Gender analysis of the top classic papers in otolaryngology head and neck surgery
Abstract
Objective
The aim of this study was to identify and analyze the gender breakdown of first authorship contributing to the most-cited papers in the field of otolaryngology, with a goal of identifying trends in gender representation in publishing.
Methods
The top 150 most-cited papers were identified using the Science Citation Index of the Institute for Scientific Information. Among the first authors, gender, h-index, percentage of first, last, and corresponding authorship positions, total publications, and citations were analyzed.
Results
The majority of papers were in the English language, from the United States, of clinical nature, and on otologic topics. Eighty-one percent of papers (n = 122) had men who were first authors, although there was no difference in h-index score, authorship position, number of publications, citations, and average citations/year between men and women first authors. Upon subgroup analysis by decade (1950s–2010s), there was no difference in the number of articles by women first authors (P = 0.11); however, there was a statistically significant increase in the percentage of women authors (P = 0.001) in papers published later compared to those published earlier.
Conclusions
While a promising number of women otolaryngologists are publishing high-powered articles, future initiatives to promote academic inclusivity of women should be considered.
INTRODUCTION
Over the years, the field of otolaryngology has rapidly expanded due to technological advancements and groundbreaking discoveries. This has in turn led to an increased number of scientific innovations and publications. The measure of scientific work is often reflected in the number of peer citations an article receives, which demonstrates the gravity and relevance of an article's contribution. While citation frequency does not always correlate with quality, bibliometric analysis is a popular method to assess the impact of articles and scientific journals. Several surgical specialties, such as plastic,1, 2 orthopedic,3 general,4 and neurologic surgery,5 have utilized this strategy to determine the most frequently cited articles in their domain.
An objective indicator of scholastic contribution can be measured by authorship order—a system designed to reward intellectual contribution while reflecting the hierarchical structure of the research community. The most prestigious positions are first and last authorship, which are responsible for the primary production and supervision of the paper, respectively. While the number of women first authors has increased significantly over the past decades, there continues to be persistent underrepresentation of women authors in high-profile manuscripts.6, 7 Even among coauthors who contributed equally to a manuscript, more men authors were in the first position compared to their women colleagues.8 Since the number of publications and peer citations factor into academic promotion, assessing authorship of the most impactful articles within a field can improve understanding of gender inequality.
The aim of this study was to identify the top 150 most-cited articles in the field of otolaryngology over the past 75 years and to analyze trends in articles, journals, and first authorship. Determining the characteristics of the seminal articles in otolaryngology can help identify and assess the influential contributors to the field while guiding future research pursuits.
MATERIALS AND METHODS
Search strategy
The Web of Science Journal Citation Reports (Clarivate Analytics) was used to determine the top 20 otolaryngology journals sorted by impact factor. The journals were then individually searched using the Science Citation Index of the Institute for Scientific Information (Thompson Reuters) database to identify the most-cited articles from 1945 to 2020 in these publications. Inclusion criteria included all types of articles without any restrictions on language, human or animal subject matter, or date of publication.
Data analysis
Data extraction and level of evidence according to Ackley et al.9 were performed independently by two authors and any disagreement was resolved by a third author.10 Total citations, as well as citation density (defined as total citations/number of years since publication), were recorded per article. Within the top-cited articles, authorship, country of origin based on first author address, year of publication, institution, journal name, research nature (basic science, clinical research), and subspecialty were extracted. First author characteristics including h-index, percentage of first and last author position, total publications, times cited with and without self, citing articles with and without self, average citations/publication, and citations/year were analyzed. The percentage of women authors per manuscript was calculated and analyzed based on the type of publication (i.e., basic science, prospective cohort, retrospective review, review articles, clinical practice guidelines). A subgroup analysis by decade (1950–1959, 1960–1969, 1970–1979, 1980–1989, 1990–1999, 2000–2009, 2010–2019) assessing first authorship, percentage of women authors, and overall citation density was conducted to examine overall trends.
