Trends in the Scientific Literature on Sudden Hearing Loss: A Bibliometric Study

Objective: To analyze the scientific publications, from a bibliometric approach, on the sudden hearing loss since the first documents appear (1966) until the year 2017. Methods: The Scopus database was used to carry out the study, and the descriptors "sudden hearing loss", “sudden sensorineural hearing loss" and "sudden sensorial hearing loss" were used, limited to the title, keywords and summary field. The most common bibliometric indicators were applied for the selected publications. Results: A total of 2,509 original articles were recovered between1966-2017. The growth of the publications was adapting better to a linear adjustment (r = 0.84). The time of duplication is 10.4 years. Bradford's core (most productive journals) formed by thirteen journals, some of the most productive are Otology and Neurotology, Acta Oto-Laryngologica and Laryngoscope, with an Impact Factor of 2.024, 2.471 and 1.116, respectively. Regarding geographical distribution, the United States, Japan and Germany present the highest production. Conclusions: The scientific production on the sudden hearing loss follows a linear growth, without evidence that should have reached a saturation point. In this case, the postulates of the Law of growth of Price's science are not fulfilled. In addition, a high index of transience exists.

under 14 years [8] and only 1.2% under 9 years [10]. The global prevalence of this disorder ranges from 5/20 patients per 100,000 inhabitants (data for the United States) (1), and there may be 15 million affected worldwide [11].
The etiological diagnosis of SSNHL is usually controversial [12] and only around 10% of patients diagnosed provide a justifiable cause. Some of the most frequent disorders are infections, labyrinthine hemorrhage, and disorders of the microcirculation, autoimmune diseases, neurological disorders (multiple sclerosis), Menière's disease or idiopathic causes [13,16]. However, the viral etiology is increasingly accepted by many authors. The infection by herpes virus type I or II, varicella and zoster may be behind its etiology [17]. For this reason it is advisable to associate antivirals with standard treatments.
Unfortunately in most cases (between 70 and 90%), and despite an exhaustive search of the causes, none are found [18].
If there are symptoms of SSNHL, such as full ear, dizziness, ringing in the ears or tinnitus, it should be ruled out that the loss is conductive, performing otoscopy to check that there is no obstruction due to fluid or wax, pure tone audiometry and tympanometry. If sudden deafness is diagnosed, other tests will be done to try to determine the cause, such as blood tests, diagnostic imaging tests (magnetic resonance imaging) and balance tests (dynamic posturography and videonystagmography). The otoacoustic emissions as well as vestibular potentials (VEMPS) can also provide us with additional information.
There is no standard treatment protocol, but the most accepted therapeutic regimen is the use of corticosteroids [19,21]. The mechanism of action in the inner ear is not clear, but a higher perilymphatic concentration of corticosteroids is associated with greater recovery of hearing [22]. The treatment has to be focused on increase the irrigation of the inner ear to reduce the inflammation. It has been demonstrated that intratympanic steroid injections are as effective as oral [23,24], and is an option for people who cannot take oral steroids, or who do not obtain the expected results with oral treatment.
Corticosteroids should be used as soon as possible, to obtain the best results [6,7], even before obtaining all the tests, because if it is delayed, the chances of reversing the hearing loss decrease. The SSNHL constitutes a real therapeutic urgency and everything that delays it will be to the detriment of the result.
Another treatment that has proven effective is hyperbaric oxygen therapy when it starts in the first weeks after the symptoms [25,26]. It is thought that SSNHL produces hypoxia of the cochlear zone and this treatment would improve the oxygenation of the external hair cells [27,28]. However, the hyperbaric chamber is only valid if it is used at 48 or 72 hours, maximum after five days. The experience with it is accepted by most authors, but in these conditions. The work we have done is a bibliometric type study. Bibliometry can be defined as the science that studies the nature and course of a discipline (as long as it gives rise to publications), through the computation and analysis of several facets of written communication [29]. Bibliometric analysis includes the collection, processing and management of quantitative bibliographic data from scientific publications [30].
The term bibliometry refers to a group of bibliographic searches and scientific information that allows us to analyze the production and dispersion of a specific topic in a statistically quantitative way [31].
Thanks to the use of statistics, bibliometric indicators allows us to measure the amount of scientific publications made by researchers, and a series of characteristics of them, with which we can analyze scientific literature [32].
The parameters that are used in the evaluation process of any activity can be defined as "indicators". Both production and dispersion are indicators of impact [33]. The production allows us to know if a specific topic is booming or not and which authors write more about the topic and the dispersion shows which magazines publish these articles. The impact factor (IF) is used to measure the visibility and usefulness of the sources, and allows to determine the interest that documents have in terms of their greater or lesser use by different users. The IF, devised by Eugene Garfield [34], is published annually as part of the Journal Citation Report.
This type of bibliometric studies are of great help to know the social and scientific importance of a discipline or a specific field. Our group has studied, with a bibliometric approach, the evolution of scientific literature in different areas of biomedicine [35,44].

