Abstract

Background

Women are under-represented in the academic oncology workforce in general and even more strikingly in senior leadership positions.^1,2 There are many potential causes of gender disparities in academic oncology, including bias, and harassment, which have downstream effects on retention, recruitment, and representation in leadership.³ Among both men and women, pre-pandemic rates of burnout ranged from 28% to 38% in different oncology subspecialties (radiation, medical, and surgical oncology) and have been attributed in part to increasing burden of administrative tasks, professional-personal life imbalance, and caring for patients who are critically ill.^4,5

Alarming data regarding the effects of the pandemic on careers of women in academia are rapidly emerging, many of which have likely been exacerbated by intersectional factors such as race, ethnicity, and/or caregiver status.^6,7 For example, female researchers and those with young dependents saw the greatest reductions in time devoted to research during the COVID-19 pandemic.⁸ This translated into a reduction in female authorship of grant submissions and publications contemporaneous with the start of the pandemic, with disproportionate effects on Black women and mothers.^7,9-11 This occurred at the same time that women took on greater service roles for academic journals.^9,12

According to one 2020 survey of 667 female oncologists, 22% of those working in an academic setting were considering leaving academic oncology in the next 5 years.¹³ Among respondents, approximately 40% felt that the largest sacrifice of a career in academic oncology was less time spent with loved ones, and more than half felt that they were less likely than male colleagues to be promoted.¹³ However, because this survey was conducted only with female respondents, data are lacking on how the experiences of men and women compared at academic cancer centers during the COVID-19 pandemic.

In 2021, the UCSF Helen Diller Family Comprehensive Cancer Center (HDFCCC) formed a Gender Equity Committee (GEC) in response to emerging data on the exacerbation of gender inequities during the COVID-19 pandemic. The GEC sought to investigate the scope of gender-related workplace issues by assessing the climate for all faculty at HDFCCC.

Methods

Data Collection

Given the sensitive nature of the topic, demographic variables were intentionally limited to protect the identities of respondents and to improve response rates. Respondents were asked to self-report on their current gender identity (man, woman, non-binary/third gender, transwoman, transman, genderqueer, additional category, or prefer not to disclose); their promotion schedule at UCSF (on time, accelerated, or delayed); primary work responsibilities (clinical/patient care, education, research, and/or administration/management); and whether they self-identified as a primary caregiver for children <18 years old (yes or no). Respondents were asked if they self-identified as part of a racial or ethnic group that is under-represented in health professions or medicine (yes or no), according to the NIH definition.¹⁴ Response selections for all demographic questions included an option: “Prefer not to disclose.” Respondents were asked to report their primary academic appointment, with the option of 5 oncology-specific Departments or Divisions, prefer not to disclose, or other (with the option to include a write-in response). For the purposes of maintaining respondents’ anonymity and enhancing sample size, faculty were not asked to report their current rank.

The Culture Conducive to Women’s Academic Success (CCWAS) tool was previously developed and validated to define, measure, and analyze how culture constrains women’s career success in academic medicine.¹⁵ The CCWAS consists of 46 questions across 4 distinct domains (Fig. 1): (1) equal access to opportunities, (2) work-life balance, (3) freedom from gender bias, and (4) leadership support.¹⁵

Open in a separate window

Figure 1.

Four domains of the original Culture Conducive to Women’s Academic Success (CCWAS) instrument. Adapted with permission from Wolters Kluwer Health, Inc.: Westring et al.¹⁵ (2012). The Creative Commons license does not apply to this content. Use of the material in any format is prohibited without written permission from the publisher, Wolters Kluwer Health, Inc. Please contact permissions@lww.com for further information.

We adapted the original CCWAS tool to be relevant for individuals of all gender identities, making minor modifications only to remove gender specificity from the questions. For example, instead of “Women faculty have equal access to career development opportunities,” the modified instrument stated: “I have equal access to career development opportunities.” As in the original survey, questions were organized according to the 4 dimensions, and all sections were scored using a 5-point Likert-type scale from 1 (strongly disagree) to 5 (strongly agree). At the end of each section, a separate question provided an option for respondents to share relevant experiences and comments in free text.

Following the CCWAS tool, 2 additional sections were added to the survey: (1) burnout and (2) bias reporting. With regard to burnout, respondents were asked to rank their level of burnout on a scale of 1-5 based on a non-proprietary single-item burnout measure previously described.¹⁶ The section on reporting included questions adapted from a previous survey, soliciting prior experiences regarding gender bias or harassment at work, reporting, impact, and institutional response.¹⁷ In addition, this section contained open-ended questions asking the respondents about their ability to voice concerns about gender bias and recommendations for policies or directives that should be implemented by institutional leadership to address gender bias.