Statistical analysis
All descriptive statistics were performed using RStudio (RStudio Inc.). The Pearson's correlation coefficient was employed to test for correlation between nonparametric variables. The independent-samples t test was used to assess first author characteristics based on gender and research nature. Differences in total citation based on article characteristics (country of origin, year of publication, subspecialty) were analyzed using an independent-samples t test or one-way analysis of variance (ANOVA) accordingly. Subgroup analysis was conducted with Fisher's exact test for categorical variables or one-way ANOVA for continuous variables with post hoc Tukey's honest significant difference when appropriate. P < 0.05 were considered significant.
RESULTS
Article characteristics
The top 150 otolaryngology articles by total citations are listed in Supporting Information: Table S1. Six of the top 150 otolaryngology papers have been cited over 1000 times. The most-cited article was Facial-Nerve Grading System by House and Brackmann, which had a total of 2178 citations, whereas the least cited articles both had 258 citations each.11, 12 The most recent article in the top 150 was the International Consensus Statement on Allergy and Rhinology: Rhinosinusitis, which has been cited 364 times to date.13 Laryngoscope was the journal with the highest number of articles in the top 150 (39 articles, 26.0%) (Table 1). The average number of citations was 445 ± 277 (258–2178) and the average citation density was 25.4 ± 26.5 (4.5–253.3), with European Position Paper on Rhinosinusitis and Nasal Polyps 2012 by Epperson et al.14 generating the highest citation density (Supporting Information: Table S1). There was no correlation between the journal impact factor and the number of articles among the top 150 most-cited (R2 = 0.009) (Table 1).
Rank | Full journal title | Number of records | Number of top 150 articles |
---|---|---|---|
1 | JAMA Otolaryngology—Head & Neck Surgery | 1282 | 1 |
2 | Hearing Research | 6681 | 24 |
3 | Ear and Hearing | 2974 | 19 |
4 | Rhinology | 3018 | 7 |
5 | Journal of Vestibular Research: Equilibrium and Orientation | 833 | 3 |
6 | Journal of the Association for Research in Otolaryngology | 900 | 3 |
7 | International Forum of Allergy & Rhinology | 1453 | 1 |
8 | Dysphagia | 2608 | 6 |
9 | Head and Neck | 1397 | 3 |
10 | Laryngoscope | 30,138 | 39 |
11 | Otolaryngology—Head and Neck Surgery | 11,721 | 28 |
12 | Trends in Hearing | 295 | 0 |
13 | Clinical Otolaryngology | 4246 | 4 |
14 | Clinical and Experimental Otorhinolaryngology | 614 | 0 |
15 | American Journal of Rhinology & Allergy | 1348 | 0 |
16 | Journal of Voice | 3098 | 2 |
17 | Journal of Otolaryngology—Head & Neck Surgery | 995 | 0 |
18 | International Journal of Audiology | 1959 | 5 |
19 | European Archives of Oto Rhino Laryngology | 7023 | 4 |
20 | Otolaryngologic Clinics of North America | 3542 | 1 |
Demographics
The majority of papers were from the United States (n = 99, 66.0%) with the top contributors being Massachusetts Eye and Ear Infirmary (n = 9) and University of Pennsylvania (n = 6) (Table 2). Most of the published papers were from the 2000s (n = 54, 36.0%) followed closely by the 1990s (n = 43, 28.7%). The most commonly represented subspecialty was otology (n = 76, 50.7%), followed by rhinology (n = 37, 24.7%). Of the top 150 articles, 28 were basic science (26 otology, 2 rhinology). The average level of evidence was 4.7 ± 1.6 (1–7), with seven Level 1 papers (4 otology, 1 head & neck, 1 laryngology, 1 general). Types of publication included reviews (n = 28, 18.7%), clinical practice guidelines/expert opinion (n = 24, 16.0%), prospective cohort studies (n = 34, 22.7%), retrospective reviews (n = 13, 8.7%), and basic science studies (n = 28, 18.7%). There were no differences in total citation based on country of origin, first author gender, year of publication, research nature, and subspecialty.
