INTRODUCTION

Prior to the COVID-19 pandemic, gender inequities permeated medical careers, including salaries1, 2, leadership roles,3 academic rank,4 distribution of domestic and caregiving responsibilities,5 speaking opportunities at national meetings,6 and authorship.7,8,9 Hospital medicine is no exception.10, 11 The COVID-19 pandemic exacerbated gender inequities—publications by women physicians disproportionately declined during the pandemic,12,13,14,15 women shouldered a disproportionate burden of pandemic childcare and eldercare duties,16 and women physicians’ career promotion and leadership opportunities suffered.17 These trends threaten the viability of the physician workforce in general and hospitalist workforce specifically, given women represent a large proportion of the field.18

Much of the literature to date on gender inequities12,13,14,15 has focused on describing and analyzing existing disparities. Comparatively less literature has focused on the solutions to mitigate these inequities. Given the pandemic’s impact on gender disparities overall, we aimed to (1) understand the pandemic’s impact on women hospitalists and to (2) uncover possible solutions. To accomplish this, we held semi-structured focus groups conducted through an existing research consortium19 in order to capture the experiences and perceptions of hospitalist clinicians, leaders, and scholars spanning institutions, positions, and ranks.

METHODS

Study Design

On 5/21/2021, we conducted 6 semi-structured virtual focus groups with hospitalists from Hospital Medicine Reengineering Network (HOMERuN), a national consortium of over 50 academic hospital medicine programs focused on improving quality and implementation science.19 During the pandemic, HOMERuN held regular conference calls to discuss challenges and responses. The Institutional Review Board at the University of Texas at Austin approved this study as IRB exempt (STUDY00003031).

Setting and Participants

Focus group participants were recruited through HOMERuN email solicitations.19 Participants were notified prior to and at the beginning of the HOMERuN call of the intent to conduct and record focus groups (via Zoom). HOMERuN call attendees typically include physicians, advanced practice providers, researchers, and leaders with varying degrees of professional experience. No incentives to participate were offered and refusal to participate was the only exclusion criteria. Up to 4 participants were distributed to each group with the intent to optimize diversity of unique sites in each group.

Participants were provided the following discussion prompts: (1) In what ways did the pandemic impact women physicians uniquely? (2) What issues need to be addressed to better support women in medicine in the future? The focus groups lasted 30 min. At the end of the session, all participants reconvened and reported major discussion points to the group at large. Twenty-two participants (21 physicians, 1 advanced practice provider) from 13 unique academic medical centers across 12 states contributed to focus group discussions. Fourteen self-identified as women and 8 as men. Eight were Assistant Professors, 6 were Associate Professors, 6 were Professors, and 2 did not disclose. Five were division chiefs (Appendix 1).

Research Team and Reflexivity

Members of our research team included physicians and nonphysicians (A.K.) with expertise in health information technology. Several team members have experience with both traditional (MB, AK, LL, GA) and rapid qualitative methods (MB, LL, AK). All team members that moderated the focus group sessions were trained to follow rapid qualitative methods.20

Moderator Preparation

The moderator guide and discussion prompts (Appendix 2) were developed by the authors. These prompts were developed from authors’ lived experiences and subject matter expertise and as well as a review of the literature.

Methodological Orientation and Theory

We used a phenomenological qualitative methodological approach, identifying the essence of participants’ reported experiences and applying a pragmatic framework to identify solutions for the challenges described by the participants.21

Data Collection

All focus groups were recorded with participants’ permission, and facilitators (GA, JB, MB, SK, AL, KN, and RP) took field notes to supplement recordings. Field notes and recordings were used for analysis. Due to the concurrent nature of the focus group discussion, we were unable to formally evaluate for data saturation during the virtual session.