Material and Method
the database used in this study, Scopus, is the largest summary and citation database that evaluates scientific literature equally (scientific journals, books and conferences). It covers almost 22,000 titles from more than 5,000 publishers, of which 21,500 are scientific, technical, medical and social science journals (this includes arts and humanities). Scopus is the most complete and easy to use database in the biomedical field in comparison with any other tool used in scientific literature research, it is considered the largest database in the world, and is commonly used by researchers to perform bibliometric analysis [45,46].
Through the use of remote downloading techniques, documents published since 1966 (year in which the first records appear) have been retrieved up to 2017. The search has been limited in the summary field, title and keywords, using the specific descriptors: "sudden hearing loss", "sudden sensorial hearing loss" and "sudden sensorineural hearing loss".
In this study, the most common bibliometric indicators were used: Price Index, doubling time, annual growth rate, Price transience index, Lotka Law and the Bradford Zones.
Among the production indicators we have applied the Price law [31]. This law is the most used indicator in the analysis of the productivity of a specific discipline or a particular country, to reflect a fundamental aspect of scientific production, which is its growth. To assess if the scientific production in the SSNHL follows the Price law, we have created a linear trend curve that is expressed in the following way y = 0.2111x -414. 49 and another for the exponential trend line according to the equation y = 2E -75e 0,086x .
Related to the growth, is the time of duplication and the annual growth rate. The first is the amount of time required by the subject of study to double his production. The form of growth was studied from the equation of Egghe and Ravichandra [47], this function is represented mathematically as: Where C (t) is the total number of documents produced at time t; c and g represent estimated constants of the observed data, taking into account that c> 0, g> 1, and t ≥0; t is the number of chronological years studied in the research period (t = 0, 1, 2, ..., n). The model not only provides an average rate of growth, but also offers a rate of duplication.
Among the bibliometric indicators of dispersion, we have applied the Bradford zones. Bradford evidenced the concentration in a small number of journals the majority percentage of the pertinent bibliography in a subject, which implies a rapid fall in the performance of expanding the search for references outside this nucleus [48].
To know the influence of publications, the Impact Factor (IF) is used. This indicator is calculated taking into account the times a journal has been cited in the Science Citation Index database in the last two years and the total number of articles published in this journal in those same years. The list of scientific journals of the Journal Citation Report, divided into law establishes that in the entire scientific community, the number (A) of authors who have published a number (n) of works, in the course of a period of several years of activity, i.e. A (n) authors, is equal to the number of those who have published a single work, A (1), in the same period of time, divided by the square of n. In mathematical terms the original law is expressed by the formula: n 2 It is also interesting to know the number of authors with only one work. It is what is known as a transience or price index. Its calculation is given as the percentage of the quotient of the authors with a single job among the total. Mathematically it would be expressed: different areas, attributes to each one an IF and establishes its prestige range [49]. distribution of authors based on the number of articles published [50], also known as the "inverse quadratic of scientific production". It studies the publication volume of the authors and states that the number that publishes few works is greater than the amount published by many. This Through the search criteria used, 2,509 records (articles, magazines, letters to the director, etc.) have been recovered, during the period spanning from 1966, the year of the first registrations, until 2017. The chronological distribution is illustrated in the Fig. 1.
To determine whether growth follows Price's exponential growth law, we have ordered the data in a linear fashion according to the formula y = 2.8685x -5664.4 and another that fits the exponential curve according to the formula y = 8E-83e 0.0965x . In this way, the value of the coefficient of determination is greater in the linear function, with a value of 0.83, which shows the power of representation compared to the exponential which has a value of 0.78. With these data we can conclude that the analyzed repertoire is better adapted to a linear adjustment than to an exponential one and that the postulates of the Price Law, in this case, are not fulfilled. Table 1 shows the parameters and values obtained from the application of the exponential model by the non-linear regression method.  From this method it is inferred that the literature on SSNHL grows at a rate of 6.9% per year with a production that doubles its size every 10.4 years. We have obtained that the model is explained at 95.6% (Supplementary Figure).
As illustrated in Figure 2, the period between the years 2017-2013 is the one that contains the largest number of documents with more than 30% of the total of them. The most productive countries are the US, Japan and Germany. These 3 countries account for more than 40% of total production (Fig. 3). As regards the scientific journals in which the articles on sudden deafness have been published, the Bradford model was applied. The average number of articles per Bradford area was 627.  Table 2A. Distribution of journals in the Bradford Zones. The core is composed of 7 journals, 1.16 of them, with more than 25% of the articles published. Table 2B shows the IF data (impact factor) of the most productive journals, according to data from JCR 2017, as well as other data of interest.   Table 3A. Classification of authors based on productivity. If we continue with the analysis of the authors, we have been able to highlight, as illustrated in Table 3B, that 6 authors, among the 10 most productive, are Japanese. Another interesting fact is the number of signatures of the articles, a fact that indicates the degree of collaboration of the authors. In this study we have verified that the most common co-authorship index is of 3 firms and that the article with the most collaboration has 31 signatures (Supplementary Table).