Characteristic | Number (%) | Total citations (mean ± SD, range) | P value |
---|---|---|---|
Country | 0.27 | ||
Argentina | 1 (0.7) | 914 | |
Australia | 3 (2.0) | 338 ± 90 (267–439) | |
Austria | 3 (2.0) | 363 ± 46 (317–409) | |
Belgium | 2 (1.3) | 334 ± 51 (298–370) | |
Canada | 7 (4.7) | 318 ± 80 (261–456) | |
Finland | 2 (1.3) | 442 ± 68 (394–490) | |
France | 1 (0.7) | 289 | |
Germany | 7 (4.7) | 435 ± 223 (258–862) | |
Ireland | 1 (0.7) | 422 | |
Italy | 1 (0.7) | 259 | |
Japan | 1 (0.7) | 275 | |
Netherlands | 4 (2.7) | 862 ± 783 (331–2026) | |
Spain | 1 (0.7) | 311 | |
Sweden | 2 (1.3) | 305 ± 1 (304–306) | |
Turkey | 1 (0.7) | 287 | |
United Kingdom | 14 (9.3) | 553 ± 374 (281–1638) | |
United States | 99 (66.0) | 438 ± 246 (258–2178) | |
First author gender | 0.56 | ||
Men | 122 (81.3) | 439 ± 252 (258–2178) | |
Women | 28 (18.7) | 473 ± 369 (261–2026) | |
Publication date | 0.52 | ||
1950s | 1 (0.7) | 323 | |
1960s | 2 (1.3) | 362 ± 47 (328–395) | |
1970s | 5 (3.3) | 361 ± 80 (293–489) | |
1980s | 30 (20.0) | 437 ± 348 (293–489) | |
1990s | 43 (28.7) | 506 ± 291 (258–1638) | |
2000s | 54 (36.0) | 398 ± 154 (259–914) | |
2010s | 15 (10.0) | 506 ± 435 (283–2026) | |
Research type | 0.25 | ||
Basic | 28 (18.7) | 391 ± 260 (258–1638) | |
Clinical | 122 (81.0) | 458 ± 281 (258–2178) | |
Subspecialty | 0.60 | ||
General | 4 (2.7) | 339 ± 111 (265–505) | |
Head & Neck | 11 (7.3) | 391 ± 130 (258–731) | |
Laryngology | 19 (12.7) | 456 ± 229 (261–1111) | |
Otology | 76 (50.7) | 423 ± 288 (258–2178) | |
Rhinology | 37 (24.7) | 512 ± 317 (260–2026) | |
Sleep | 3 (2.0) | 478 ± 261 (297–777) |
- Abbreviation: SD, standard deviation.
Authorship characteristics
The average number of authors was 5.1 ± 7.7 (1–75). There was no correlation between the number of authors and the number of citations (R2 = 0.009) or year of publication (R2 = 0.17). The average percentage of women authors was 18.4 ± 26.0% (0%–100%). One hundred and twenty-two papers (81.3%) had a man as the first author, while 28 papers (18.7%) had a woman as the first author. The most commonly represented subspecialties with a woman as the first author were otology (11 papers, 39.3%), rhinology (8 papers, 28.6%), and laryngology (6 papers, 21.4%). There was no difference in h-index score, authorship position, number of publications, and citations between men and women as first authors (Table 3).
