In this study, we used bibliometric analysis to produce the top-cited articles on Chiari malformations since 1900. Using this data, we reviewed how our knowledge of this condition has evolved over the past century, both in terms of diagnosis and treatment.
Published in 1938, the first high-impact work of the 20th century presents a case of Chiari which was managed surgically and adds suggestions for future operations. At the time of this article’s publication, Chiari was not considered a surgical problem; instead, it was a rare finding to be reported in literature. With their groundbreaking work, Penfield and Coburn introduced Chiari into the world of neurosurgery; however, key publications were limited in the decades that followed, and the majority focused on anatomic features and morphogenesis rather than treatment. Though no frequently cited papers were published in the 1940s, there was a small uptick in the next two decades, with 8 articles published between 1950 and 1969. These numbers steadily increased to 7 papers in the 1970s and 9 in the 1980s. The bulk of high impact Chiari research came in between 1990 and 2009, years which contributed a total of 69 top-cited papers. It was during this time that there was a shift in the content of high impact publications, with surgical techniques flooding the literature.
After the influential Milhorat publication in 1999, which described radiographic and clinical findings in hundreds of patients with symptomatic Chiari I, there was an outpouring of high-impact papers. This paper, aided by several landmark publications describing the possible pathophysiology of Chiari[5; 11; 12], allowed for a new area of focus. With the characterization of Chiari malformations established in the neurosurgical community, articles focusing on treatment options came to the forefront. Between the years 2000 and 2009, over a third of the top 100 papers were published, the majority of which discussed surgical techniques and clinical outcomes.
Given the limitation that older studies have in the realm of total citations, we also focused on the mean number of citations each study received per year. This provided insight into articles which were consistently relevant over time and brought newer, less-cited publications to our attention. Though the two most cited overall articles are also the two most cited per year, the lists differ from here, with more recently published papers tending to appear higher on the most cited per year list (Table 2) than the top cited overall list (Table 1). For example, when looking at the top 10 for each category, the earliest published content for citations per year is from 1994, with 6 being published during or after the year 2000. The top 10 for citations overall includes only 2 papers from the 2000s. Among the top cited overall, the most common topic was the pathophysiology of Chiari with six papers; the other four publications focused on clinical descriptions with or without radiologic supplementation. In the top cited per year category, 3 surgical experience papers took the place of pathophysiology and clinical feature publications. This demonstrates the important content shift which took place at the turn of the century.
Among the top 100 publications, the majority presented the clinical features of Chiari malformations. A subset of these discussed management of patients with Chiari, with an emphasis on imaging becoming apparent over the last few decades. Starting in the 1990s, some papers started addressing Chiari I in the pediatric population, acknowledging that the clinical features and subsequent management differed for younger patients. Articles focused on surgical techniques and outcomes, most of which appeared in the late 1990s-2000s, made up a substantial portion of the top 100. These publications were assisted by numerous studies devoted the pathophysiology of Chiari, which both contributed to the clinical picture of the disorder and introduced new possible targets for surgical management.
The highest total citation count belonged to Milhourat et al., a publication which still receives an average of 32.9 citations per year, and is currently sitting at 724 citations overall. A landmark article, it established a clinical picture of Chiari I using a prospective cohort of 364 patients who presented with symptoms. The authors addressed the pathophysiology of Chiari I, using MRI data from their participants to support the theory that the malformation arises from a defect of the para-axial mesoderm, and demonstrated a possible genetic component to the disorder using familial aggregation studies. They also determined that a volumetrically diminished posterior cranial fossa was the most consistent finding among their cohort. The most common clinical symptoms fit with this finding, most of them related to overcrowding of the hindbrain—headaches, visual changes, spinal cord disturbances, etc. In addition to being a groundbreaking piece of work in 1999, this paper holds its clinical relevance today as modern imaging techniques are resulting in higher rates of detection.
The next most cited publication, both in total citations and citation frequency, was the work of Oldfield et al., which is cited an average of 19.8 times per year and has been cited 535 times overall. Published 5 years earlier in 1994, this article delves into the pathophysiology of the syringomyelia that can develop with Chiari I. Using 7 cases, the authors proposed a new mechanism, based on the pressure acting on the surface of spinal cord rather than pressure from within the cord, as previous theories suggested. The next most cited publications followed in 1997 and 1999 with similar content, studying the pathogenesis of Chiari itself and the syringomyelia associated with it, respectively. These papers, both of which are hold a place on the top 10 cited per year, not only contributed to a better understanding of this disorder but continue to help develop treatment options for patients today.
The case report was the type of publication observed most frequently in our analysis, closely followed by retrospective studies. Despite the significant presence of Chiari in the literature over the past few centuries, review articles breaking into the top 100 were relatively limited, with only 5 total meta-analyses or literature reviews. The University of California system contributed the most articles to the top-100, with 8 total. This was followed by the Children’s Hospital of Philadelphia, Harvard University, and the University of Pennsylvania, which each contributed 5 papers. The United States published the majority of the top-100 articles, many of which were published either in Neurosurgery or Journal of Neurosurgery.
Studies which use bibliometric analyses are predisposed to several limitations. The Web of Science results only went as far back as 1900, potentially excluding the work of the earliest Chiari publications, such as those of Hans Chiari in the 1800s. Additionally, the classification of Chiari vs. syringomyelia was unclear for many decades, leaving the possibility that some influential works were excluded for semantic reasons. Our study was limited to English articles, which could have omitted high-impact articles published in other languages. Another limitation lies in the use of citations as a unit of measurement for impact—self-citation as well as omission bias, or the exclusion of articles which do not support a given hypothesis, produce some amount of error with this system. Given the substantial history of Chiari in the world of neurological research, it is possible that impactful papers from the early 1900s are no longer cited if the information they initially presented has become common knowledge. It is also more challenging to obtain accurate citation data on these studies.