Fewer women overall and in first, middle, and last authorship positions in COVID-19 ophthalmology research publications
In this study, women did not exceed the 40% mark in any authorship position and category. Overall women represented 33.1% middle, 31.2% first, and 24.6% last authorship positions. Women were most likely to be middle, followed by first and then last author in all categories, including clinical, basic and clinical science, general ophthalmology, and subspecialty journals. Conventionally, the first listed author is considered the lead investigator, the last author is the senior (or corresponding) author, and the remaining authors are middle authors. First and last authors are considered leadership positions [
23]. This study shows that women occupy fewer research leadership positions in COVID-19-related ophthalmology studies than their male colleagues.
These results are in line with early findings in other medical fields, notably public health, internal medicine, and radiology [
8,
24]. The pandemic has brought challenges to all researchers, due to ongoing social isolation measures leading to predominant work from home. With limited access to childcare, early and mid-career women are particularly affected by COVID-19 [
25]. Late-career women are also affected, as their age group may predispose them to retire early due to circumstances surrounding the pandemic [
26]. Studies examining the gender gap in academia have proposed multiple theories: the historical workplace marginalization of women, the lack of female research leadership opportunities, and unblinded peer review bias [
27]. While our observational results cannot conclusively correlate COVID-19 challenges with a decrease in female authorship, they suggest that women are unequally burdened by societal changes occurring during the pandemic.
COVID-19 ophthalmology authorship compared to previous ophthalmology authorship
COVID-19 has increased the authorship gender gap in ophthalmology. The gap between the proportion of female authors in COVID-19 ophthalmology research and the predicted 2020 proportion of female authors based on the trend of previous years (2002–2019) is 6.1% for overall authors, 7.8% for first authors, and 5.5% for last and middle authors (Fig.
3b).
A gap in first authorship positions is found when comparing COVID-19 ophthalmology authorship to the 2019 comparator group in clinical journals (significant difference of 7.3%), clinical and basic science journals (9.2% difference), general ophthalmology journals (significant difference of 7.5%), and subspecialty journals (6.1% difference). The differences are also high for senior authorship position in clinical journals (significant difference of 5.7%), clinical and basic science journals (significant difference of 11.3%), general ophthalmology journals (significant difference of 5.9%), and subspecialty journals (11.3% difference). The gap is therefore larger for women occupying leadership (first and last) positions during COVID-19.
Limitations
The main study limitation was that the authors did not self-identify their gender, which would have confirmed gender assignments and allowed for a non-binary gender spectrum. Indeed, we had to use an imperfect gender predictive algorithm, albeit correct 98% of the time (Gender-API’s estimated accuracy). Furthermore, while we acknowledge that gender exists on a spectrum and is socially produced, we were constrained by Gender-API’s binary gender output.
Possible solutions to address female authorship underrepresentation
Implicit and unconscious biases could be responsible for this 2020 gender gap. These biases can start to be overcome through adequate management training, mentorship, and sponsorship, which are strategies to promote women’s place in academia [
28].
On a larger scale, these biases can be reinforced by national, regional, local, and institutional policies. Female underrepresentation seems to occur to a greater degree in countries with high gender inequality indices: India with a high GII value (0.501) has 31.6% women authors, while the USA with a lower GII value (0.182) has 35.0% women authors (Fig.
2c). When excluding geographical regions with a small number of authors (e.g., Sub-Saharan Africa with 4 authors), a similar trend can be noted: North America (lower GII, 35.3% women) has a higher proportion of female authors than South Asia (high GII, 31.5% women) and Latin American and the Caribbean (high GII, 26.3% women) (Fig.
2a).
COVID-19 may exacerbate these pre-existing challenges for women. Reliable access to childcare has been listed as a major source of anxiety for healthcare professionals, especially with school closures and increased work hours [
29]. During the pandemic’s first peak of cases, countries like the UK (GII = 0.119) and Canada (GII = 0.083) provided temporary emergency childcare to essential workers. These social policies supporting workers, in addition to long-term job security and workplace re-entry support plans, can be particularly beneficial to early and mid-career women [
30].
Public policies that help bridge the gender gap in research involve ensuring equal pay, granting basic legal rights and reforms (e.g., right to education, freedom of choice, countering practices leading to sex imbalance at birth), and efficiently implementing and encouraging couples to take advantage of shared parental leaves [
31]. Housework and childcare more often are the female partner’s responsibility (when there is one), and a shift toward balanced sharing of these duties would allow for greater female academic productivity and career advancement [
32].
Publication and submission processes should be examined and potentially reshaped [
33]. While double masking may reduce reviewer bias, a recent study demonstrates that the double-blind review process (where authors nor reviewers did not know each other’s identity) did not increase the incidence of female authorship [
34]. An alternative solution could be to disclose and monitor author and reviewer genders by editorial teams, which would actively encourage gender-diverse teams [
10]. A more aggressive stance could be the use of quotas, which have been proven to be an effective solution to diminish demographic gaps in politics and economics [
35]. Quotas could lead to an increase in female leadership positions, such as more female academics and more women in the senior authorship position [
36]. The submission, review, and publication process has been shown to have bias and has only slowly evolved over the past 50 years. Further efforts are needed to ensure that this process is fair, based on academic merit, and gender-blind [
37].