Characteristic | Men (mean ± SD, range) | Women (mean ± SD, range) | P value |
---|---|---|---|
H-index | 35 ± 20 (1–96) | 32 ± 17 (3–63) | 0.42 |
First author (%) | 36 ± 19 (7–100) | 37 ± 19 (20–100) | 0.84 |
Last author (%) | 31 ± 15 (0–100) | 25 ± 14 (0–45) | 0.06 |
Corresponding author (%) | 28 ± 16 (0–100) | 33 ± 19 (0–72) | 0.20 |
Publications | 146 ± 137 (1–999) | 122 ± 122 (3–379) | 0.39 |
Times cited | 6399 ± 6164 (4–34,232) | 5984 ± 6111 (374–20,059) | 0.75 |
Times cited without self | 5980 ± 5699 (4–32,556) | 5670 ± 5760 (374–19,153) | 0.80 |
Citing articles | 4206 ± 3653 (4–23,890) | 3993 ± 4006 (374–13,396) | 0.78 |
Citing articles without self | 4153 ± 3568 (4–23,437) | 3914 ± 3925 (374–13,171) | 0.75 |
Average citation/publication | 52 ± 51 (1–442) | 53 ± 21 (22–125) | 0.88 |
Average citations/year | 161 ± 146 (2–733) | 163 ± 144 (7–489) | 0.95 |
- Abbreviation: SD, standard deviation.
There was not a significant effect of type of publication on the average percentage of women authors for reviews, retrospective reviews, clinical practice guidelines, cohort studies, and basic science studies, F(4, 122) = 0.97, P = 0.43. When conducting the subgroup analysis by decade, there was no difference in women first authorship (P = 0.11), but there was a statistically significant difference in the percentage of women authors, F(6, 143) = 3.87, P = 0.001 (Table 4). There was also a significant effect of decade on citation density, F(6, 143) = 17.4, P < 0.001, with post hoc comparisons indicating that the mean citation density for articles published in the 1950s (n = 1, 4.5), 1960s (M = 6.5, SD = 0.3), 1970s (M = 8.3, SD = 1.2), 1980s (M = 12.4, SD = 9.9), 1990s (M = 20.0, SD = 11.0), and 2000s (M = 26.2, SD = 12.1) significantly differed from the mean citation density for articles published in the 2010s (M = 73.8, SD = 56.7). In addition to the 2010s, the mean citation densities for articles published in the 1950s, 1960s, and 1970s also significantly differed from those of articles published in the 2000s.
Decade | Total, N (%) | Women first authors, N (%) | Women authors (% mean ± SD, range) | Citation density (mean ± SD, range) |
---|---|---|---|---|
1950–1959 | 1 (0.7) | 0 (0) | 0 | 4.5 |
1960–1969 | 2 (1.3) | 0 (0) | 0 | 6.5 ± 0.3 (6.3–6.7) |
1970–1979 | 5 (3.3) | 1 (20.0) | 10.0 ± 22.4 (0–50) | 8.3 ± 1.2 (7.0–9.8) |
1980–1989 | 30 (20.0) | 1 (3.3) | 3.2 ± 10.5 (0–50) | 12.4 ± 9.9 (6.75–62.2) |
1990–1999 | 43 (28.7) | 8 (18.6) | 18.4 ± 26.4 (0–100) | 20.0 ± 11.0 (9.4–54.6) |
2000–2009 | 54 (36.0) | 12 (22.2) | 24.2 ± 28.0 (0–100) | 26.2 ± 12.1 (13.0–66.3) |
2010–2019 | 15 (10.0) | 6 (40.0) | 34.6 ± 27.5 (0–100) | 73.8 ± 56.7 (32.4–253.3) |
- Abbreviation: SD, standard deviation.
DISCUSSION
Bibliometric analysis is commonly used to determine the most impactful papers in a scientific discipline. While citations do not always correlate with the quality of work and may be dependent on external factors, academic productivity is an objective measure of influence and contribution to the field. In this study, we describe the top 150 papers in otolaryngology. The majority were from English-speaking countries, of clinical nature, and in the fields of otology and rhinology. In the subgroup analysis by decade, there was no difference in female first authorship, but there was a significant increase in the percentage of female authorship. This analysis provides valuable information on the influential authors and topics in otolaryngology while identifying which qualities make an article important to the community.
