Cascading Mentorship as Advocacy Training in Undergraduate Medical Education: A Mixed Methods Evaluation

Physicians are in a position of great inuence to advocate for health equity. As such, it is important for physicians-in-training to develop the knowledge and skills necessary to full this role. Although various undergraduate medical programs have implemented health advocacy training, they often lack experiential learning and physician involvement. These aspects are foundational to the Advocacy Mentorship Initiative (AMI) which utilizes cascading mentorship as a novel approach to advocacy training. Medical students develop advocacy competency as peer mentors to youth raised in at-risk environments, while also being mentored themselves by physician residents. We aim to determine whether there are specic advantages to utilizing cascading mentorship to facilitate the attainment of advocacy competencies in undergraduate medical education.


Introduction
The role of a physician as Health Advocate is a central pillar in the development of competencies deemed necessary to practice medicine effectively, as described by the CanMEDS framework [1]. The CanMEDS framework has been adopted by medical programs in dozens of countries [2]; while other countries have included advocacy to established competency frameworks. For example, as "Systems-based Practice" by the Accreditation Council for Graduate Medical Education (ACGME) Core Competencies in the United States [3]. By developing skills in the Health Advocate role, medical students learn to identify and address inequities in the social determinants of health and the manner in which these in uence health outcomes in various populations [4]. Importantly, preclinical advocacy training provides the unique opportunity for students to develop their identity as advocates, unrestricted by the realities of clinical responsibilities that can overshadow patient care [5]. Nevertheless, the effective integration of advocacy training in undergraduate medical education presents various challenges [6], and advocacy training has historically been an underappreciated component of medical training [7].
Experiential learning and community engagement in health advocacy training has been implemented in various undergraduate medical education programs [8]. Several medical schools, including the University of Toronto (UofT), have integrated community-based service learning (CBSL) as a tool for medical trainees to engage in experiential learning opportunities while also addressing community needs [9]. CBSL has been described as an effective method to teach social determinants of health [10] and provides trainees with the opportunity to directly observe health inequities [11,12]. Appreciating social determinants of health and health inequity are foundational to engage in health advocacy [4].
One example of CBSL involves mentoring youth raised in at-risk environments. Various studies have shown signi cant positive impact upon mentees [13], particularly for at-risk youth [14]. Furthermore, mentors gain valuable skills and insights, which can extend to areas pertaining to effective advocacy training. One study found that through mentoring at-risk youth, university students had an increased sense of civic responsibility, and a belief that "it is every person's responsibility to use their time and talents to help solve social problems" [15]. A thematic analysis of medical students' re ections while mentoring youth from a Native Hawaiian community found that the experience developed their skills in establishing relationships, self-re ection, communication, compassion, as well as led to a greater understanding of health inequity [16]. Literature pertaining to the extent to which mentoring may serve as an opportunity to develop advocacy skills and competencies remains relatively lacking, especially in the context of medical education.
Medical students are also often mentees in mentoring relationships with senior medical trainees and physicians. A review of such mentorship programs identi ed bene ts for medical student mentees in clinical knowledge and skills attainment, professional development, communication skills development, exposure to subspecialties, career guidance, and opportunities for networking and research involvement [17]. However, development of health advocacy competencies in medical students through physician mentorship remains a poorly described area of study. Recognizing this gap, Luft [18] calls upon medical schools to utilize physician advocates as mentors, teachers, and role models for medical trainees to develop advocacy skills, especially within a formal curriculum.
Here we describe and evaluate an innovative advocacy training program, called the Advocacy Mentorship Initiative (AMI). AMI utilizes a cascading mentorship model that enables preclinical medical students to become peer mentors for youth raised in at-risk environments, while also receiving mentorship from postgraduate medical trainees (resident physicians) (see Figure 1). We aim to determine whether there are speci c advantages to the utilization of a cascading mentorship model to facilitate the attainment of advocacy skills in undergraduate medical education.

