Female Representation in Ophthalmic Oncology and Pathology: An Analysis of Scholarship and Leadership

Abstract

Introduction

Female representation in ophthalmic oncology and pathology is unknown. We evaluated gender-based trends in authorship and leadership in these fields, focusing on intraocular tumors.

Methods

Articles containing “uveal melanoma” (UM), “retinoblastoma” (Rb), “primary intraocular lymphoma” (PIOL), “choroidal metastasis” (CM), and articles in Ocular Oncology and Pathology (OOP) were searched on PubMed and author names obtained. Leadership/awards from 4 international ophthalmic oncology and pathology societies were recorded. GenderAPI was used to predict gender.

Results

Among 4,245 UM, 9,660 Rb, 141 PIOL, 353 CM, and 568 OOP authors, more males were first (range, 52–67%) and last authors (range, 69–76%). Analysis of the 4 keywords showed overall female authorship increased from 2011 to 2016 to 2017–2022 (35%–40%, p < 0.001), with an increase in both first (45%–47%, p = 0.03) and last authorship (26%–33%, p < 0.001). Female first authors were more likely to publish with female last authors (Rb OR 1.72, p < 0.001; OOP OR 2.04, p = 0.01). Males held more leadership roles/awards in societies.

Conclusion

Female representation in ophthalmic oncology and pathology has grown but gender disparities remain. Female first authors were more likely to publish with female last authors. Most society leaders and awardees were male. Cultivating mentorship through professional organizations may help achieve gender parity.

Introduction

Ophthalmology, like many specialties in medicine, has traditionally been male dominated. In the last 5 decades, however, there has been a substantial increase in the number of female physicians in ophthalmology, with an analysis showing proportion of women ophthalmologists increased from 4% in 1969 to 27% in 2019 in the USA [1, 2]. The Association of American Medical Colleges 2021 data reports that 27% of all actively practicing ophthalmologists in the USA are female, but approximately 42% of ophthalmology trainees are female [3]. Disparities in female representation within subspecialties exist [4]. Females are more likely to choose medical subspecialties such as medical retina or uveitis over surgical subspecialties like surgical retina [2]. Societal and cultural factors may also be at play in these decisions. Although the increasing trend in the number of female ophthalmologists is encouraging, gender parity has yet to be achieved [1].

A consequence of fewer females in the field is a lower number of women in ophthalmology leadership positions. Within ophthalmology, representation in leadership roles, including medical school faculty, researchers, journal editors, society leaders, and award recipients, are skewed toward males [5–8]. Approximately 90% of academic ophthalmology department chairs in the USA are male, and females have been recipients of only 25% awards given by major ophthalmology and subspecialty societies since 1970 [7, 9]. Research publications mirror this disparity, though there have been positive trends over the last decade. Within ophthalmology, female first authors have increased by 21–34% from 2000 to 2009 in one study [10]. However, the same study found no increase in the number of last authors, indicative of a senior role, in ophthalmology from 2002 to 2014 [10].

Ophthalmic oncology and pathology are among the smallest subspecialties in ophthalmology, and the representation of females in these related fields has not previously been studied. Our objective was to examine patterns in academic article authorship related to the study of intraocular tumors over the past decade, as well as leadership roles within four of the largest ophthalmic oncology and pathology societies over recent decades. Based on our findings, we aim to identify opportunities for improving female representation within the subspecialties.