The survey was administered electronically using the Qualtrics Software Platform. The introduction to the survey offered assurance of anonymity with the following statement: “All responses will be reported anonymously. If any particular demographic category has fewer than 5 responses, data from this subgroup will not be reported but will be analyzed as part of the larger dataset.”

Results

CCWAS Domains

Median scores and interquartile ranges (IQRs) were calculated for each of the 4 CCWAS domains by faculty gender (Table 2). The overall median total score for the CCWAS was lower for women as compared to men (women: 108.50 [IQR 86.00, 133.25] vs. men: 128.00 [IQR 107.00, 157.75], P = .001) (Fig. 2). Within each primary work responsibility category, there were similar trends for the CCWAS total score (as well as scores for each CCWAS domain; Supplementary Table 1). Optional free text responses are summarized in Table 3, according to domain.

Table 2.

Wilcoxon rank sum test results of CCWAS domain scores by gender in each of 4 domains and total combined score.*

	Equal access		Work-life balance		Freedom from gender bias		Leadership support		Total score
	Median (IQR)	P-value	Median (IQR)	P-value	Median (IQR)	P-value	Median (IQR)	P-value	Median (IQR)	P-value
Women	37.00 [31.00, 45.00]	<.001	20.00 [15.25, 27.00]	.002	10.00 [6.00, 13.00]	.4	43.00 [30.25, 52.00]	.01	108.50 [86.00, 133.25]	.001
Men	49.00 [42.00, 54.00]		25.00 [19.25, 32.00]		11.50 [0.50, 17.75]		51.00 [39.00, 60.00]		128.00 [107.00, 157.75]

Open in a separate window

*Higher CCWAS scores indicate the respondent perceived the work culture to be favorable.

Table 3.

Illustrative quotes from free text answers for the 4 CCWAS domains.

CCWAS domain	Free text answer
1.Equal access
“There are many inequities in opportunities given—be it protected time, vacation, advancement opportunities. When this is brought up, I am told ‘it will get better.’ However, I see others getting opportunities, and I feel I am still lacking.”
“I constantly have to fight for space for myself and my lab, but at the same time, I am criticized for not growing the lab fast enough.”
“I’ve noticed I’m most often asked to take minutes at meetings, which precludes me from participating as much as I would like.”
“I have lost traction and opportunities during each pregnancy. People have been reluctant to give me opportunities once they know I’m pregnant, but it can take years to recover from the missed opportunity.”
“I am chosen for a lot of leadership work but do not have enough protected time or compensation for these roles.”
“Compensation is opaque … It is also unclear how to negotiate for increases. When asked, I was told from leadership that getting another job offer is the best way to negotiate salary.”
2.Work-life balance
“There is no formal support. A competitive environment favors overwork.”
“We are ridiculously overworked, undervalued, and there is no time for the core academic mission.”
“The culture is that everyone emails and Zoom calls at all hours of the day/night and on weekends; We are expected to join calls even when we are driving our kids to school or during family dinner time. There are no boundaries that are respected, including when we take vacation. And when we do return, there’s a disgruntled sense that we have to make up the time.”
“There should be support in place for faculty who need to take time off, like hiring someone for that temporary amount of time off. That would ensure the faculty stays productive while still taking on family responsibilities, without jeopardizing productivity and promotions.”
“It is standard in my division to make up the call you miss on maternity leave.”
“I don’t have any children and feel that I am expected to be flexible and able to cover the clinic significantly more often than those who have kids.”
“Loss of administrative support has added hours of extra work to my plate each week, taking away from productivity or valuable time with my family.”
3.Supportive leadership
“Our Chair/Chief is making an effort to improve things but is stymied by senior faculty intent on preserving the old-school ways.”
“While I think leader support is important and the tone set by the Chair/Chief is influential… I think systemic, structural, and cultural changes are needed to address a lot of these things, it cannot be up to one person.”
“There are multiple toxic members in our workplace that the Chair/Chief does not have the power to fire.”
“The Chair/Chief has removed leadership opportunities from faculty who have challenging family situations. Rather than working with the faculty members to help figure out how to be supportive, they just categorically decided to remove them and has commented that they doesn’t think they are able to do it, when evidence shows otherwise.”
4.Freedom from gender bias
“I have been yelled at by my department chair in front of my lab members. Have also experienced other faculty belittled because of ethnicity.”
“I feel like I can report gender bias, but often that nothing will be done.”
“I was attacked verbally, insulted and undermined for being openly gay.”
“I have had patients say inappropriate things to me based on my gender.”
“When I have had interactions with abusive patients, I routinely receive no support in setting limits, and the manager typically handles the situation by disparaging me to others.”
“I received an inappropriate recommendation for promotion into an education rather than a research track.”

