The value of diversity and representation across healthcare professions has been extensively reported. Reference AlShebli, Rahwan and Woon1–Reference Marcelin, Manne-Goehler and Silver4 Inequitable representation across various leadership domains can affect opportunities for promotion and academic career development across different specialties, including the field of infectious diseases. Reference Abdul-Mutakabbir, Arya and Butler5–Reference Ruzycki, Fletcher, Earp, Bharwani and Lithgow16 Persistent inequities are often discussed in the context of sex, race, and ethnicity, however equity gaps can exist across other domains, and are often intersectional. Reference Harris17 Professional medical organizations have varied approaches to addressing representation gaps; intentional initiatives have resulted in more equitable representation in some target groups but not others. Reference Marcelin, Khazanchi and Lyden18 The Society for Healthcare Epidemiology of America (SHEA) is a professional society with >2,000 members that aims to improve public health by supporting antibiotic stewardship among healthcare clinicians and establishing infection prevention measures. 19 The annual SHEA Spring conference (SHEA Spring) highlights healthcare epidemiology, antibiotic stewardship, long-term care, clinical microbiology, public health, and patient safety and quality. It provides unique networking opportunities for people at various stages of career development within their fields of interest. 20 Presenters of oral, poster, and plenary sessions discuss and disseminate data on implementation science, research methods, and novel scientific topics.
The SHEA Board of Trustees reviews the organization’s strategic goals every 5 years. At the most recent review, the board flagged diversity, equity, and inclusion (DEI) as a key priority area. The 2022–2026 SHEA Strategic Map aims to embed DEI within the society’s volunteer structures to advance DEI principles with an emphasis on speaker, author, and leadership roles. 21 This includes the SHEA DEI pledge that “SHEA is committed to diversity within its leadership, membership, committees, conference speakers, and guidelines panels that fully represents the communities served by SHEA members.” 22 The new SHEA Strategic Map and DEI pledge highlight the importance of characterizing the current landscape of selected speakers for SHEA Spring to identify opportunities for improvement. This descriptive study evaluated speaker demographics at SHEA Spring conferences from 2019 to 2022.
Methods
The study was a retrospective descriptive analysis of the demographic composition of speakers at the annual SHEA Spring conferences between 2019 and 2022, excluding 2020 when SHEA Spring was canceled due to the COVID-19 pandemic. Descriptive statistics were performed on summary data. We used the Fisher exact test to compare speaker and member demographics for each year the conference was held. The Fisher exact test was used to compare number of times speaking (1 time vs >1 time) across the period. The unit of analysis was speaker slot rather than individual (because speakers may have spoken several times during a conference) for comparisons. Each speaker slot was considered an independent event. If a speaker spoke more than once within an individual session, the speaker was counted only once for that session.
Self-reported demographics were obtained through responses to conference and member registration surveys. During the study period, SHEA surveys recorded gender as a self-reported binary female or male indicator. Registration surveys allowed people to self-identify race and ethnicity in 1 or more category, including American Indian/Alaska Native (AIAN), Asian, Black/African American, Hispanic/Latino, Native Hawaiian/Pacific Islander (NHPI), White, Other, and No Answer. Indigenous groups were combined as AIAN/NHPI for analysis due to small sample size. Age was captured in 10-year increments and was divided into 4 analytic groups: <40, 40–49, 50–59, and ≥60 years. Specific practice locations submitted by speakers and members were grouped into 1 of 4 US Census regions (Northeast, South, Midwest, and West) or were labelled as international. Professional degree was categorized into 4 main categories: physician, pharmacist, other doctorate, and nondoctoral degree. Speaker professional degree was not available for 2022. All analyses were performed using SAS version 9.4 software (SAS Institute, Cary, NC). P < .05 was considered statistically significant.
This work was evaluated and classified as exempt non–human-subjects research by the University of Nebraska Medical Center Institutional Review Board.
Results
In total, 447 speaker slots were filled by 305 unique speakers over the 3-year period. The SHEA average annual membership between 2019 and 2022 with self-reported demographics included was 55.2% female, 44.8% male, 69.3% White, 21.4% Asian, 6.0% Hispanic/Latino, 2.9% Black, and 0.4% American Indian/Alaska Native/Native Hawaiian/Pacific Islander (AIAN/NHPI). However, almost half of the members did not report a race or ethnicity (Supplementary Table 1 online). SHEA Spring speakers during the same period were 63.5% female, 36.5% male, 68.2% White, 13.3% Asian, 3.8% Black, 3.4% Hispanic/Latino, and 0.8%AIAN/NHPI. Only 13.4% of speakers did not report race or ethnicity. Each year, there were <6 speakers in each Black, Hispanic/Latino, AIAN/NHPI race or ethnicity category.