Methods

Research articles were identified on PubMed using the search terms “uveal melanoma” (UM), “retinoblastoma” (Rb), “primary intraocular lymphoma” (PIOL), and “choroidal metastasis” (CM) from 2011 to 2022. All articles published by the journal Ocular Oncology and Pathology were also included in the search. Ocular Oncology and Pathology is the official journal of the American Association of Ophthalmic Oncologists and Pathologists and the International Society of Ocular Oncology, launched in 2015. A validated global artificial intelligence software program, GenderAPI (www.gender-api.com), was used to assign perceived gender to individuals, based on their name, with a specified level of certainty. Authors assigned a gender by the software with <90% certainty were excluded. To validate the accuracy of the gender classifications for UM and Rb, a random subset of 25 authors from each cohort was sampled and assignment accuracy determined through online contact via Google (Google Inc., Mountain View, CA), supplemented by personal connections. To validate the accuracy of the gender classifications for PIOL and CM, the Gender-API output for all authors associated with these keywords was reviewed and confirmed through online contact and personal connections, given the limited number of authors in these cohorts. PIOL and CM authors whose gender could not be assigned based on these methods were excluded. Furthermore, names of the 66 most cited UM, Rb, and intraocular lymphoma authors in PubMed ranked by Expertscape (www.expertscape.com), an online database on professional expertise, were obtained and accuracy of their GenderAPI outputs confirmed.

Data for ophthalmic oncology and pathology society leadership were compiled using public information on websites of the American Association of Ophthalmic Oncologists and Pathologists (AAOOP), the International Society of Ocular Oncology (ISOO), Pan-American Ocular Oncology Society (PAOOS), and Asia-Pacific Society of Ocular Oncology and Pathology (APSOOP). Data included named lecture speakers, award recipients, and society leaders. Named lectures included the Stallard, Ellsworth, Frezzotti, Judith Kingston, Jerry A. Shields, Lex van der Eb, President’s, Murphree, and Zimmerman Lectures. The only award studied was the ISOO Pioneers in Ocular Oncology award. Society leaders included President/President-Elect, Secretary-Treasurer, and Board of Directors for AAOOP (all 2-year positions); President, Vice-President, Secretary, and Treasurer for ISOO (all 2-year positions); President for PAOOS (1-year position); and President, Vice-President, Secretary General, Scientific Secretary, Treasurer, and Council Members for APSOOP (all in 2021). Female representation of the current editorial board of Ocular Oncology and Pathology was also conducted. Names and photographs (as available) of leaders were collected and gender assigned through online contact and personal connections of the authors.

Python 3.12.0 (Python Software Foundation) was used to collect unique journal articles from PubMed and identify first and last full author names (online suppl. File; for all online suppl. material, see https://doi.org/10.1159/000542560). Chi-square tests were used to compare proportions of female authors and leaders between specified time intervals: 2011–2016 versus 2017–2022 for UM/Rb/PIOL/CM articles, 2015–2018 versus 2019–2022 for Ocular Oncology and Pathology articles, 1979–2005 versus 2006–2022 for AAOOP leadership roles, 2009–2015 versus 2016–2022 for ISOO leadership roles, and 2011–2016 and 2017–2022 for the PAOOS President role. The ISOO Biannual Stallard Lecture (2009–2015 vs. 2017–2022) and AAOOP Annual Zimmerman Lecture (1986–2003 vs. 2004–2021) were the only awards with long-term data, analyzed with chi-square (if all fields had a sample of ≥5) or Fisher’s exact test (if any field had a sample of <5). The cut-off year for publications were pre-selected to investigate the past decade of the literature (2011–2022), whereas the cut-off year for leadership roles and awards were preselected to bisect the range of years provided on the publicly available websites up to 2022. Given the paucity of long-term publicly available data for APSOOP, female prevalence was assessed without prespecified time intervals. Gender correlation and odds of female first authors publishing with female last authors for UM, Rb, PIOL, CM, and Ocular Oncology and Pathology were assessed using chi-square testing and logistic regression analyses. All statistical analyses were performed using XLSTAT v2021.1.1 (Addinsoft, Paris, France), with a two-tailed significance level set at 0.05.