Open in a separate window

Open in a separate window

Figure 2.

Boxplots demonstrating distributions of CCWAS total combined scores according to gender of respondents. Box limits indicate the range of the central 50% of the data, with a central line marking the median value.

Of the 17 questions in the CCWAS domain of Equal Access to Opportunities, 7 resulted in statistically significant differences in responses from men and women faculty (Supplementary Table S2, Section A). Women faculty were less likely to report equal access to career development opportunities (P < .001), recognition for work productivity (P = .003), receipt of feedback (P = .005), awards/honors (P = .012), equitable pay (P < .001), leadership opportunities (P < .001), and attention to comments in meetings (P < .001), than men faculty. In addition, the median total scores from women faculty respondents were significantly lower than men (Table 2).

The Work-life Balance domain also had a lower median overall score for women as compared to men (Table 2). Of the 8 questions in the CCWAS domain of Work-life Balance, 3 resulted in statistically significant differences when analyzed by gender (Supplementary Table S2, Section B). Women faculty were more likely to report that a reduction of workload for family reasons results in an expectation to take on extra work upon their return (P = .018). Second, women faculty were more likely to report that they are viewed as less committed to their careers by other faculty (P = .033). Finally, women faculty were more likely to report losing out on career opportunities due to having children (P = .02).

Of the 13 questions in the CCWAS domain of Leadership Support (Supplementary Table S2, Section C), one resulted in a statistically significant difference in responses from men and women respondents. Women faculty were less likely than men faculty to state that their chief or chair encourages faculty to take advantage of institutional policies for managing work and family (P = .009). Overall, there was a lower median score in this domain for women as compared to men (Table 2).

Of the 4 questions in the domain of Freedom from Gender Bias (Supplementary Table S2, Section D), all 4 resulted in statistically significant differences in responses from women and men respondents, despite no significance difference in median score for this domain (Table 2). Women respondents were less likely to report that faculty feel comfortable raising concerns about the supportiveness of the work environment (P = .004), that faculty are encouraged to raise concerns about gender biases (P = .004), and that concerns about gender bias and harassment are handled confidentially and effectively (P = .002). Finally, women were more likely to feel that when a faculty member raises concerns about gender bias, the person would be seen as a “whiner” (P < .001).

Discussion

We report a study conducted at an NCI-designated Cancer Center measuring the extent to which work culture is conducive to faculty success. Although this assessment focuses on dimensions of culture which have been established as most directly affecting the success of women faculty, we surveyed all faculty across genders in order to establish a basis for comparison. Analyses of reported experiences among faculty according to gender revealed a significant difference in questions in all 4 domains of the CCWAS and by total median score; differences were most pronounced in the domain of Equal Access to Opportunities. This was consistent across a variety of work roles including education, clinical, research, and administration.

Within the domain of Equal Access to Opportunities, significant differences were detected in access to career development and leadership opportunities. This finding is consistent with prior reports in the literature that women are under-represented in oncology leadership positions.¹ In addition, women respondents were more likely to report that they were not recognized for productivity or substantive contributions in meetings and less likely to receive awards or honors. This is also consistent with previous reports that women in oncology are under-represented in awards from professional societies.¹⁸

Significant differences were detected in several questions in the Work-Life Balance domain, which may reflect that women shoulder a higher burden of extra-professional responsibilities.¹⁹ This study was conducted in late 2021, following a period when many childcare facilities were closed and prior to the COVID-19 vaccine approval for children <12 years, and therefore at a time when extraprofessional demands were heightened among primary caregivers.^20-22 Burnout was also demonstrated to be higher in those with responsibilities as a primary caregiver. This analysis could not cover all work-life related factors, such as caregiving for elders and other family members and, therefore, may have under-ascertained extraprofessional demands.

In the Freedom from Gender Bias domain, significant differences in comfort, perceived freedom to raise concerns, and expectations for appropriate handling related to actual experiences of gender bias were all detected. These findings are in contrast with fewer statistically significant differences in the Leader Support domain, which may reflect that while leaders may outwardly espouse values of diversity, equity, and inclusion, the ability of faculty to self-advocate to leaders when concerns arise may be in fact be more problematic. Although statistical significance was not achieved in the total score for the Freedom from Gender Bias domain, it is noteworthy that the results of each individual question were statistically significant. The failure to detect a statistically significant difference in the total domain score could be related to the small number of questions (n = 4), a higher level of missingness for responses in this section, or a reluctance to share experiences of bias despite survey anonymity.

