The 100 most influential spine fracture publications

Introduction

Given the importance and traction of evidence-based medicine in recent years, research studies are instrumental not only in the understanding and advancement of disease management, but in clinical decision-making. Unfortunately, the task of searching through medical literature can be daunting and overwhelming given the abundance of information available. While new studies are being published at increasing rates, few studies make lasting contributions to a field.

Citation analysis, also known as bibliometric study, is a quantitative analysis of written publications that has become a popular method to identify studies that have a high impact on a particular field of medicine (1). While many fields in medicine and multiple fields of orthopaedics have undergone extensive literature reviews to identify the most influential articles published, there is no comprehensive review of the most cited articles related to spine fracture management.

Advancements in the management of spine fracture have had a beneficial effect on the long-term prognosis of these potentially devastating injuries. Mismanagement of certain injury patterns could result in further injury to the patient. The goal of this study was to identify and analyze the 100 most cited publications related to spine fracture management through an extensive literature search method, used and validated in previous bibliometric analyses (2-6). An appreciation and understanding of these influential publications can help clinicians with treatment decisions as well as aid researchers in future project development.

Methods

The Preferred Reporting Guidelines for Systematic Review (PRISMA) guidelines and checklist items were used to guide the methods. These items included stating our objectives and rationale (to isolate the most cited publications related to spine fracture), detailing search methods via Web of Science, eligibility criteria (peer-reviewed articles and review articles related to spine fracture), a multi-step screening of search results (Figure 1), and data collection. The search method was conducted as follows: three investigators determined the appropriate Boolean search statement and search parameters based on the intended objectives of the paper and Web of Science capabilities. Web of Science was utilized for the search because of its breadth of content and analytical capabilities. A Web of Science query was carried out on May 5, 2018. The query involved the following search statement: ((Fracture OR rupture OR displacement OR fixation OR repair OR classification OR chance OR burst OR injury) AND (Spine OR vertebra* OR cervical OR thoracic OR thoracolumbar OR lumbar OR lumbosacral OR sacral OR sacrum)). This search statement was utilized because it contains the various terms associated with: (I) the location (spine); and (II) the type of injury (fracture of any given classification); the investigators concluded that these search terms would appropriately narrow the search to the topic of interest while ideally capturing all articles related to spine fracture. The “*” indicates all endings to this particular word are included in the search.

Figure 1 Modified PRISMA flowchart.

Regarding specific search filters and criteria, articles from both medical and non-medical journals were included. Only articles written in English were included. This search was not restricted to any particular data range. The search was restricted to peer-reviewed articles and review articles; thus, all other paper classifications such as letters, editorials, and conference papers were excluded. No restrictions on the time of publication were placed on the original search. No restrictions were placed on study type, such as clinical trials versus cohort studies versus cross-sectional studies. No restrictions were placed on the specific cohorts of included study. No restrictions were placed on authors, source titles, funding agencies, organizations, editors, research areas, regions of origin, or paper categories. The resulting articles were sorted by descending number of citations. The search results were saved and the first 12,000 results along with all corresponding citation and journal source data were exported to Endnote Online and Microsoft Excel for screening.

The screening process involved a multi-step approach. Two investigators sequentially reviewed the results by title and abstract when necessary to gather only articles that pertained to the treatment, prognosis, or management of spine fractures. The initial screen was stopped when 100 articles were isolated for further review. The second screening step consisted of two investigators reviewing the abstracts of the 100 articles to ensure that all inclusion criteria were met. Full manuscripts were reviewed if necessary to verify if an article met inclusion criteria. Publications pertaining to injuries of the spinal cord, vertebrae, or cauda equina were permitted for inclusion. Additionally, studies pertaining to osteoporotic compression fractures were included as these fractures often result from minor traumas and potentially have the same treatment options as non-osteoporotic compression fractures. Articles that discussed treatments for other spinal conditions or injuries, such as tumors or infections, were excluded. Publications in question for inclusion were reviewed independently by the lead author (CJD). As a final screening step and a primary step of data collection, all included publications were reviewed by one investigator for LOE and article topic data. No articles were determined to fail to meet inclusion criteria; therefore, no articles were excluded in this final screen. Articles for which LOE or article type data was unclear were discussed with a second investigator.