Results

Authorship

A total of 2,858 UM articles representing 5,716 first and last authors, 7,242 Rb articles representing 14,484 first and last authors, 76 PIOL articles representing 152 first and last authors, and 193 CM articles representing 386 first and last authors between 2011 and 2022 were identified. Among these articles, 1,067 UM, 3,649 Rb, 52 PIOL, and 108 CM articles were identified between 2011–2016 and 1,791 UM, 3,593 Rb, 24 PIOL, and 85 CM articles were identified during 2017–2022. Subsequently, after filtering based on gender classification confidence, 4,245 UM authors (74%), 9,660 Rb authors (67%), 141 PIOL authors (93%), and 353 CM authors (91%) were included in the analysis. In randomly selected sets of UM and Rb authors, mean percentage of correct assignments was 95% (range 92%–100%).

Analysis of all articles in UM, Rb, PIOL, and CM showed a significant increase in female authorship from 2011–2016 to 2017–2022 (35%–40%, p < 0.001). For overall first authorship, there was a significant increase from 45% in 2011–2016 to 47% in 2017–2022 (p = 0.03). Similarly, for overall last authorship, there was a significant increase from 26% in 2011–2016 to 33% in 2017–2022 (p < 0.001).

Female UM authorship significantly increased from 2011–2016 to 2017–2022 (37% vs. 41%, p = 0.01). A total of 1,001 UM first authors were female (48%), with 371 females among 789 authors in 2011–2016 and 630 females among 1,292 authors in 2017–2022. The proportion of female first authors did not change significantly between the two time periods (46% vs. 49%, p = 0.4). A total of 680 female last authors were identified in UM-indexed articles, with 227 among 825 authors in 2011–2016 and 453 among 1,339 authors in 2017–2022. There was a statistically significant difference in the percentage of female last authors between the two periods (p = 0.002), which increased from 28% in 2011–2016 to 34% in 2017–2022. Female Rb authorship significantly increased from 2011–2016–2017–2022 (45% vs. 47%, p < 0.001). A total of 2,131 first authors were female (46%), with 1,086 among 2,442 authors in 2011–2016 and 1,045 among 2,240 authors in 2017–2022. The proportion of female first authors increased from 44% to 47%, but this did not meet statistical significance (p = 0.1). Analysis of last authorship showed 1,457 females (29%) were last authors in Rb articles, with 657 among 2,545 authors in 2011–2016 and 800 among 2,433 authors in 2017–2022. Similar to percentage of UM last authors, the percentage of female Rb last authors significantly increased from 26% to 33% (p < 0.001). Female PIOL authorship was not significantly different between the two time periods (40% vs. 26%, p = 0.2). Thirty-two first authors were female (46%), with 22 among 46 authors in 2011–2016 and 10 among 24 authors in 2017–2022. The proportion of female first authors did not change significantly between the two time periods (48% vs. 42%, p = 0.1). A total of 17 females (24%) were last authors in PIOL articles, with 14 among 45 authors in 2011–2016 and 3 among 26 authors in 2017–2022. The percentage of female PIOL last authors did not change significantly between the two time periods (31% vs. 12%, p = 0.06). Female CM authorship was not significantly different between the two time periods (27% vs. 39%, p = 0.08). A total of 58 CM first authors were female (33%), with 29 among 104 authors in 2011–2016 and 29 among 73 authors in 2017–2022. The proportion of female first authors did not change significantly between the two time periods (28% vs. 40%, p = 0.1). Analysis of last authorship showed 55 females (30%) were last authors in CM articles, with 27 among 100 authors in 2011–2016 and 26 among 76 authors in 2017–2022. The proportion of female first authors rose over time but did not change significantly between the two time periods (27% vs. 34%, p = 0.3). The trend in authorship for UM, Rb, PIOL, and CM are shown in Figures 1–4.

Fig. 1.

UM authorship over time.

Fig. 4.

CM authorship over time.

Fig. 2.

Rb authorship over time.

Fig. 3.

PIOL authorship over time.

