Cerebrospinal fluid (CSF) leak describes the discharge of CSF from the intracranial cavity through an osseous defect within the skull base. The underlying dura mater and adherent pia-arachnoid mater are disrupted, resulting in a communication between the intracranial cavity, the subarachnoid space and either the nasal or middle ear cavity. The condition was first described as a pathologic entity in 1899 by Clair Thompson.1 While CSF leaks may occur spontaneously, common etiologies of CSF rhinorrhea include trauma, neoplasms, and prior surgery, while CSF otorrhea is usually associated with craniocerebral trauma (e.g. skull fracture involving the temporal bone), neurosurgical procedures, or other conditions.2 Patients with CSF leak can present with a variety of symptoms such as clear nasal discharge and headache or complications such as pneumocephalus, meningitis or brain abscess.

As scientific knowledge has grown in biomedicine, it has also become necessary to develop tools to manage and understand the body of evidence. In that sense, bibliometrics has become a consolidated discipline for analyzing scientific activity, enabling the characterization of a particular field or area of knowledge by means of the quantification of the bibliographic characteristics of scientific publications.2,3 Nowadays, identifying citation classics and top-cited papers is one of the key methodologies used to systematically evaluate research performance. This information can help optimize the allocation of resources, reorient research support, rationalize research organizations, restrict research in particular fields, and augment research productivity.3 Across different fields, the scope of scientific literature has broadened to achieve a more multidisciplinary vision. This expanded focus justifies the review of the most important papers to help guide future research and practice. In that sense, several medical specialties have ranked articles within their fields by citation frequency.4–7

In otorhinolaryngology, several factors have contributed to a growth in research output: the academic development of the field and training programs; significant advances in information and communication technology, which allow studies and experiments to be rapidly performed, written, reviewed, published, and cited; the increase in cooperative practices and multidisciplinary research approaches; the growing need to publish in order to secure academic promotions and research funding; and the existence of a larger critical mass and evidence base, driving further research advances in the field.8,9 Research on the etiology, diagnosis and treatment for CSF leak has also increased over the last few decades, but while several studies have analyzed the top cited documents in otorhinolaryngology,4,10–12 none have specifically focused on CSF leak.

The present study identifies and analyzes the characteristics of the 100 most cited articles on CSF rhinorrhea and otorrhea. This information could help researchers and professionals understand the research areas that are generating the most impact on the field, the countries that are making the largest contribution, and the main journals used to disseminate advances.

We opted to identify documents about CSF leak by means of the Medical Subject Headings (MeSH) thesaurus, a detailed instrument for controlled terminology. The MEDLINE database included the terms “Cerebrospinal fluid rhinorrhea” and “Cerebrospinal fluid otorrhea” in 1966 to define respective CSF discharges through the nose or through the external auditory meatus/eustachian tube into the nasopharynx. However, it was not until 2005 that the database included a composite descriptor for “Cerebrospinal fluid leak” to define any discharge of CSF through a hole in the skull bone.

The study period was limited to 1945–2018. The searches took place on the Clarivate Analytics Web of Science (WoS) platform, which includes MEDLINE database, on January 21, 2019. The WoS Core Collection provides information on the number of times a particular article has been cited by other articles. This strategy yielded a total of 4155 documents from the MEDLINE database. After excluding 25 editorials and 1021 documents that were not indexed in the WoS Core Collection, we had a total of 3109 articles, from which we aimed to identify the 100 most cited papers. In fact, we included 101 articles because the papers ranked 100 and 101 received the same numbers of citations. Data collected for each article included the full reference (author’s names, journal title and publication year), impact factor (in the Journal Citation Report [JCR] 2017), WoS category of the journal, countries of authors, and type of document (article, review, case report, meta-analysis, trial). Citation density was determined by dividing the number of citations received by the position of the journal in the JCR category.

Table 1 lists the full references and citation data for the 101 top-cited articles in the literature on CSF leak. Laryngoscope was the most prolific journal, publishing 20 articles, followed by the Journal of Neurosurgery (n = 17), Neurosurgery (n = 12), Annals of Otology, Rhinology & Laryngology (n = 6), and Otolaryngology-Head and Neck Surgery (n = 6). These 5 journals covered 60.4% of the highest-impact documents (Table 2).

Table 3 lists the JCR categories of the top-cited articles. The leading category was surgery (47.5% of the documents), followed by otorhinolaryngology (46.5%) and clinical neurology (41.6%).