According to 2019 data from the Association of American Medical Colleges, women now comprise the majority of enrolled US medical students for the first time ever, yet there continues to be disproportionate underrepresentation of women in otolaryngology within residency and fellowship positions.15-17 Among women medical students, gender discrimination and perceived incompatible lifestyle have all been cited as barriers to pursuing a surgical career.18 Career progression in women surgical residents has largely been attributed to organizational factors, such as rigid program culture, work–family conflicts, and lack of mentorship, although younger otolaryngologists reported feeling more supported by their departments.19, 20 As of 2020, 18% of all practicing otolaryngologists in the United States are women.21 The number of full professor positions nearly reflects this demographic (85/531, 16.0%), although women continue to be underrepresented in department chair positions (3/87, 3.4%).22
In regard to high impact research, 28 (18.7%) of the papers on the top 150 list were written by women, a percentage that closely mirrors an 18% women otolaryngology workforce. When comparing articles by a decade of publication, papers published in the 2010s had the highest percentage (n = 6, 40.0%) of women as first authors, although this did not ultimately differ from the previous decades. Within academics, the field of otolaryngology has grown significantly, and over the course of the last decade alone, the number of women residents and fellows expanded from 365 (28%) to 623 (38%), while the total number of female otolaryngologists increased from 1034 (11%) to 1784 (18%).23 Compared to other surgical specialties over the past five decades, otolaryngology is the third fastest-growing field for women surgeons next to obstetrics and gynecology and ophthalmology.24 As of 2020, 36% of all otolaryngology academic faculty are women, an encouraging percentage that may lead to increasing numbers of women first-authored papers in the future. While the majority of the top 150 most-cited articles are authored by men, these findings demonstrate a decreased gender gap and may reflect a time lag of the changing demographic and increase in women leadership. A growing emphasis on publications for matching into otolaryngology residency and for career advancement may also play a role in the increased academic output of women.
There are valuable implications for the results of this study from an international perspective. In England, nearly a fifth of all surgeons happens to be women. Otolaryngology is increasingly well represented among those in training, with women making up 43% of higher specialty trainees in the field. However, only 12% of consultant surgeons are women, suggesting that attaining senior leadership may be challenging for a number of reasons.25 Similarly, the proportion of women in otolaryngology residency training programs in Canada has effectively doubled from 20% to 40% between 1990–1994 and 2010–2014.26 Although the international growth in women trainees is promising and may result in greater representation in leadership roles, striving to improve gender diversity still remains an important goal. Increased recruitment for national/international societies and journal editorial boards can help younger trainees advance in their academic careers.
In specialties historically dominated by men, a slowly rising trend of women first authorship has been observed, although the overall percentage continues to be modest. In three high-impact orthopedic journals, the percentage of women first authors increased from 11% in 2006 to 17% in 2017, corresponding to a 14% women workforce in 2017.27 Among the top neurosurgical journals, the proportion of women first authorship significantly increased from 2003 to 2018 (12% to 16.5%, respectively) with women representing 17.5% of all neurosurgery residents.28 While these results may reflect the percentage of women in the field, a discrepancy still exists, particularly in high-impact surgical journals and among senior author positions.29
Our study indicates that the latter trend also exists in otolaryngology. Although both the proportion of women in the field and female authorship has increased, female first authorship has not significantly increased. Similar trends have been reported across multiple medical specialties, warranting further exploration of how increases in the first authorship may translate to senior authorship in the future, and thereby imply career advancement.30, 31
Classically, the first authorship role indicates substantial inception and creation of the project idea as well as the data analysis, extraction, and manuscript writing. In scenarios where multiple authors have relatively equal amounts of involvement, situations of gender inequality may arise. Given the hierarchical nature of medical training, slight differences in rank may lead to relegation to a lesser authorship position for women who contributed more than their male counterparts.8
