The Future of Physician Advocacy: Attitudes and Intentions of U.S. Medical Students

Background: Advocacy is a core component of medical professionalism. Medical organizations frequently emphasize advocacy, particularly around just resource distribution and social determinants of health, but it is unclear how educators can best prepare trainees for this professional obligation. Objective: We sought to assess medical students’ attitudes toward advocacy, including activities and issues of interest, and to determine congruence with professional obligations. Design, Participants, and Measures: A cross-sectional, web-based survey probed U.S. medical students’ attitudes toward advocacy around 7 issues directly related to health (e.g. nutrition/obesity, addiction) and 11 indirect determinants of health (e.g. housing, transportation, education). Descriptive statistics, Kruskal-Wallis tests, and regression analysis investigated associations with gender, race, political identi�cation, and intended future �eld. Key Results: Of 240 students completing the survey, 53% were female; most were white (62%) or Asian (28%). Clinical/non-primary-care was the most common intended future �eld (61%). Most agreed it is very important that physicians encourage medical organizations to advocate for public health (76%) and provide health-related expertise to the community (57%). More participants rated advocacy for medical issues as very important (e.g. drug addiction [83%], nutrition [81%]), compared to issues with indirect connections to health (e.g. national security [22%], transportation [36%], criminal justice [40%]) (p<0.001). Generally, liberals and non-whites were more likely than others to value advocacy. Conclusions: Medical students reported strong interest in advocacy, particularly around health issues, consistent with professional standards. Many attitudes were associated with political a�liation and race. To optimize future physician advocacy, educators should provide opportunities for learning and engagement in issues of interest.


Introduction
2][3] Medical organizations and codes of conduct frequently emphasize the importance of physician civic engagement. 4,5The American Medical Association (AMA), for example, urges physicians to "advocate for social, economic, educational, and political changes that ameliorate suffering and contribute to human well-being." 6Similarly, the Physician's Charter asserts that "the medical profession must promote justice in the health care system, including the fair distribution of health care resources." 7And physicians themselves almost unanimously agree that community participation, political involvement, and collective advocacy are important professional duties. 8t few physicians actually engage in these tasks. 1 In a 2004 survey, only a quarter of U.S. physicians reported participating politically (apart from voting) on local health issues. 8,9Indeed, physicians take part in community and political activities less frequently than the general population and other professionals with similar levels of education and income. 10,11While reasons for physicians' low levels of engagement likely vary, 1 there is clearly more work to do in equipping physicians to participate in and contribute to civic life.
Educating medical students about their professional responsibility to advocate for health-related issues is essential to promoting more robust physician civic engagement in the future.Yet relatively little is known about students' awareness of or interest in these vital topics.We therefore set out to understand medical students' attitudes around civic engagement, including speci c issues of interest, and to determine congruence with professional obligations.We hypothesized that students would express stronger interest in advocacy around issues directly related to health and medical care (e.g.nutrition, addiction, care access) but lesser support for engagement around indirect determinants of health (e.g.transportation, education, economic inequality).

Survey Administration
We conducted a cross-sectional, web-based survey of U.S. medical students.Participants were recruited from Student Doctor Network (SDN), a non-pro t, online forum for current and future healthcare students and professionals.The survey link was posted on SDN's Facebook and Twitter pages and on the SDN website's homepage, online forums for allopathic and osteopathic medical students, and blog; the blog post was also distributed to self-identi ed medical students who had previously opted to receive SDN emails.The survey was anonymous, but participants could opt to provide their email address to enter a lottery for 1 of 20 $100 gift cards.
The survey launched on August 13, 2019, and closed on October 15, 2019.Responses were collected and managed using Research Electronic Data Capture (REDCap), a secure, web-based software platform hosted at Memorial Sloan Kettering Cancer Center (MSK) 12,13 .Our ndings are reported according to the Checklist for Reporting Results of Internet E-Surveys 14 (Supplemental Digital Appendix 1).The study was reviewed by the MSK Institutional Review Board and deemed exempt.