Among 1,822 UM articles (64%) with first and last authors who both met inclusion criteria, 591 (32%) had female last authors. A total of 621 (34%) were published by male first and last authors, whereas 285 (16%) belonged to female first and last authors. Female first authors of UM articles were no more likely to be a co-author with female last authors as with male last authors (OR 0.95, 95% CI 0.78–1.15, p = 0.6). Similarly, from 3,925 Rb articles (54%) with both authors who met inclusion criteria, 1,199 (31%) had female last authors. There were 1,559 (40%) articles with both male first and last authors, compared to 674 (17%) with both female first and last authors. Female first authors of Rb articles were significantly more likely to publish with female last authors (OR 1.72, 95% CI 1.50–1.97, p < 0.001). Among 65 PIOL articles (86%) with first and last authors who both met inclusion criteria, 15 (23%) had female last authors. A total of 27 (42%) were published by male first and last authors, whereas 6 (9.2%) belonged to female first and last authors. Female first authors of PIOL articles were no more likely to be a co-author with female last authors as with male last authors (OR 0.78, 95% CI 0.36–3.79, p = 0.7). From 161 CM articles (83%) with both authors who met inclusion criteria, 49 (30%) had female last authors. There were 76 (47%) articles with both male first and last authors, compared to 15 (9%) with both female first and last authors. Female first authors of CM articles were no more likely to be a co-author with female last authors as with male last authors (OR 0.91, 95% CI 0.35–1.49, p = 0.8).

Among 310 articles and 620 first and last authors, a total of 568 (92%) first and last authors were included from Ocular Oncology and Pathology published 2015–2022 after excluding 52 authors based on gender classification (Fig. 5). Overall, the majority of first (57%) and last (72%) authors were male, with no significant difference in female authorship between 2015–2018 and 2019–2022 (p = 0.2). Although the number of female first authors increased from 56 (40%) to 62 (47%), the change in proportion did not reach statistical significance between 2015–2018 and 2019–2022 (p = 0.2). Despite an increase in the number of female last authors from 38 (26%) to 42 (30%), the increase in proportion did not meet statistical significance (p = 0.5). Of 256 articles (83%) with first and last authors who both met inclusion criteria, only 70 (27%) had a female last author. One hundred and fifteen (45%) articles had both male first and last authors, compared to 39 (15%) articles with both female first and last authors. Female first authors in Ocular Oncology and Pathology were significantly more likely to publish articles with female last authors (OR 2.04, 95% CI 1.17–3.56, p = 0.01).

Fig. 5.

Ocular Oncology and Pathology authorship over time.

Leadership and Awards

Analysis of 17 ISOO leadership positions from 2009 to 2022 showed 3 out of 10 leaders (30%) were female in 2009–2015, compared to 5 out of 9 leaders (56%) in 2015–2022. However, this increase did not reach statistical significance (p = 0.5). A total of 81 leadership positions held within the AAOOP spanning the years 1979 through 2022 were identified. During 1979–2005, 49 out of 64 leaders (75%) were men (Fig. 6). Female leaders increased from 15 out of 64 (25%) in 1979–2005 to 7 out of 17 (41%) in 2006–2022, but there was no statistically significant difference (p = 0.2). No data was available for the President role in 2012–2014, Secretary-Treasurer role in 2008–2009 and 2012–2021, and Board of Director roles in 1997 and 2004–2021 for AAOOP. Among 12 PAOOS Presidents identified from 2011 to 2022, 2 out of 12 (17%) Presidents were female and were elected in 2021 and 2022. Analysis of 11 APSOOP leaders, all elected in 2021, revealed 3 were female (27%) and served as co-Vice President, Treasurer, and a Council Member (Fig. 6). A total of 2 editors-in-chief and 11 editorial board members comprise the current editorial board of Ocular Oncology and Pathology. Among these members, 4 (36%) are female.

Fig. 6.

Leadership roles in ophthalmic oncology and pathology societies over time. AAOOP, American Association of Ophthalmic Oncologists and Pathologists; ISOO, International Society of Ocular Oncology; PAOOS, Pan-American Ocular Oncology Society.