The proportion of speakers in the age-group categories relative to their proportion of members changed over time (Table 1). In 2019, 49.2% of speakers were aged 41–50 years, compared with 28.6% of members in that age group (P = .0042). By 2022, 35.6% of speakers were aged 41–50 years, compared with 29.3% of members in that age group (P = .051). We detected a statistically significant association between primary-practice setting and speaker or member representation (Table 1) Community and private-practice speakers were underrepresented relative to their proportion of membership, with 7.5% speakers versus 14.3% members in 2019; 7.1% speakers versus 15.2% members in 2021; and 4.3% speakers versus 15.7% members in 2022 (P < .05). Additionally, we detected a statistically significant association between the speaker’s primary practice setting and number of speaking times in a 3-year period (P = .018). For example, among the participants who spoke more than once in the 3-year period, 75.3% were identified as working primarily in an academic setting, compared with just 54.7% of those who spoke only once during the 3-year period (Table 2).
Table 1. Comparison of Demographics Between Speakers and Membership, 2019, 2021, and 2022
Note. AIAN/NHPI American Indian/Alaska Native and Native Hawaiian/Pacific Islander; N/A not available.
a Fisher exact P value for differences between member and speaker proportions, each year.
b Demographics of gender only reported as binary during observation period.
c 2022 membership data did not include highest degree therefore comparison could not be made.
Table 2. Comparison of Demographics with Speaker Times Across the 3 Years
Note. AIAN/NHPI, American Indian/Alaska Native and Native Hawaiian/Pacific Islander
a Fisher exact test P value for differences between 1 year and >1 year proportions.
b Demographics of gender only reported as binary during observation period.
c 2022 Membership data did not include highest degree.
We detected variations in representation of professional degree holders among speakers across the years studied (Table 1). In 2019, there was no association between highest degree and speaker or member representation. However, in 2021 there was a statistically significant association between highest degree and speaker or member representation (P < .0001). For example, pharmacists represented 11.6% of speakers compared with 2.9% of members, and individuals with other doctorate degrees represented 8.5% of speakers compared with 4.6% of members. Although 21.5% of members identified nondoctoral degrees as their highest degree (eg, BSN, RN, MSN, etc), this group represented 11.6% of speakers in 2021. These data were not available for 2022. Additionally, there was a statistically significant association between the speaker’s highest degree and number of speaking times in a 3-year period (P = .0008). For example, among the participants who spoke more than once in the 3-year period, only 4.3% were identified as having “other doctorate degree” compared with 13.4% of participants who spoke only once in the 3-year period (Table 2).
Discussion
This analysis of speakers at SHEA Spring conferences revealed several key findings. The SHEA Spring conference speaker panel demonstrates commendable gender equity in representation; more than half of the SHEA membership are women, and the proportion of women speakers reflects the membership. The organization should continue to monitor these data to ensure that this equity continues in the absence of active sustainable interventions. However, the number of persons self-identifying as minoritized racial or ethnic groups, including Hispanic/Latino, Black, and AIAN/NHPI, are markedly underrepresented in both SHEA membership and in speaker demographics over the 3 years, neither of which approach that of the 2021 US population (19.1% Hispanic/Latino, 13.6% Black, 1.6% AIAN/NHPI). 23 In 2022, 9 (9.6%) of 94 speakers combined identified as Hispanic/Latino, Black, or AIAN/NPHI.
These findings represent an opportunity for SHEA Spring Planning Committee. The SHEA Strategic Map and pledge commitment to diversify the membership and conference speakers will require intentionality to achieve equitable racial or ethnic representation. With intentional efforts (Fig. 1), gradual strides in equitable representation can be attained. The Infectious Diseases Society of America (IDSA) achieved equitable gender representation of invited speakers at the IDWeek national conference through intentional gender-equity initiatives implemented by the program committee. Reference Marcelin, Khazanchi and Lyden18 However, although they observed gradual improvements in racial and ethnic demographic representation of speakers over a 9-year period, these were not statistically associated with the intentional gender-equity initiative, which had a much more dramatic impact on the proportions of male and female speakers. Reference Marcelin, Khazanchi and Lyden18
Figure 1. Putting it all together: organizational and individual actions to enhance equity in speaker representation at national conferences.