Survey Instrument
We adapted some survey items from prior studies 8,9,15−19 and developed additional, novel questions focusing on study objectives (Supplemental Digital Appendix 2).A response to each question was required in order to proceed to the next.The survey included demographic items and several measures assessing participants' interest in following or becoming involved in healthcare policy.Two measures tested their views of physicians' responsibilities to patients around healthcare access and costs (providing care regardless of patients' ability to pay, being aware of the overall costs of care they provide). 9,19ditional items gauged participants' attitudes toward 3 forms of physician civic engagement: 8, 19,20 Community participation (providing health-related expertise to local populations), individual political engagement (being politically involved around health issues at the local, state, or national level), and collective advocacy (encouraging medical organizations to advocate for public health).
The survey also assessed participants' support for individual or collective advocacy by physicians around 18 public priorities, adapted from recent national surveys of the U.S. population. 17,18Seven issues directly related to health and healthcare: healthcare costs, healthcare coverage for the uninsured, Medicare/Medicaid/Social Security, drug addiction and treatment, abortion laws/reproductive issues, nutrition/obesity/food safety, and disability rights.Eleven additional issues had connections to or implications for health: education, 21,22 housing/homelessness, 23,24 transportation, 25,26 immigration, 27,28 LGBTQ (lesbian, gay, bisexual, transsexual, and queer/questioning) issues, 29,30 racial issues, 31,32 economic issues, 33,34 environmental issues, 35,36 human rights, 37,38 crime/criminal justice, 39,40 and military/national security issues. 41,42sponse options included Likert scales for agreement (strongly agree, agree, disagree, strongly disagree) and importance (very important, somewhat important, not important).We piloted a preliminary version of the survey with 15 medical students and internal medicine residents at Weill Cornell Medical Center and made minor changes to the survey based on their feedback and responses.

Analysis
We used descriptive statistics to summarize participants' demographics and attitudes.We used census zones to determine geographic region of participants' schools.We used the Kruskal-Wallis test to evaluate associations between demographic characteristics (including gender, race, year in school, political identi cation, and anticipated future eld) and attitudes around healthcare policy and forms of civic engagement.We also created a composite civic-mindedness score for each participant by averaging the strength of their responses (using scores of 1 for "not important," 2 for "somewhat important," and 3 for "very important") to all 18 public-priorities questions (overall score); we similarly generated composite scores for the 7 issues directly related to health and healthcare (medical score), as well as for the 11 issues addressing indirect or social determinants of health (social score).We used univariate tests and multiple linear regression to evaluate associations between demographic characteristics and overall, medical, and social scores.All analysis was performed in Stata 14.2. 43

Results
There were 815 visitors to the SDN postings linked to the survey.Three hundred sixty-one unique individuals accessed the survey link, and 356 completed the rst page to determine eligibility (based on attendance at an accredited U.S. medical school).Of 277 eligible participants, 240 completed the survey (view rate 44%; participation rate 77%; completion rate 87%). 14(Supplemental Digital Appendix 1) Participant characteristics are shown in Table 1.Eighty-seven percent of participants were enrolled in MD programs.The majority was 25-34 years old (63%) and white (62%), and about half were women (53%).Participant gender and race were similar to characteristics of U.S. medical students overall. 44All geographic regions and years of medical school were represented, with slight overrepresentation of rst-year students.

Interest and intentions around healthcare policy
Most participants in our study (181, 75%) were members of at least one medical organization addressing healthcare policy issues (e.g.American Medical Association, American Medical Student Association, American Medical Women's Association).
A large majority reported following healthcare policy in the news (82%) (Table 2).Eight-seven percent somewhat (28%) or strongly disagreed (59%) that healthcare policy will have little or no effect on their care of patients, with liberals twice as likely as conservatives to hold these views (p<0.001).Most also planned to become involved (80%) or take leadership (65%) in healthcare policy issues as physicians; liberals and those intending to enter primary care were 2 and 6 times likelier, respectively, than others to express strong interest in policy involvement (p<0.05).

Responsibilities around healthcare costs and access
The survey probed participants' beliefs about physicians' responsibilities around healthcare access and costs (Table 3).
Three-quarters agreed it is very important for physicians to know the overall cost of the care they provide.A large majority (81%) believed that it was very important for physicians to provide necessary care regardless of the patient's ability to pay, although liberals (87%) were more likely than independents (72%) and conservatives (57%) to hold this view (p<0.001).