Authors’ country of origin was most commonly the USA (75.2%), followed by the UK (6.9%) and Italy (4%), as shown in Table 4. Most documents were original articles (n = 88), while the other 13 were reviews (n = 12.9). According to the clinical document types assigned in MEDLINE, there were 12 case reports; 4 meta-analyses, and 2 clinical trials.

The articles were published between 1954 and 2018, but scientific activity was concentrated between the two decades of 1990–1999 (n = 30) and 2000–2009 (n = 30). Seven of the top papers were published in both 1997 and 1999, while the years 2004 and 2012 each saw the publication of another six top articles (Table 5).

The MeSH terms for the 101 top documents are listed in Table 6. These were led by “Cerebrospinal fluid rhinorrhea” (n = 81). “Cerebrospinal fluid otorrhea” appeared in 33 documents, while “Surgery” was in 62 documents, “Endoscopy” in 35, “Cerebrospinal fluid” in 9, and “Cerebrospinal fluid shunt” in 7 (Table 6).

Although several methodologies exist for determining the impact of journals and articles, the number of citations and citation rank lists are still the dominant methods used for identifying influential work in areas including neurosurgery, otolaryngology, ophthalmology, and others.13 The analysis of the top-cited articles illustrates how knowledge accumulates over time, therefore this study aimed to determine which articles on CSF leak have had the most influence by ranking the 100 most cited works since 1945. In addition, we analyzed the characteristics of these articles to determine the factors contributing to situating them as the most relevant to other researchers working in the specialty.

The top 101 documents were published in 36 different journals, but half the articles were concentrated in just 5 journals. Otorhinolaryngology and neurosurgery were the dominant specialties, while other disciplines made only nominal contributions to our population of high-impact studies of CSF leak.

Seventy percent of the 101 top-cited articles have been published since 1990, confirming the heavier influence of documents published towards the end of the century. This result is associated with the phenomenon of obsolescence and the concentration of researchers’ interest in more recent studies, as measured through the “half life” of publications, among other citation indicators.14

In that regard, scientific evolution in the field of CSF leak has been driven by improvements in diagnostic, imaging, and surgical techniques for its management. One of the most important advances has undoubtedly been the introduction of sinonasal endoscopy for treating the condition.15 Endoscopy for fistula closure substantially decreased the morbidity associated with the craniotomy approaches used until the 1940s and increased the closure success rate, which until then had not exceeded 60%.16

Since its introduction in the 1970s by Messenklinger and Stammberger,17–19 endoscopic surgery has progressed tremendously. Beyond its adoption as a treatment for sinonasal pathologies, use of the technique has expanded to other areas, and it is now a major tool for treating skull base pathologies.

Wigang15 first discovered endoscopic closure technique in 1981. From then on, the use of this approach spread, and several other authors described their experiences in case series that demonstrated the feasibility and advantages of the method in terms of decreasing morbidity and increasing successful closure.20–22 These experiences probably explain why a large number of the most cited papers we identified were published over 30 years ago (29%): these studies were pioneering or landmark contributions to the field.

This evolution has been the reason for the high scientific productivity over the years, which corresponds to two different periods: from 1990 to 1999, and from 2000 to 2009. In each of these decades, 30 of the 101 top-cited articles were published.

The first period reflects changes in the management of the fistulas, wherein endoscopic sinus surgery is increasingly favored. Over this decade, articles in high-impact journals described authors’ experiences with innovative techniques.2,20,23 This inflection point in the management of the pathology was encapsulated in a meta-analysis published by Hegazy et al. in the year 2000: “Transnasal endoscopic repair of cerebrospinal fluid rhinorrhea: a meta-analysis” in Laryngoscope. In it, the review authors defend and confirm endoscopic surgery as a safe method for CSF fistula closure. This paper had a high impact on the field; the 272 citations it received in the study period make it the second most influential article in our population of documents, with a citation density of 16.24

During the second period of highest production (from 2000 to 2009), the topic attracting the most research interest was the expanded endonasal approach to the skull base. Indeed, several articles were published in relation to endoscopic transphenoidal surgery for skull base lesions.25–27 But an important drawback of this endoscopic approach was the difficulty in reconstructing large dural defects, which often led to complications such as CSF leaks, meningitis or pneumocephalus. For many years, the typical method of closing the dural defects was by means of onlay and inlay grafts, but this technique was associated with very high rates of postoperative leaks.