Notably, almost half of the SHEA membership did not report demographic data on race and ethnicity. The absence of this data makes assessing the impact of organizational changes more challenging. For example, 3.4% of speakers in 2022 identified as Black, compared with 2.8% of members (Table 1). However, in that year, 49% of member race or ethnicity data were missing; therefore, the available race and ethnicity data likely underestimate these proportions significantly. In response to these identified data gaps, SHEA plans to update and streamline their membership surveys. The trend toward more even distribution in speaker age categories during the observation period is commendable and suggests that perhaps even a generalized DEI pledge may result in better representation among age groups because program committees may search outside their typical professional circles to identify new speakers.
As an interprofessional organization, SHEA is unique in that it attracts multidisciplinary membership and conference attendees. However, the data demonstrate that although there was commendable growth of pharmacist speakers from 2019 to 2021, individuals from community and private-practice settings and those with degrees other than medical or pharmacy degrees (eg, epidemiologists or infection prevention nurses) had fewer opportunities to speak at the SHEA Spring conferences. Individuals from these groups were additionally less likely to be repeat speakers. These are missed opportunities to enhance the engagement of individuals from community and private-practice settings and those with other doctorate degrees, whose valuable experiences may differ vastly from academic physicians and should be represented.
Recommendations
Figure 1 describes recommendations that can be implemented to enhance equity in speaker representation at national conferences, which comprise overlapping organizational and individual actions. Collecting and reporting self-identified demographic data of the society membership and conference participants (including professional degree or practice setting, age, race, ethnicity, gender, and gender identity) and ideally, following the data over a specified period, should be an essential component of any organization’s action plan for equity. Reference Martin12,Reference Marcelin, Khazanchi and Lyden18,Reference Allen, Abdul-Mutakabbir, Campbell and Butler24 In cases where the membership is not reporting this data, the organization should explore and mitigate reasons for missing data and commit to educating their members on both the value of accurate and complete demographic reporting, as well as the anticipated use of these data in driving future organizational growth through a lens of equity (Fig. 1). Within the surveys or requests for demographic information, select demographics should require a response and include several options for maximal inclusivity (eg, “multiple races” or “prefer not to answer” or “does not apply to me”) so that these options are also recorded. Persons from historically excluded groups are underrepresented as infectious diseases specialists, and until this disparity is mitigated, recruitment of racial and ethnic minoritized hospital epidemiologists and antimicrobial stewards, will be challenging. Reference Marcelin, Manne-Goehler and Silver4 Additionally, in professional spaces like medical societies, these individuals may have access to fewer networks and mentorship and decreased likelihood of being approached for opportunities to serve on national committees or to serve as speakers. Reference Allen, Abdul-Mutakabbir, Campbell and Butler24,Reference Titanji, Abdul-Mutakabbir, Christophers, Flores, Marcelin and Swartz25
Although this study did not include program committee demographic data, studies have reported that the makeup of persons on conference planning committees are impactful in influencing the selection of subsequent speakers. Reference Arora, Kaur, Dossa, Nisenbaum, Little and Baxter6,Reference Marcelin, Khazanchi and Lyden18,Reference Casadevall and Handelsman26,Reference Nittrouer, Hebl, Ashburn-Nardo, Trump-Steele, Lane and Valian27 Others who have evaluated conference program committee demographics have reported underrepresentation of speakers from minoritized communities with a predominantly White-identifying program committee. Reference Marcelin, Khazanchi and Lyden18,Reference Earp, Mora and Rozental28 Proactive recruiting efforts from historically excluded groups as speakers, moderators, and program committee members can mitigate these gaps and increase speaker representation at the conference (Fig. 1). Reference Marcelin, Khazanchi and Lyden18,Reference Allen, Abdul-Mutakabbir, Campbell and Butler24 For example, during program committee meetings, intentional conversations about the demographic balance of the proposed speaker list before they are invited would support the mutually beneficial goals of equity and excellence in each session. If the proposed session does not achieve these goals, program committee members (and/or society leaders) should widen the net to identify individuals whose contributions to the session would achieve them. Furthermore, adding established DEI goals to organizational dashboards linked to publicly reported equity data (including demographics of members, speaker, committee membership/leadership, journal editors, etc) can reinforce accountability and allow organizations to identify evolving opportunities to center equity in their strategic vision and mission (Fig. 1). Reference Martin12,Reference Allen, Abdul-Mutakabbir, Campbell and Butler24
Organizations and conferences can implement additional structural changes to support the stated goal of achieving equity across speaker representation. These include developing and enforcing a transparent speaker policy accessible on the organization or conference website, clearly incorporating goals and outcomes as part of the case for equity in conference proceedings (Fig. 1). Reference Martin12,Reference Marcelin, Khazanchi and Lyden18 At the conference itself, additional policies to establish an inclusive environment may encourage potential speakers to accept future invitations, Reference Martin12,Reference Marcelin, Khazanchi and Lyden18,Reference Allen, Abdul-Mutakabbir, Campbell and Butler24,Reference Sheffield, Marcelin and Cortés-Penfield29 particularly if they may belong to a historically minoritized group (Fig. 1). These may include zero-tolerance antiharassment or antidiscrimination policies, Reference Martin12,Reference Marcelin, Khazanchi and Lyden18 affordable conference prices or reasonable reimbursement/honorarium, and supportive parental accommodations including childcare and lactation facilities (Fig. 1). Reference Sheffield, Marcelin and Cortés-Penfield29 The equity-minded recommendations for improving racial/ethnic representation of speakers apply to professional diversity, and likewise an additional individual action item for physicians working in teams with other health professionals in antimicrobial stewardship and healthcare epidemiology can nominate their colleagues to speak on their shared topics or request copresentation to provide space for others to have access to the stage (Fig. 1).