Methods
Program Description: The Advocacy Mentorship Initiative (AMI) The Advocacy Mentorship Initiative (AMI) is a supplemental advocacy curriculum at the UofT that has provided medical students with focused advocacy training since 2014. AMI pairs pre-clerkship medical students (herein referred to as mentors) to youth raised in at-risk environments, (herein referred to as mentees). The mentees are identi ed as having signi cant emotional, social, and/or behavioral concerns, which are implicitly connected to their social determinants of health. The curriculum runs for approximately one year, starting in March of the rst year of medical school and extends to completion of the second year.
AMI follows a cascading mentorship model in which mentors are provided with teaching and supervision by a UofT Psychiatry Resident and an Adolescent Medicine Fellow, (herein referred to as residents or resident mentors) in groups of three or four medical students to two residents (herein referred to as mentoring groups). The primary author, a dually quali ed child and adolescent and forensic psychiatrist, provided over-arching support and supervision to all participants (see Figure 1). Due to unforeseen circumstances, the 2019-2020 cohort did not include Adolescent Medicine mentors and thus proceeded with Psychiatry Residents only.
All mentors and residents met together monthly for large group supervision and teaching related to advocacy. Flipped-classroom teaching methods were utilized. The large group sessions were overseen by the primary author. Mentors and residents provided updates regarding pertinent matters related to the matches through facilitated group discussion. Con dentiality of the mentees was preserved; other mentors, residents and staff were not made aware of the full names of mentees. Each resident also provided at least one 45-minute didactic teaching session during these monthly meetings. The sessions were facilitated by e-modules [19] and evaluations were completed by the entire group following each of the sessions. Social justice, humanism, and critical consciousness informed the topics discussed, as well as principles related to child development and advocacy.
In addition, each mentoring group met on a biweekly basis and residents provided their assigned mentors with supervision and mentorship to foster the mentor's peer relationship with their youth and to discuss various social determinants of health.
Mentors completed mandatory CBSL training or Enriching Educational Experiences requirements, thus ful lling curricular requirements related to their medical doctoral degrees. Residents completed evaluations and provided summative feedback at the completion of the program. Mentors were required to attend at least 75% of the teaching sessions to obtain a letter of completion with comments provided by their residents. This letter contributed to their undergraduate medical education portfolio. Residents were also provided with letters of completion outlining their involvement in the program. Evaluations and comments from mentors assigned to each resident were included in these letters, which were also added to their postgraduate training portfolios.
Mentors were engaged in AMI as registered volunteers of The Peer Project, a community-based organization (CBO), which matches vulnerable youth aged 6-15 to young adult mentors between the ages of 16-29. The Peer Project is a non-denominational and non-pro t organization supported by the United Way, a national non-governmental organization. Youth who were deemed to be raised in at-risk environments were matched to mentors in AMI. Staff members from the Peer Project were included in monthly meetings to inform discussions from a CBO perspective.

Participants
Ethics approval was obtained from the UofT Research Ethics Board. Inclusion criteria consisted of active enrollment in the UofT medical school and AMI program, and informed consent to participate (see Table  1). Medical students were enrolled in March of their rst year of undergraduate medical education in 2017, 2018, 2019, and 2020. Table 1: Population, intervention, comparison, and outcome information pertaining to this study.

POPULATION
Second year medical students at UofT enrolled in AMI who provided informed consent to participate in the AMI program.
INTERVENTION Weekly 1.5 hour peer mentorship, monthly discussions with psychiatry and adolescent medicine residents, and monthly large group teaching sessions.

COMPARISON
Pre-exposure knowledge of youth mental and physical health topics and con dence levels in skills vs. post-exposure knowledge and skill levels.

OUTCOME
Unique experiential learning opportunity to gain advocacy skills.

Recruitment of medical students (mentors)
The primary author recruited students through a lunch presentation in January of the rst year for each of

Data collection
Data was collected through pre-and post-exposure questionnaires (see Additional File 1). Questionnaires were sent to participants via e-mail prior to their match with their mentees and following completion of the AMI program. Questionnaires included both Likert scales and short-answer questions. Each participant was assigned a unique and anonymous identi er by an administrative assistant; the investigators remained blinded to these codes. The questionnaire asked participants to rate their level of knowledge and skills regarding various topics and their level of con dence regarding skills on a ve-point Likert Scale (1=Poor, 2=Fair, 3=Good, 4=Very good, 5=Excellent). The pre-exposure questionnaire also included open-ended questions asking students to list their learning goals. In the post-exposure questionnaire, they were asked open-ended questions regarding skill development, bene t of involvement in AMI and resident mentors, and the impact of AMI on their career path.