In this regard, the introduction of the first endonasal pedicle flap, the nasoseptal flap, represented a major impetus to the rapid development and progress of endoscopic skull base surgery. This innovation decreased initial CSF leak rates from 20% to less than 5%, spurring greater expansion of the endoscopic approach.28 Thus, the most cited article over the entire study period was by far “A novel reconstructive technique after endoscopic expanded endonasal approaches: vascular pedicle nasoseptal flap”, by Hadad et al., published in 2006 in Laryngoscope.29 It also ranked as the article with the highest citation density, at 69.7. The 767 citations it received over the following 12 years illustrates the study’s influence on the evolution of the endoscopic skull base surgery, marking a turning point in the history of the CSF leak surgery, endoscopic endonasal surgery, and endoscopic skull base surgery.

Although the most frequent topic of the top-cited articles was endoscopic surgery, other high-impact articles dealt with other subjects, such as clinical and diagnostic features of CSF leak, neurinoma surgery, or the association between CSF leak and temporal bone fractures. The impact of these articles in the literature peaked from the 1970s to the 1990s; however, after that the important evolution of endoscopic skull base surgery eclipsed the publications focusing on CSF fistulas. In any case, most of the articles presented in this study dealt with surgical aspects of CSF leak, reflecting the wider interest in surgical papers compared to diagnostic or clinical studies.

The USA was the largest contributor to CSF leak research. The vast majority (75.2%) of our high-impact publications come from authors and institutions from that country. This finding is consistent with results reported in other surgical fields such as maxillofacial and plastic surgery.30 In addition to the concentration of resources in the USA and its mainstream position in biomedical research, the Matthew effect of accumulated advantage could also be a factor that favors the concentration of citations among journals and authors from this country. Other countries contributing influential papers to the literature include the UK (n = 7), Italy (n = 4), and Canada (n = 3).

The most significant aspect to point out regarding documentary types is the predominance of case studies (35%), with similar values compared to other surgical areas, such as maxillofacial surgery (31%).31 Taken alone, clinical case studies do not provide enough evidence for guiding treatment decisions, but when they are collectively considered, appropriately codified and properly integrated into structured information systems, physicians can use the information gleaned as a solid evidence base for comparing cases and checking diagnoses.

The top cited articles were mostly published in otorhinolaryngological and neurosurgical journals, but as in other areas of knowledge, a few generalist journals also stood out for contributing high-impact articles. This is the case of the New England Journal of Medicine, with one document that received 140 citations and had a citation density of 46.7 and an impact factor of 79.260 (2017 JCR).32 The Lancet also contributed a document to the list, which garnered 81 citations and showed a citation density of 1,8 and an impact factor of 53.254.33

The journal of publication is an important factor determining the potential for an article to be cited.4 General medical journals have a wider audience and larger circulation, so they obtain a higher impact factor than smaller specialist journals. Thus, it is difficult to make individual comparisons between journals or to compare the impact of journals from different thematic categories.

There are several limitations to this type of study. First, although citation analysis is one of the most widely used bibliometric parameters, providing a measure of scientific activity, visibility, use, dissemination, and impact, it does not represent a measure of scientific quality.31 Second, our search was based on journals with impact factors or under tracking for impact factors. This criterion preferentially favored Western articles, especially those from the USA, the UK, and Canada. Most papers in non-English journals were cited by other papers published in the same language. Therefore, we might have missed a number of highly cited articles related to CSF leak. Third, this study was based on objective citation data, but some landmark CSF leak papers may not have figured among the top-cited papers, as they were cited only until their findings became well known. This phenomenon, termed ‘obliteration by incorporation’, has been observed in other fields. Finally, other factors could also affect the citation rates, such as the journal’s year in review, authors’ self-citations, incomplete citing, and omission bias.6 Despite these limitations, citation analysis and impact factor are widely used to rank and evaluate articles and journals. However, these assessment methods should be complemented by others such as peer survey and specialist opinion of citation analysis.

Our findings offer information related to the dissemination of knowledge in recent decades about the cerebrospinal fluid leaks. Two well-defined periods of maximum scientific activity were driven by surgical innovations. This study also shows that the major specialties contributing to the field of CSF leak were otorhinolaryngology and neurosurgery, which were almost equally represented among the most cited documents.

The authors declare no conflicts of interest.