The 100 most cited articles were isolated for additional analysis. The number of citations in the most recent year was used to systematically establish the order of articles when an equal total number of citations occurred. On May 22, 2018, 17 days after the initial search was conducted, the final 100 articles were further reviewed via Web of Science for the following: publication year, country of origin, journal of publication, article type, frequency of citations, average citations per year, citations in 2017, and authorship frequency. May 22, 2018 was effectively the last access date as far as data collection.

Each publication’s institutions of origin were determined by manual article review. This institutional information was cross-referenced with the Web of Science organizational information, and the combined data was utilized. The level of evidence and article type was assigned to each publication after careful review of the abstracts and manuscripts.

Our initial list is subject to bias in that the more recent articles had a shorter time window between the date of publication and the date of our search; therefore, these articles had a more limited opportunity for peer citation. In an attempt to mitigate this bias, we identified the top 10 most influential articles of the past decade. Our original Web of Science search results were refined to include only articles published between 2008 and 2018. Citation data for these articles was gathered, and the articles were sorted by decreasing number of average citations per year, as opposed to total citations.

Results

The initial search yielded 110,809 publications that met the criteria of our search term statement. After sequentially reviewing the first 1,178 results, ordered by decreasing number of citations, we found 100 articles that specifically pertained to spine fracture management (Figure 1).

The most productive decade was 2000–2009 (41%); the second most prolific decade was 1990–1999 (31%). Between 1950 and 2010, a positive trend between the year of publication and the number of publications was noted (Figure 2). Based on our methods, we documented every country affiliated with a single publication, as opposed to the country of the first or last author only. In all, 16 countries contributed to these 100 publications with the United States as the greatest contributor (n=65). The second greatest contributor was Canada (n=12), followed by France (n=10). Of note, 10 countries contributed to three or more articles (Figure 3).

Figure 2 An analysis of articles published from 1950 to 2018.

Figure 3 A summary of the countries of origin of the articles.

The range of publication years for the 100 articles was 1953 to 2010. The articles were collectively cited a total of 30,129 times. The most cited paper was cited 1,076 times; the least cited paper was cited 156 times. The average number of citations in 2017 was 18.5. The average number of citations per year since the year of publication was 15.9. The average total number of total citations was 301 (Table 1). The first ranked article by F Denis from 1983 explores the middle osteoligamentous complex and classifies major and minor spinal injuries to discuss treatment based on a three-column approach. The second ranked article is “A Randomized Trial of Vertebroplasty for Osteoporotic Spinal Fractures” by DF Kalmes from 2009. The third ranked article is “A Randomized Trial of Vertebroplasty for Painful Osteoporotic Vertebral Fractures” by DF Buchbinder. The earliest publication is Cloward’s explanation of back pain associated with ruptured discs and the utility of spinal fusions for this condition (Table 1).

Table 1 100 most cited spine fracture publications
Full table

Spine contained the most articles, accounting for 27% of the publications. The second most recurring journal was the Journal of Bone and Joint Surgery- American Volume, accounting for 13% of the publications. Journal of Neurosurgery and New England Journal of Medicine were tied as the third most recurring journal, each accounting for 5% of the articles. Interestingly, articles within the list are associated with 32 different journals (Table 2). We of Science provided the impact factor of each listed journal.

Table 2 Journal of origin
Full table

The most common level of evidence was IV (n=36). Within this article list the following numbers of articles were at each level of evidence: 9 (level I), 11 (level II), 16 (level III), 36 (level IV), and 28 (level V) (Figure 4).

Figure 4 Number of articles at each level of evidence.

Regarding types of articles, clinical outcomes (n=39; 39%) was the most common. Surgical technique (13%) was the second most common article type, and 10 articles pertained specifically to biomechanics (Figure 5). There were 13 authors contributing to at least 3 publications in this list. H Deramond (n=4; 4%) and A Vaccaro (n=4; 4%) were tied as the most productive author (Table 3).