Among 36 AAOOP and 18 ISOO named lectures and award recipients from 1986 to 2022, 4 (13%) and 4 (29%) were female, respectively. All ISOO Stallard Lecture awardees were male. Female awardees for the AAOOP Zimmerman Lecture increased from 1 out of 18 (6%) to 3 out of 18 (17%), though this difference was not statistically significant (p = 0.6).

Discussion

In this study, we found that female scholarship has increased in the fields of ophthalmic oncology and pathology over time, yet gender disparities remain with more men than women in prominent authorship positions, leadership roles, and award recipients. We identified a significant increase in the proportion of female authors in articles on intraocular tumors, including a significant increase in both female first and last authorship. Analysis of individual keywords showed a significant increase in the percentage of female last authors in topics on UM and Rb over the past decade. However, a similar trend was not seen for increased female first authorship in UM and Rb or with either authorship position for PIOL and CM. Female first authors were more likely to publish with female last authors in articles on Rb and within Ocular Oncology and Pathology. There was no significant difference in the proportion of women in professional society leadership roles and more men received awards. Our findings highlight both the promising growth of female scholars in ophthalmic oncology and pathology, as well as ongoing disparities in these fields.

Our analysis of authorship trends revealed modest increases in the proportion of females serving as first authors. Even though female first authorship in combined keyword analysis significantly increased over time, analysis of individual keywords did not show statistical significance. As first authors are often junior, one might have expected to see growth of female first authors before that of female last authors, who are typically established senior researchers. This would have been supported by the greater proportion of female ophthalmology residents and fellows compared to active female ophthalmologists. Surprisingly, our study showed a different finding. Females last authorship demonstrated a notable increase over time in UM and Rb, indicating a positive trend toward greater senior roles within the field. It is important to note that our study did not differentiate the type of scholar in authorship and likely includes basic scientists and other professionals engaged in oncology and pathology research, which may potentially explain this finding. The smaller sample size of PIOL, CM, and Ocular Oncology and Pathology articles dataset did not provide sufficient evidence to conclude significant changes in last authorship proportions. Females were 1.7 and 2.0 times more likely to publish with female last authors in Rb and Ocular Oncology and Pathology articles, respectively. Though this was not significant for UM, PIOL, or CM articles, this suggests mentorship from senior female researchers is key for junior female researchers and may help increase female representation.

Our conclusions on authorship are similar to what was reported by another study investigating female first authors and last authors of articles published from 2012 to 2021 in three large general ophthalmology journals: Ophthalmology, JAMA Ophthalmology, and American Journal of Ophthalmology [11]. Researchers found a significant increase in the proportion of female last authors over the last decade from 23% to 31% but did not find a significant increase in the proportion of female first authors. Together with our findings, this points to the need for a critical analysis of mentorship for females early in training. In contrast, another study reported a significant increase in the proportion of both female first authors and last authors of pediatric ophthalmology articles in the Journal of AAPOS from 2011 to 2019 [12]. These findings may partially be explained by the greater proportion of women (53%) than men (47%) in pediatric ophthalmology. A meta-analysis of studies published prior to 2021 assessing authorship in the surgical literature showed the proportions of female first authors and last authors were 12% and 21%, respectively [13]. We found a higher proportion of female first and last authors in this study, ranging from 12% to 48% depending on the keyword and journal studied. Hence, compared to other surgical subspecialties, ophthalmic oncology and pathology appear to have high female representation. Our findings suggest that the number of established senior female researchers in these fields is rising, and more work is needed to support junior female researchers.