Number of documents
*IP: Participation Index Table 3B. Most productive authors.

Supplementary Table.
Signing authors per document.

Discussion
Bibliometric studies have become essential tools for evaluating scientific activity, allowing an overview of the growth, size and distribution of scientific literature associated with a particular discipline [33,51,53].
However, despite the benefits of this sociometric approach, its limitations must also be taken into account. For example, bibliometric studies do not take into account neither the quality of the publications nor the fact that the results of the scientific activity are only measured by publications. The aspects that are not considered include teaching, applied research or scientific dissemination [54,55]. However, these studies are an effective complement to the opinions and judgments of experts in each field, and are useful and objective tools to evaluate the results of scientific activity, offering a more realistic view of the general situation and an indication of trends, as well as a prediction of how the subject of analysis could evolve [56].
The results obtained, from this bibliometric analysis, allow us to understand the development of the scientific literature in recent decades on sudden deafness. In this regard, it should be noted, as shown in Fig. 1, that the number of scientific publications has experienced a linear growth, in the last 52 years, and that, therefore, Price's Law of Growth of Science is not fulfilled, theory of the expansion of scientific literature [36]. We can highlight the years 2015, 2013 and 2016 as the most productive ( Fig.  1) and the period 2017-2013 which contains the largest number of publications (Fig. 2). The increase in documents is due both to studies that attempt to analyze the prevalence of this pathology [1, 8, and 11] and studies about its treatment. In 2013, we found a study on the incidence of sudden deafness in the United States [57], where they sought to link SSNHL with age, ethnicity, gender..., demonstrating a relationship with age and an increase in patients over the years, both in young people and in adults. In 2017 these data are updated [58], showing a slightly higher prevalence in women, where the male-female ratio stands at 1:1.35. Similarly, it was observed, in relation to age, that it is higher in people over 55 years of age and finally, a relationship with climate was found, that is, in autumn (when the temperature is colder) there are more cases of sudden deafness.

Signing authors
Due to the increase in patients that is occurring annually, the need arises to investigate to find effective treatments. In this sense, in recent years there has been an increase in the number of publications dealing with corticosteroids [59], intratympanic injections [60] and hyperbaric oxygen therapy [25,26]. These studies demonstrate the efficacy against sudden deafness, either in combination therapy [61] or individually [62][63][64], reaching the conclusion that early and successful treatment obtains very favorable results [6,7].
The most productive countries on sudden deafness are the USA, Japan and Germany. The United States being the country with the most documents, however, one institution in Germany is the most productive, Ludwig-Maximilians-Universitat Munchen. Among the ten most productive institutions, we only find two from the US, Harvard Medical School and Massachusetts Eye and Ear Infirmary.
The transience index is quite high, with a value of 84% which indicates that most authors have written on the subject occasionally. The medium producers make up 15.86% while the great authors only represent 0.07%.
In this sense we find that 10 authors of the 7,028, who make up the study, are responsible for a high percentage of these works, more than 7%. Among these authors we find five Japanese, T. Nakashima Table 4. Classification of authors based on productivity.
The interest of such important researchers demonstrates the relevance of the subject of study. Another fact, which reinforces this argument, is that the time of duplication of the scientific literature is only 10.4 years.
To analyze the quality of the articles, the impact and excellence indicators of the publications on the subject in question were used. The fact that prestigious journals such as Laryngoscope (IF = 2.471), Otolaryngology -Head and Neck Surgery (IF = 2.440) or Otology and Neurotology (IF = 2.024) will publish articles on SSNHL is an important fact that indicates the relevance (both clinical and social) that this pathology has acquired in recent years.
Finally, and to conclude, it can be affirmed that, despite the limitations of bibliometric studies, and thanks to the design of this study, we have been able to offer an image of the representativeness and evolution of sudden deafness, observing quality parameters and diffusion more commonly used at international level. Research in this field will probably continue to grow, due to the need presented by an aging population.

Conclusions
By way of conclusion, it can be asserted that, despite the limitations characteristic of bibliometric studies, and thanks to the design of this study, we have been able to offer a picture of the representativeness and evolution of international research on SSHNL. Research in this field will possibly continue to grow in the coming years due to the continuous ageing of populations and associated pathologies.

Conflicting interests
The authors declare that there is no conflict of interest.

Funding
This research received no grant from any funding agency in the public, commercial, or not-for-profit sectors.