Figure 5 Frequency of each article type.

Table 3 Author frequency
Full table

Moreover, 10 institutions contributed to at least four articles on the list. University of Maryland and University of Toronto, accounted for the largest contributors with seven articles each. The next most contributive institution was University of British Columbia with six articles (Table 4). Two particular special topics were discussed in 41% of our top 100 articles: osteoporosis (n=34) and pedicle screws (n=7).

Table 4 Contributing institution
Full table

Web of Science assigned one or more publication categories to each article. According to this analysis, 51 articles were related to orthopaedics; 45 articles were related to clinical neurology; 32 articles were related to surgery.

With regards to the second search, which was focused on publications from the past decade relating to spine fractures, the highest number of citations per year was “A Randomized Trial of Vertebroplasty for Osteoporotic Spinal Fractures” by DF Kalmes in 2009. The second ranked article was “A Randomized Trial of Vertebroplasty for Painful Osteoporotic Vertebral Fractures” by DF Buchbinder from 2009. The third ranked article was “Mortality Risk Associated with Low-Trauma Osteoporotic Fracture and Subsequent Fracture in Men and Women” by D Bliuc in 2009 (Table 5). Interestingly, the first and second ranked articles in the second search list were ranked second and third in the original search list (Table 1), indicating that these two articles had both a very high number of total citations and a very high number of average citations per year.

Table 5 Most cited publications per year from the past 10 years
Full table

Discussion

This study highlights the articles and authors that have had some the greatest impact on spine fractures in the last century and the beginning of this century. The findings of this study identify many of the articles responsible for the development of current management strategies for spine fractures, including biomechanical concerns. The most cited paper in spine fracture literature is the classic work of Denis from 1983, which established the 3-column model of the spine (7). This paper introduces the middle osteoligamentous complex and classifies major and minor spinal injuries to discuss treatment based on a three-column approach. This model is often used to characterize fractures as stable or unstable injuries, depending on the involvement of 2 or more columns.

The second most cited work is from Kallmes et al. within the last decade involving a randomized controlled trial of vertebroplasty for osteoporotic spinal fractures (8). In this study, 131 patients were randomized to either vertebroplasty or simulated vertebroplasty without introduction of polymethylmethacrylate (PMMA). The findings were remarkable that both groups showed clinical improvement with regards to pain and disability, which were sustained at one month, without statistical significant differences between the groups. These results lead authors to conclude that factors outside of injection of PMMA in osteoporotic fractures may account for the clinical improvements observed in the placebo group.

Similarly, the third most referenced paper, Buchbinder et al., is another randomized trial of vertebroplasty for painful osteoporotic fractures published within the last decade (9). This study evaluated 71 patients who were randomized to vertebroplasty versus sham procedure. This study and the findings were similar to Kallmes et al. with the major difference being in study design (length of follow up of 6 months) (10). This article also failed to show significant difference of pain reduction in vertebroplasty versus sham procedure.

Osteoporosis was the most common topic in the top 100 papers, with 34 publications focused on the topic. Osteoporotic vertebral fractures are widespread in the population with an estimated 550,000 to 700,000 fracture occurring annually (9,11). These fractures have a vast economic implication on the population as the prevalence is expected to continue to increase with the aging population. Burge et al. estimated cost of vertebral compression fractures to account for about 1.1 billion dollars in 2005, with estimated growth of approximately 50% by 2025 (1). Two of the most cited papers in our top 100 focused on the controversial issue of vertebroplasty versus conservative treatment for vertebral compression fractures. These papers have had dramatic impact on the rate of vertebroplasty, which has decreased since these papers were published in 2009 according to two large database studies (12,13).