Rapid Qualitative Analysis Methods

We have previously described the rapid qualitative methods utilized in detail20 and discuss their application here in the “Discussion” (“Strengths and Limitations”). We used standardized summary templates20, 22, 23 to identify the main discussion points from each focus group (Appendix 3). To optimize for consistency, GA, MB, JB, SK, AL, KN, and MS independently created a summary template of a single focus group’s session, utilizing the recording and the moderator’s field notes to summarize each unique discussion point. The authors compared templates to ensure consistency in how discussion points were documented across reviewers. Once consensus was achieved, the remaining focus groups sessions were distributed amongst GA, MB, SK, AL, KN, and MS and summary templates were generated for each focus group. JB double-checked all templates and compared them against the original recordings for consistency.20

Summary templates were organized (KN) into an analysis matrix.22, 24 Each row in the matrix represented a focus group and each column referred to a unique point explored during the focus group. The authors then met multiple times to discuss findings and reach consensus regarding major themes identified from the matrix. Member checking25 was also conducted; 2 focus group participants who did not moderate or analyze focus groups reviewed the themes to confirm these reflected their experiences as a participant. A written summary was also emailed to the research network.

RESULTS

Four key themes emerged from rapid qualitative analyses of the focus group sessions: (1) the pandemic exacerbated existing gender inequities, (2) women’s academic productivity and career development were negatively impacted, (3) women held disproportionate roles as caregivers and household managers, and (4) institutional pandemic responses were often misaligned with workforce needs, especially those of women hospitalists. Multiple interventions were also proposed. The main themes are discussed below with accompanying exemplar quotes (Table 1). Proposed solutions and exemplar quotes are provided (Table 2).

Table 1 In what ways did the pandemic impact women physicians uniquely?
Full size table
Table 2 What issues need to be addressed to better support women in medicine in the future?
Full size table

Theme 1: COVID-19 Exacerbated Existing Gender Inequities

Participants noted that pre-existing gender inequity worsened during the pandemic. Some participants noted that women hospitalists were disproportionately represented at junior faculty levels and in clinical tracks, and therefore were deployed for more clinical work during COVID-19 surges, detracting from research, teaching, and leadership. Some noted a subset of women hospitalists did step into “crisis” leadership roles, but viewed these as undervalued as academic currency for promotion. Others perceived that physicians nominated as visible experts at both local and national levels were disproportionately men, despite availability of equally qualified women.

Theme 2: COVID-19 Negatively Impacted Academic Productivity and Career Development for Women Physicians

Participants referenced data, personal experiences, and observations that academic productivity of women physicians during the pandemic decreased, with fewer publications, grants, and other research endeavors. Some participants perceived that individuals with greater dedicated non-clinical time, often predominantly men, built academic momentum by producing scholarship related to work done by frontline physicians, often predominantly women.

Some participants felt the pandemic decreased women physicians’ ability to find mentorship, noting women’s networks were already smaller before COVID-19. Participants reported that virtual meeting platforms had a mixed impact on professional networks. Some felt these were less effective than in-person meetings in providing visibility, credibility, and career advancement. Many participants commented the shift to virtual work negatively impacted work-life balance. However, others acknowledged virtual meetings allowed greater flexibility, which could allow expansion of one’s professional network.

Theme 3: The Disproportionate Role of Women as Caregivers and Household Managers Exacerbated the COVID-19 Burden on Women Physicians

While participants noted that school closures and disruption of caregiving services affected all faculty with such responsibilities, many noted that women were more significantly impacted than their male counterparts. Several noted this issue predated but was augmented by the pandemic, citing the “3rd shift problem,”26, 27 a concept referenced by many participants—the pandemic increased clinical work burden (1st shift), drastically increased domestic/caregiving responsibilities for women physicians (2nd shift), resulting in even less time for the “3rd shift,” i.e., time allocated to career and professional development.

The level of caregiving burden was felt to be influenced by the number and age of dependents, financial resources, partner’s occupation, and availability of outside support. Participants noted the higher cost of caregiving support for hospitalists, because affording services to accommodate a hospitalist’s schedule was challenging due to 12-h shift requirements. In addition, the necessity for isolation during the pandemic limited access to usual networks of support from family and friends.

Theme 4: Institutional Initiatives Were Often Misaligned with Women’s Needs

Many participants noted that institutional interventions were often misaligned with women physicians’ needs and did not address root causes. Examples of interventions included concierge services such as grocery procurement, dog walking, and child or eldercare centers. These interventions were short-lived and often did not align with typical hospitalist schedules—e.g., care centers closed before the end of a shift. Moreover, repeated surge staffing that hospitalist groups implemented during the pandemic were out of sync with fixed availability of support services. Some participants viewed institutional interventions as public relations measures that did not meet their actual needs.