Public roles: Collective, community, and individual
The survey assessed participants' attitudes toward 3 forms of civic involvement by physicians: collective advocacy, community participation, and political engagement (Table 3).Most (76%) said that it was very important for physicians to encourage medical organizations to advocate for the public's health.This attitude correlated with political identi cation, with 82% of liberals, 64% of independents, and 57% of conservatives holding this attitude (p<0.001).More than half (57%) reported that it is very important for physicians to provide health-related expertise to local community organizations.Firstand second-year students and those intending to enter primary care were more likely than others to express this opinion (p<0.05).Fewer than half (45%) agreed that it is very important for physicians to be politically involved in health-related matters at the local, state or national level.Liberals were more likely than independents and conservatives to hold this view (p<0.001).
Overall civic-mindedness scores, averaging the strength of participants' responses (using scores of 1 for "not important," 2 for "somewhat important," and 3 for "very important") to all 18 issues, had a mean of 2.5 and median of 2.6 (IQR: 2.28-2.83).Medical scores, based on participants' assessments of the 7 issues directly related to health and healthcare (healthcare costs, healthcare coverage for the uninsured, Medicare/Medicaid/Social Security, drug addiction and treatment, abortion laws/reproductive issues, nutrition/ obesity/food safety, and disability rights), had a mean of 2.7 and a median of 2.9 (IQR: 2.57-3.00).Social scores, based on participants' responses to the 11 issues with indirect connections to or implications for health (education, housing/ homelessness, transportation, immigration, LGBTQ issues, racial issues, economic issues, environmental issues, human rights, crime/criminal justice, and military/national security issues), had a mean of 2.4 and a median of 2.5 (IQR: 2.00-2.82).(Table 4) Regression analysis indicated that liberal participants had higher overall, medical, and social scores than conservatives (p<0.01).Nonwhite participants had higher medical scores than whites (p<0.05);they also had higher overall and social scores, although this trend was not statistically signi cant.Kruskal-Wallis tests showed women and those intending to enter primary care to have higher overall and social scores than other participants (p<0.05).However, our regression analysis did not nd gender or intended future eld to be a signi cant predictor, suggesting that political identi cation was driving these differences and capturing the variance in our model.(Table 5)

Discussion
Our ndings can assist medical schools in preparing the next generation of physicians to more actively engage in civic life.
Nearly all participants in our study expressed nascent interest in healthcare policy and civic engagement.Anticipating its relevance to their future practice, most stayed abreast of healthcare policy news and anticipated some degree of policy involvement as physicians.These results suggest that most students would welcome greater opportunities in medical school to learn about and participate in health policy issues.
Participant views of speci c civic roles were more variable.Large majorities expressed a strong sense of professional obligation around the just provision of clinical care, from being cost-aware to treating patients who are unable to pay.
Notably, despite efforts to develop novel curricula, [45][46][47] these fundamental public health issues remain underemphasized in medical school environments.Our ndings should encourage further efforts, as they suggest that many students would appreciate instruction around the challenges of caring for patients in a system where costs are frequently high and nontransparent and many patients are un-or under-insured.
Fewer than half of participants saw civic engagement by individual physicians, whether through community service or political involvement, as crucial.Medical student attitudes, then, stand in contrast to broad recognition of the important role of individual physicians in driving health system changes that will better meet the health needs of society. 48More transparent role modeling of civic engagement by educators and better student access to meaningful advocacy opportunities could lead to greater student awareness of the importance of individual engagement and greater commitment to future action. 49st participants strongly supported public health advocacy by medical organizations.This nding indicates that medical students would likely be interested in learning about how medical organizations have advocated around public health issues, as well as the obstacles (internal and external) to success.This would give students crucial insight into organized medicine's mixed record in challenging or perpetuating the inequalities that pervade the U.S. healthcare system.It could also not re ect these potential shifts; future studies should evaluate the in uence of these events on the attitudes of students and physicians.

Conclusion
Medical students report keen interest in civic engagement and advocacy, particularly around issues directly involving health or healthcare services, which is generally consistent with professional standards.Many attitudes and interests are associated with political a liation, race, gender, and intended future eld.To optimize future physician advocacy, educators should provide opportunities for student learning and engagement in these vital matters.

Table 5
Univariate associations between student characteristics and civic engagement is a list of supplementary les associated with this preprint.Click to download. This