Our study has various limitations. Primarily, we used the total number of citations as the main surrogate for article importance. This is a controversial measure for the impact a paper has as it can over- or underestimate the true impact of a study (14,15). An additional limitation of our study involves the discussion of spinal fracture classification systems. We could not thoroughly discuss the most common classification systems such as AO, Magerl, and McAfee as the publications within our list were from a large timespan over which classification systems changed significantly due to the increased use of imaging techniques such as CT and MRI. Lastly, older publications have a greater opportunity to be cited more frequently; hence, these publications may be inaccurately viewed to be of greater importance. To account for this limitation, we included a separate list of the most important citations of the last ten years with average citations per year since publication to eliminate this bias. Furthermore, it is evident from our overall list that the oldest papers were, in fact, not the most cited papers.

Conclusions

This study is the first to identify the most cited spine fracture papers. It provides insight that the 2000s contain the plurality of the influential publications indicating how some of the most important changes to spine fracture management pertain to better imaging modalities and surgical technologies. Knowledge of the influential publications can serve as a guide for a comprehensive understanding of the historical and current literature pertaining to spine fractures.

Acknowledgements

None.

Footnote

Conflicts of Interest: The authors have no conflicts of interest to declare.

References

1. Burge R, Dawson-Hughes B, Solomon DH, et al. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Miner Res 2007;22:465-75. [Crossref] [PubMed]
2. Ellegaard O, Wallin JA. The bibliometric analysis of scholarly production: How great is the impact? Scientometrics 2015;105:1809-31. [Crossref] [PubMed]
3. Steinberger J, Skovrlj B, Caridi JM, et al. The top 100 classic papers in lumbar spine surgery. Spine (Phila Pa 1976) 2015;40:740-7. [Crossref] [PubMed]
4. Eshraghi A, Osman NA, Gholizadeh H, et al. 100 top-cited scientific papers in limb prosthetics. Biomed Eng Online 2013;12:119. [Crossref] [PubMed]
5. Skovrlj B, Steinberger J, Guzman JZ, et al. The 100 Most Influential Articles in Cervical Spine Surgery. Global Spine J 2016;6:69-79. [Crossref] [PubMed]
6. De la Garza-Ramos R, Benvenutti-Regato M, Caro-Osorio E. The 100 most-cited articles in spinal oncology. J Neurosurg Spine 2016;24:810-23. [Crossref] [PubMed]
7. Kelly JC, Glynn RW, O'Briain DE, et al. The 100 classic papers of orthopaedic surgery: a bibliometric analysis. J Bone Joint Surg Br 2010;92:1338-43. [Crossref] [PubMed]
8. Denis F. The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine (Phila Pa 1976) 1983;8:817-31. [Crossref] [PubMed]
9. Buchbinder R, Osborne RH, Ebeling PR, et al. A randomized trial of vertebroplasty for painful osteoporotic vertebral fractures. N Engl J Med 2009;361:557-68. [Crossref] [PubMed]
10. Kallmes DF, Comstock BA, Heagerty PJ, et al. A randomized trial of vertebroplasty for osteoporotic spinal fractures. N Engl J Med 2009;361:569-79. [Crossref] [PubMed]
11. Kondo KL. Osteoporotic vertebral compression fractures and vertebral augmentation. Semin Intervent Radiol 2008;25:413-24. [Crossref] [PubMed]
12. Rosenbaum BP, Kshettry VR, Kelly ML, et al. Trends in Inpatient Vertebroplasty and Kyphoplasty Volume in the United States, 2005-2011: Assessing the Impact of Randomized Controlled Trials. Clin Spine Surg 2017;30:E276-82. [Crossref] [PubMed]
13. Sayari AJ, Liu Y, Cohen JR, et al. Trends in vertebroplasty and kyphoplasty after thoracolumbar osteoporotic fracture: A large database study from 2005 to 2012. J Orthop 2015;12:S217-22. [Crossref] [PubMed]
14. Cheek J, Garnham B, Quan J. What's in a number? Issues in providing evidence of impact and quality of research(ers). Qual Health Res 2006;16:423-35. [Crossref] [PubMed]
15. Smith R. Beware the tyranny of impact factors. J Bone Joint Surg Br 2008;90:125-6. [Crossref] [PubMed]