SOLUTIONS FOR THE FUTURE

Participants reflected that the challenges of gender equity in medicine were indicative of overall societal inequities. However, multiple participants felt that health care systems should be proactive in addressing root causes of gender inequities so that local change could help spur broader societal changes. Participants shared various strategies employed by their institutions and brainstormed high-level solutions as summarized in Table 2 and discussed below.

Solution Theme 1: Implement Benefit and Workforce Solutions to Relieve the Disproportionate Burden of Caregiving Duties

Participants suggested institutions reduce disproportionate caregiving burdens by designing benefits that align better with a typical hospitalist schedule—e.g., backup caregiving services that can flex to hospitalists’ shifts. Some institutions implemented onsite daycare with flexible hours; others gave funds towards household management services.

Participants also suggested redesigning the hospitalist work model, noting that working 7 days on, 7 days off is difficult for those with caregiving duties. Some suggested work schedules structured around 8 h shifts instead of 12, or scheduling providers based on caregiving or partner needs. Participants also proposed institutions schedule networking opportunities at times that respect typical caregiving hours.

Solution Theme 2: Redesign Workforce Structures to Standardize Leave and Increase Resiliency

Multiple participants noted the continued stigma around leave such as maternity leave and raised concerns that women solely taking parental leave could be detrimental to their career advancement. Participants suggested that leave be normalized and built into workforce structures for all hospital medicine physicians, whether this be for caregiving responsibilities, family, or medical leave. Participants noted most hospitalist groups deploy their workforce with an assumption that each person fulfills 100% of planned shifts. The clinical work of a person on leave falls on colleagues already working 100%—essentially overtime. A solution employed by some institutions was the implementation of an external workforce, such as locum tenens or additional back-up providers, to cover clinical duties for providers on leave. Setting aside funds to offer vacant shifts as moonlighting opportunities was another solution. Participants noted the pandemic highlighted the need to restructure the depth, number, and flexibility of the hospitalist workforce for increased future resiliency.

Solution Theme 3: Incorporate Metrics to Better Value Clinical and COVID-19 Contributions in Pathways to Promotion

Participants advocated that the impact of the pandemic be incorporated into modified frameworks of academic promotion pathways. One institution adopted a COVID-19 section in curriculum vitaes, highlighting increased clinical responsibilities undertaken during the pandemic as a new form of currency for promotion. Other institutions allowed members to pause the promotion clock, although some participants worried this could inadvertently cause a cohort of women to fall behind. Many participants advocated that excellent clinical work and education should have similar academic weight as traditional research.

Solution Theme 4: Implement Transparent Processes to Promote Equity in Salaries and Non-clinical Time

Participants emphasized the need for transparency to combat salary inequities, with concern that inequities would be exacerbated by the pandemic as women’s career trajectories stalled. Some focus groups discussed the need for transparency with regard to non-clinical time allocation (e.g., dedicated research time), arguing that increased transparency could spur interventions to address root causes of any uncovered discrepancies.

Solution Theme 5: Advocate for Increased Diversity, Including Greater Representation of Women Physicians, in Leadership Positions

Participants noted the traditional model of academic career advancement relies on the concept of “sweat equity”—extra time and effort devoted to academic endeavors, often outside of full clinical schedules. Participants noted this model may preferentially favor those who have the means and flexibility to put in extra hours, selecting against those with caregiving responsibilities who are disproportionately female.

Focus groups discussed the need for emphasis on gender equity in career promotion and leadership as well as for proactive sponsorship of women physicians. Participants noted that ongoing intentionality was needed in appointing women to decision-making committees, promoting them to leadership roles, and highlighting their expertise on panels, as keynote speakers, and editorial boards. Participants recommended reinvigoration of women’s leadership initiatives and mentorship programs. Others suggested implicit bias training to help decrease biases.

DISCUSSION

A convenience, cross-sectional sample of academic hospitalists noted that the pandemic highlighted and exacerbated existing gender inequities, increased clinical and caregiving responsibilities at the expense of academic endeavors, and had an overall negative impact on women hospitalists’ academic productivity, career development, and networking opportunities. Systemic institutional interventions with iterative stakeholder feedback were felt to be needed to address these inequities.