Potential limitations of this study must be acknowledged. First, this study was conducted in the midst of the COVID-19 pandemic, and no prior data from our institution existed for comparison. Thus, it is unknown whether results reflect pre-existing conditions at our institution, or if the results reflect factors related to or exacerbated by the pandemic.^23,24 However, many of the free text answers summarized in Table 3 reflect issues that are not directly related to the pandemic. Second, the majority of respondents were women, with dramatically different response rates by gender (70% vs. 25%), which may reflect a self-selection effect and may bias the results based on those who are impacted negatively by work culture being more likely to respond. High non-response rates from men may reflect, distrust, lack of engagement related to issues of gender equity, and/or reluctance to participate in survey completion since they may not be directly affected by the topic.^25,26 The 2:1 participation of women to men is, however, consistent with previous studies showing that women are generally more likely to respond to surveys^27,28 and are more likely to respond to a survey related to gender. Finally, our analyses were limited by the intentional omission of data on faculty rank from our survey, which could have potentially shed light on the directionality of any association of gender bias with career stage. Due to limited sample size, we did not evaluate associations of race, ethnicity, and sexual orientation; therefore, the intersectional issues at the crossroads of gender and other identities were not addressed in the current study and remain priorities for future inquiry.

These data have potential to inform the development of targeted interventions to reduce inequities in academic oncology. Related to findings from the Equal Access to Opportunities domain, we propose development of faculty leadership and coaching programs particularly targeting early and mid-career women faculty, evaluation of composition of leadership positions by gender and race/ethnicity, and adhering to best practices for recruitment/hiring to ensure equity in the selection process.^29-32 Setting term limits for leadership positions is another proposed solution to increasing diverse representation in academic medicine leadership.³³ Based upon findings from the Work-Life Balance domain, we recommend that policies for coverage during periods of leave and support to maintain research productivity such as bridge funding opportunities are needed.³⁴ Moreover, policies about timing of work meetings and after-hours clinical coverage may serve to reduce the burden of work demands after hours and flexible work schedules should be encouraged.^35,36 A “time banking” approach can provide support for extraprofessional tasks and help balance work service demands.³⁷ In the Leader Support domain, we ensured results of this survey were disseminated to pertinent leaders along with resources for training in unconscious bias and allyship.³⁸ Finally, in the Freedom from Gender Bias domain, we recommend the creation of institutional resources and dissemination of information regarding anonymous reporting and support structures.³⁹

Without immediate action to counteract the effects of the COVID-19 pandemic on the workforce in academic medicine, there is risk of undoing previous progress toward gender parity in academic medicine.⁴⁰ Understanding specific factors that detract from the success of women and others who are under-represented in medicine is critical to designing and implementing relevant strategies to improve job satisfaction, performance, and retention in the academic oncology workforce. With these findings, we aim to raise attention to the myriad factors contributing to the under-representation of women faculty in cancer-related specialties at an NCI-designated Comprehensive Cancer Center. This survey will potentially serve as an example for other academic centers who desire to understand the workplace culture, as it impacts academic success for its faculty and as a benchmark for longitudinal measurements at our institution to evaluate trends over time. Furthermore, we propose solutions aimed at mitigating differences in workplace culture in each of the domains to improve gender equity in cancer clinical care and research workforce.

Supplementary Material

Acknowledgment

Contributor Information

Bridget P Keenan, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Amanda Sibley, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA.

Li Zhang, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA. Department of Epidemiology and Biostatistics, University of California, San Francisco, CA, USA.

Alyssa F Westring, Department of Management and Entrepreneurship, Driehaus College of Business, DePaul University, Chicago, IL, USA.

Ana I Velazquez, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Erin M Bank, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA.

Emily K Bergsland, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Lauren Boreta, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Department of Radiation Oncology, University of California, San Francisco, CA, USA.

Patricia Conroy, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Department of Surgery, University of California, San Francisco, CA, USA.

Mariza Daras, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Neuro-Oncology, Department of Neurology, University of California, San Francisco, CA, USA.

Michelle Hermiston, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Pediatric Oncology, Department of Pediatrics, University of California, San Francisco, CA, USA.

Gerald Hsu, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Pamela L Paris, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Department of Urology, University of California, San Francisco, CA, USA.

Sorbarikor Piawah, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Sumi Sinha, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Department of Radiation Oncology, University of California, San Francisco, CA, USA.

Julie A Sosa, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Department of Surgery, University of California, San Francisco, CA, USA.

Mazie Tsang, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Alan P Venook, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Melisa Wong, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

Sue S Yom, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Department of Radiation Oncology, University of California, San Francisco, CA, USA.

Katherine Van Loon, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA. Division of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA, USA.

References