Leadership roles and recognition within AAOOP and ISOO showed encouraging trends for increased female representation, however, these did not achieve significance. Our findings add to the growing literature of reports from other subspecialty societies. A study of leadership in the American Society of Retinal Specialists showed that female leadership in its board of director positions was significantly higher in 2020 than in 1983–1989 [14]. Another study from the American Society of Ophthalmic Plastic and Reconstructive Surgery found a significant increase in female early career awardees (6%–29%) from 1969–1994 to 1995–2018, but females were significantly underrepresented as late-career award winners [15]. Rivera and colleagues reported that ophthalmology journals with a higher representation of female executive board members also had more female authorship [11]. Our study found that female representation in the current editorial board of Ocular Oncology and Pathology was 36%, but there is no publicly available data on prior editorial boards. Lack of data in ophthalmic oncology and pathology makes it difficult to establish precise benchmarks for gender parity in leadership roles. Ideally, the proportions would reflect the percentage of female ophthalmology residents and fellows (42%) and active female ophthalmologists (27%), as reported for 2021 [3]. Estimating the prevalence of non-ophthalmology-trained female researchers engaged in these fields further poses challenges. As the number of females in senior leadership increases, more females are expected to receive these awards. Future work on collecting awards and long-term leadership data from ophthalmic oncology and pathology professional societies is warranted.

Our study had several limitations. The GenderAPI algorithm is a validated gender classification tool powered by artificial intelligence and utilized in 190 countries, with a large database that includes over 6 million names. A threshold of 90% was set and a high percentage of correct assignments was found on random sampling. This resulted in the exclusion of 1,471 UM (26%), 4,824 Rb (33%), 52 (8%) Ocular Oncology and Pathology authors from analysis due to gender classification uncertainty. Despite manual review of each entry, 11 PIOL (8%), and 33 CM (9%) authors were also similarly excluded. Names not as prevalent in this database may have been labeled as low certainty. Though the GenderAPI database includes names from several countries, there may be nationality bias during the gender classification process. This is particularly important as a study of articles on eye neoplasms between 1966 and 2012 found the USA was the predominant country in number of publications, followed by Germany and the United Kingdom [16]. The authors recognize that gender is a social construct and acknowledge that GenderAPI does not include individuals who identify as nonbinary. The reliance on incomplete leadership data may introduce bias, but also reflects the study’s focus on transparency of publicly available data. Similarly, data on the overall number of female members of the professional societies studied were not publicly available. Finally, given the focus on select intraocular cancers, the study did not evaluate the growing number of female researchers in other areas of ophthalmic oncology and pathology such as ocular surface tumors and orbital tumors.

In conclusion, female representation in ophthalmic oncology and pathology research and leadership is growing. Cultivating mentorship, particularly through formal programs within research and professional organizations, may be a pivotal strategy for gender parity. To this end, Women in Ocular Oncology were established in 2013 and a current effort within ISOO is development of a mentorship program pairing junior and senior female ophthalmic oncologists. Continued transparency in leadership positions and awards will also support this mission, fostering equitable leadership and recognition opportunities. Presence of female leaders will benefit female scholarship, ultimately promoting female representation in our fields.

Statement of Ethics

This study protocol was reviewed by Oregon Health & Science University Institutional Review Board and was deemed not human research and exempt for further review (STUDY00024221). Ethical approval and consent were not required as this study was based on publicly available data.

Conflict of Interest Statement

A.H.S. and P.M. are consultants for Castle Biosciences Inc. The remainder of the authors has no conflicts of interest to declare.

Funding Sources

O.E.U., L.H., and A.H.S. were supported in part by National Eye Institute P30 EY010572 from the National Institutes of Health (Bethesda, MD) and by an unrestricted departmental funding from Research to Prevent Blindness (New York, NY). Funders had no role in the design, data collection, data analysis, and reporting of this study.

Author Contributions

O.E.U., B.S., and L.H.: conceptualization, data analysis, data interpretation, and drafting of work. M.B.: data analysis, data interpretation, and drafting of work. J.L.B., P.M., and D.S.G.: conceptualization and critical review of written work. A.H.S.: conceptualization, funding support, and critical review of written work.

Funding Statement

O.E.U., L.H., and A.H.S. were supported in part by National Eye Institute P30 EY010572 from the National Institutes of Health (Bethesda, MD) and by an unrestricted departmental funding from Research to Prevent Blindness (New York, NY). Funders had no role in the design, data collection, data analysis, and reporting of this study.

Data Availability Statement

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.