Our findings mirror literature indicating that existing gender disparities in caregiving and domestic responsibilities worsened due to the pandemic.5, 28, 29 The reasons women continue to shoulder these duties disproportionately is multifactorial and includes gender and societal norms. Even in dual physician couples, these duties disproportionately fall to women.5, 29 These increased duties, in addition to increased clinical loads during COVID-19 surges for frontline hospitalists specifically, come at the expense of academic endeavors, contributing to the observed decreased academic productivity of women physicians.12,13,14 These studies paired with this work highlight the importance of addressing the disparities that impact women physicians’ career productivity and advancement, including building larger support networks, standardizing leave, and building more robust caregiving infrastructures.

The literature also suggests that inequities can be mitigated when processes are transparent and built on best practices.6, 30 For example, the Society of Hospital Medicine implemented an open call process that essentially eliminated gender inequities in speakers for lectures.6 Similarly, salary and other inequities can be eliminated through intentional, structured processes within institutions’ control as recently demonstrated at the University of Colorado.30 An additional insight from this work is that proposals for systemic change should incorporate iterative input from faculty impacted by the policies, or else run the risk of being misaligned with workforce needs.

This study also demonstrates the utility of cross-institutional focus groups to brainstorm and share possible solutions. Specific interventions shared by our cohort to help mitigate the inequities faced by women hospitalists included caregiving benefits that can flex to a hospitalist’s schedule or, instead, restructuring shift schedules to allow for more flexibility. Additionally, participants advocated for leave to be standardized and built into hospitalist schedules, noting that leave is often still stigmatized.31 Part of the stigma may stem from hospitalist groups often being understaffed at baseline for clinical workloads, which not only is associated with negative outcomes,32,33,34 but results in the clinical workload of a person on leave effectively falling on colleagues. Instead, the structure, size, and deployment of the workforce should be built around an assumption of standardized leave. Participants noted the pandemic highlighted the need overall to create buffers to build more resiliency into workforce structures, such as hiring additional providers or setting aside moonlighting funds for open shifts, so that hospitalist groups can weather stressors.

Overall, this work highlights the effects of the COVID-19 pandemic on women hospitalists, as perceived and experienced by a cross-sectional mixed-gender convenience sample. Notably, many in our cohort felt the pandemic worsened pre-existing inequities; it will be important to further explore ongoing inequities exposed by the pandemic, find solutions, and implement them to create a more resilient workforce. Many of solutions proposed would be beneficial to all hospitalists, not just women, in improving work/life balance, decreasing burnout, and providing alternate pathways to promotion.

Strengths and Limitations

This study has several strengths. It utilized rapid qualitative methods,22, 23, 35,36,37,38,39,40,41,42,43 which are useful in dynamic situations and real-world application, allowing for exploration of the themes raised during focus group discussions. Rapid qualitative analysis yields results that provide a high-level overview of the data and aids in understanding future areas of focus; it has been shown as reliable as traditional qualitative analysis.39

Additionally, this work highlights the experiences and perspectives of members of the hospitalist community from 12 academic institutions across the USA, spanning roles from junior faculty to division chiefs. We included men in our focus groups in order to capture diverse perspectives on this topic and allow for a broader understanding of the issues faculty face.

Our study is a convenience sample and subject to sampling and participation bias; the views of those who participated may differ from those who did not. While including multiple institutions, the number of participants was small and comprised of a subset of HOMERuN network members, which represents mostly academic institutions and physicians (Appendix 1 and 4). We did not collect data around years of experience. As health care workers and researchers embedded in the hospital setting, we may have brought inherent bias to the analysis. To minimize this, we used a standardized process to ensure the templated summaries were calibrated and utilized group discussion in the analysis process. We also incorporated member checking exercises as described in the Methods. Focus group approaches can be susceptible to groupthink and intimidation; however, they are a recurrent format in HOMERuN network calls and we took precautions to create an inclusive environment for all participants with moderators skilled in facilitation. We felt the focus group approach fostered rich discussion and allowed us to understand multiple different perspectives including those from men and women.

CONCLUSION

Hospitalists perceived and experienced that women physicians faced negative professional impacts from the pandemic in multiple domains including clinical workload, leadership opportunities, and career development, while also facing the brunt of caregiving duties. Many opportunities to improve workplace conditions for women physicians exist, but will require more than token initiatives. Instead policy and systemic changes are needed—with iterative engagement of women physicians.