The use of self-identified demographic data in this study mirrors the methods of the largest similar study of infectious diseases conference demographics. Reference Marcelin, Khazanchi and Lyden18 This is an important strength because many reports on speaker representation at national conferences rely on investigator assignment of characteristics. However, because these self-identified demographic characteristics were optional on membership surveys, there was a very high missing rate in racial or ethnic demographics selected by members, whereas the missing rates of the speaker demographics was or ethnicity data beyond general underrepresentation compared with the US census, or the ability to conduct intersectional analyses (eg, race and gender interactions). This is a notable limitation that suggests that a directed approach to encouraging members to self-report more accurate racial/ethnic demographics is warranted. Furthermore, the data set did not allow for disaggregation of the racial groups, and the small sample size limited the ability to analyze AIAN and NHPI groups separately. Additionally, during the observation period, the gender demographic questions were binary (male or female only), which limits the ability to include transgender, nonbinary and gender-fluid individuals. However, in 2022 the SHEA membership portal was enhanced with expanded gender identity options; therefore, future speaker representation assessments should include these data. The COVID-19 pandemic affected the conference format over the study period. The 2019 conference was held in person, the 2020 conference was cancelled, and both the 2021 and 2022 conferences had hybrid formats. Insufficient data were available to support conclusions regarding any potential shifts in demographic representation of speakers in the context of these pandemic-related conference changes. Directions for future studies include evaluating speaker equity across session categories, including high-impact sessions (eg, plenary sessions) and moderator and program committee demographics. Furthermore, as a multiprofessional organization, including speakers who are from community settings to speak on patient-centered topics is beneficial to both healthcare professionals and the communities they serve. Therefore, understanding the proportion of invited speakers that comprise this group is also important.
In conclusion, the SHEA Spring conferences demonstrated gender equity in speaker representation, had more equitable age representation, and an increase in pharmacist representation among speakers over time. However, practitioners from community settings and those with nondoctoral degrees remain underrepresented relative to SHEA membership. Racial and ethnic minoritized individuals remain underrepresented as members and speakers compared with the general US population. Findings from this assessment of speaker demographics between 2019 and 2022 offer targets for intentional interventions that are needed to consistently achieve equitable speaker representation across multiple demographic groups at the SHEA Spring conferences.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2023.207
Acknowledgements
The authors thank SHEA staff for providing us with the membership and SHEA Spring conferences demographic data.
Financial support
No financial support was provided relevant to this article.
Conflicts of interest
J.R.M. is a volunteer member of the IDSA Board of Directors and was a member of protocol leadership for the NIH/NIAID/CoVPN vaccine study CoVPN 3006/Prevent COVID U and received salary support for this activity, not related to this manuscript. I.B.K. was a volunteer member of the SHEA Board of Trustees from 2020 to 2022 and is a volunteer for the SHEA Leadership Development committee. I.B.K. reports grant funding from Center for Disease Control and Prevention (CDC) Epicenter, National Institutes of Health (NIH), Bristol Myers Squibb Foundation, and receives consultancy fees from IPEC Experts and Wayfair, all unrelated to this manuscript. Z.W. is a volunteer member and Chair of the SHEA Education Committee and is Co-chair of the 2023 SHEA-sponsored Advancing Health Equity through Antimicrobial Stewardship Workshop. She also participated in the advisory board for Entasis Therapeutics and received an honorarium in 2022. J.A.M. has participated in advisory board for Shionogi and Entasis Therapeutics and has received an honorarium and is also an appointed member of the CVSH National Health Equity Advisory Board. All other authors have nothing to disclose.
