily since 1991. There were only six articles published in 1991
and the number rose to 132 in the year 2017 (highest for any given year till date).
There have been 111 articles published in 2022 and the year 2023 has already has 21
publications till May ([Fig. 29]). The average citations per year rose from 1991 upto 2001 but has seen gradual decline
after that because of the exponential increase in the annual article production rate
([Fig. 30]). The “Bradford law” was proposed in 1948. It states that “if we arrange the journals
in decreasing order of article productivity on a given subject, then we may divide
them into a nucleus of journals more particularly devoted to the subject and several
other groups containing the same number of articles as the nucleus. The number of
articles in the nucleus and the succeeding groups will be 1:n:n[2]:…” The law is useful to assess the completeness of data, which gives a smooth curve
on the graph. The graph also helps us to know the specific journals that are most
relevant to the topic.[27] A team of authors are usually involved in the publication process; only 40 out of
a total of 1953 publications were by a single author. The collaboration was mostly
among authors from within the country as the international co-authorship was seen
in only 13.11% of the publications. United States being the leading country in terms
of number of publications has the maximum number of authors involved in publications.
The international collaboration is found only in 13.7% of the publications from United
States' similarly the “MCP ratio” for high-volume countries like China (0.08), Japan
(0.03), and Korea (0.17) is also very low, while the MCP ratio for countries with
comparatively low volume of articles published is much higher. This can be explained
by the presence of necessary expertise and equipment needed for research and innovation
within the country for United States, China, and Japan but the researchers from other
countries have to depend on the assistance and collaboration from researchers of the
above countries because their domestic set up is lacking. The researchers from United
States, Japan, China, and Germany were pioneers in the related field and contributed
to the development of cervical spine instrumentation. This explains the huge number
of citations for the publications done by researchers from these countries. Co-word
analysis gives a graphical representation of the keywords (descriptors) most commonly
and strongly associated with the documents being analyzed. It is helpful in analyzing
the content of the documents as well as identifying the popular topics and trends
in the particular field of research. The words appearing clustered together show the
strength of their association.[28]
Fig. 29 Annual scientific production.
Fig. 30 Average citation per year.
Strengths and Limitations
It is the first comprehensive bibliometric analysis of over 30 years of research on
posterior instrumentation of cervical spine. It has identified the most important
journals, most relevant authors, most cited articles, nature and amount of collaboration
among countries and authors, the landmark papers in the field, and the evolution of
trending topics on the subject. The study gains its strength from the fact that data
was collected from WOS that is a very comprehensive source and gives complete information
regarding citations, indexing, and affiliations of the contributing authors. The bibliometric
analysis was performed. Bibliometrix package of R software giving an automated and
easy to understand result with beautiful illustration of all the tables generated.
The main limitation is the time boundaries used for collection of studies that may
have led to exclusion of some important articles and authors beyond the mentioned
time period. We have also limited our study to the WOS as the only source of articles
and not considered other sources like PubMed, Scopus, Google Scholar, etc.; thus,
it may have missed on some important articles.
Conclusion
The findings of this study are important as they enhance the knowledge on the topic
of posterior instrumentation of cervical spine and helps in decision making. They
also help to highlight the gaps in present literature and could guide the trends and
directions of future research and innovation in the field.
It is evident that research and publication have grown over the years. United States
has emerged as a leader in terms of number of publications and citations. It is a
leader in terms of author's collaboration too. The university of California, San Francisco,
is the leading institute, while VKH Sonntag and RM Xu are the leading authors. The
“Spine” journal is most relevant. The article by J Harms (DOI – 10.1097/00007632-200111150-00014,
Spine, 2001) has been the most cited document. The most commonly used author's keyword
is “fixation.”