It is important to consider other underlying challenges that may contribute to differences in academic productivity. Landmark papers that change the scope of practice are uncommon. If the appropriate opportunity arises, efforts to provide equal resources and opportunities to promote research productivity should be a focus for surgical departments. During the peer review process, standard removal of institution and author names can reduce sources of bias. Institutions should make a concerted effort to encourage women trainees and junior faculty to participate in networking opportunities and specialty societies. Recent initiatives emphasizing salary equality, mentorship provision, and a diverse climate have been successful in increasing the percentage of women clinical faculty while reducing disparity in salary.32, 33
Analyzing the most-cited articles in a field allows readers to identify the most influential landmark papers, revealing insight into the innovative advances and historical developments of the specialty. This analysis demonstrates that the most-cited papers are published in the English language, originating from the United States, in journals with high impact factors, and with an emphasis on clinical research. Interestingly, every article in this study's top 150 list was considered a “citation classic,” originally defined as papers receiving at least 100 citations by Garfield in 1977.34 While the citation threshold value is arbitrary, citation classics identify important scientific contributions that push forward knowledge and guide future research pursuits. The least cited article still received over 200 citations, demonstrating widespread readership and accessibility in the field, which will only continue to progress in the future. One study reports an 11-fold increase in the number of citation classics in otolaryngology from 2002 to 2014 with a decrease in time from publication to citation classic status.35 This trend is largely attributed to technological advances and to improvements in communication and accessibility. This in turn leads to increased speed of writing, researching, referencing, and publishing. Continuing to analyze the most influential papers, therefore, is necessary to provide reference to the most impactful pieces while drawing inspiration for new directions of investigation.
There are several limitations to this study. First, a time lag puts recently published articles at a disadvantage as the older an article is, the greater the likelihood of citation. The paper with the highest citation density received the second most total number of citations. Second, the literature search was conducted within otolaryngology-specific journals, which may lead to exclusion of influential articles published in journals of other disciplines, such as Plastic and Reconstructive Surgery. However, such a search strategy guaranteed identification of the most-cited papers in otolaryngology, rather than another specialty. Third, other biases that may occur include language bias toward the English language, omission bias by purposefully not citing competitors, journal bias, in-house bias, and the snowball effect, where authors tend to cite articles that other authors are citing. Despite these biases, the top 150 most cited papers in the field of otolaryngology are a holistic representation of the major advances and achievements over the past 50 years. This analysis allows us to understand which institutions and authors have contributed landmark findings and keystone ideas that impact the way otolaryngologists practice today.
CONCLUSION
Identifying the characteristics of the most-cited articles in otolaryngology allows readers to better understand the historic advances and groundbreaking developments of the field. There continues to be a disparity among authorship in high-powered articles. Future investigations should focus on identifying targets for improvement and initiatives to promote academic productivity and inclusivity for women.
AUTHOR CONTRIBUTIONS
Beatrice Go: Conceptualization methodology; validation; investigation; data curation; writing—original draft; writing—reviewing and editing; visualization. Neeraj Suresh: Investigation; data curation; writing—original draft; writing—reviewing and editing; visualization. Cammille Go: Writing—original draft; writing—reviewing and editing; visualization. Kevin Chorath: Writing—original draft; writing—reviewing and editing; visualization. Natasha Mirza: Writing—original draft; writing—reviewing and editing; visualization. Erica Thaler: Writing—original draft; writing—reviewing and editing; visualization. Alvaro Moreira: Writing—original draft; writing—reviewing and editing; visualization. Karthik Rajasekaran: Conceptualization; methodology; validation; investigation; data curation, writing—original draft; writing—reviewing and editing; visualization; supervision; project administration.
ACKNOWLEDGMENTS
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
ETHICS STATEMENT
This study did not involve human studies or contain any protected patient health information (PHI). As a result, there was no ethics approval applicable to this study.
Open Research
DATA AVAILABILITY STATEMENT
Data without the patient's recognizable identification can be acquired by contacting the corresponding author.