Data analysis
Pre-and post-exposure questionnaire data was analyzed through quantitative and qualitative analysis. Questionnaire items using Likert scales were scored 1 to 5 (1=Poor, 5= Excellent). The distribution of differences between the pre-and post-exposure data was neither normal nor symmetrical; thus, sign tests were used to compare the median differences. Pre-and post-exposure open-ended questionnaire responses were analyzed using content analysis, employing the approach described by Taylor-Powell and Renner [20]. Upon reading through the responses numerous times, the data were categorized as themes emerged to develop a coding scheme. Categories were constructed until no new themes or subcategories were identi ed. The responses were again reviewed with the completed coding scheme. Utilizing the identi ed codes, the data was sorted into themes relevant to the focus of the study and any connections between themes were identi ed.
A triangulation protocol was adapted from Tonkin-Crine and colleagues [21] to integrate qualitative and quantitative data regarding communication skills. Pairwise comparisons were made between the quantitative and qualitative data sets to identify convergence. Pairwise comparisons were considered in "Dissonance" if the participant described that they had gained communication skills but did not demonstrate an increase in con dence in their communication skills by quantitative measures. Pairwise comparisons were considered in "Agreement" if the participant described that they had gained communication skills, as well as demonstrated an increase in con dence in their communication skills by quantitative measures. Pairwise comparisons were considered in "Silence" if the participant demonstrated increased con dence in communication skills but did not describe communication skills as skills gained during qualitative measures.
Pre-exposure self-identi ed goals The pre-exposure questionnaire asked participants to list up to four objectives they wish to achieve prior to starting the AMI program. Twelve themes emerged, as summarized in Table 2. The most common goal identi ed was to gain a better understanding of psychiatric and developmental disorders in youth. Other goals included: improve communication skills with youth and families, learn how to advocate, build relationships with youth, increased comfort working with youth, have a positive impact on mentee, become aware of community resources, contribute to the community, career development and exploration, as well as to gain a better understanding of social determinants of health, child development, and physical illness in youth. Contribute to the community 2 "Contribute and give back to the community" Self-reported con dence levels Con dence levels pertaining to 8 skills associated with the AMI program were compared and summarized in Table 3. A sign test demonstrated signi cant increases (p<0.05) in respondents' con dence regarding working with vulnerable populations and advocating for medical and non-medical needs. There was no signi cant effect on con dence regarding communicating with youth, their family members, or with staff who provide care to youth. There was no signi cant effect on con dence involving working with children who have mental illness or a chronic medical illness.

Self-reported skills gained
Thirty-one of 38 respondents (82%) answered "Yes, I have gained new skills." Identi ed themes included skills in communication, relationship building with youth, advocacy, nding community resources, cultural competency, and understanding social determinants of health. Data are summarized in Table 4, including direct quotations taken from answers pertaining to each theme. "We learn about [social determinants of health], but to go into the environment and hear rst-hand the experiences of my youth really helped me to understand these determinants." Self-reported bene ts of AMI Respondents identi ed bene ts related to their engagement with AMI. From the textual answers provided, we identi ed 7 themes. The main bene t of the AMI program was identi ed as building a relationship with a mentee. Other themes that emerged included building a relationship with the resident mentors, positive impact on mentee, learning about child development concepts and child psychiatry, learning about real-life challenges rst-hand, improving communication skills, and learning to advocate for youth. Table 5 provides representative quotes taken from responses related to the identi ed themes. Twenty-three of 38 (61%) respondents indicated that participation in the AMI program in uenced their future career. When asked to elaborate, textual responses provided codes generating ve themes. The most common theme was that AMI strengthened mentors' interest in working with youth. Other themes included awareness of social determinants of health, greater interest in psychiatry, greater interest in advocacy, and a better understanding of the advocacy role. Table 6 provides direct quotations taken from responses related to each of the themes. Awareness of social determinants of health 6 "I will now take into account social determinants of health when I recommend treatment plans for patients and their families." More interest in psychiatry

"Furthered my interest in psychiatry"
More interest in advocacy 3 "It just re-a rmed that I want to participate a lot in advocacy in whatever speciality I end up in." Better understanding of advocacy role 3 "Better understanding of how community programs and advocacy on the individual and organizational level play a role in child health and development" Bene ts of resident mentors Thirty-ve of 38 (92%) participants listed bene ts from engaging with their resident mentors. From these responses, four themes were identi ed. The most common themes identi ed were that residents provided general advice and support as well as provided advice on handling di cult situations. Other themes included career mentorship and promoted discussion regarding mentorship experiences. These themes and sample responses are provided in Table 7. "They always took the time to check in with us and provide feedback on our situations with our mentees using their knowledge from paediatrics and psychiatry. This was helpful connecting my experience to the curriculum. They also took the time to understand our matches and talk about our emotions." Advice on handling di cult situations 18 "The advice they gave and the ability to talk about any problems we had and brainstorm ways to approach di cult or challenging situations with our mentees" Career mentorship 7 "They went beyond what was required and helped support us both in the program and with our life circumstances and career aspirations." Promote discussion regarding mentorship experiences 4 "The supervising residents were able to mediate and promote highly constructive discussion amongst my classmates surrounding recurring themes."

Interpretive Analysis of Communication Skills
Of 38 ndings from 24 participants, there were 7 instances of agreement, 13 instances of dissonance, and 18 instances of silence.

Instances of Agreement
Five participants demonstrated an increase in con dence in their communication skills, as well as described that they gained communication skills. For one participant, they described gaining communication skills in all three categories (with youth, families, and staff who are in a supervisory or care provider role of the youth) and this was in agreement with increased con dence in these domains, resulting in a total of 7 instances of agreement.

Instances of Dissonance
There were 13 instances (by a total of 10 participants) in which gaining communication skills with youth and/or their families were described but were not associated with an increase in con dence in that domain. For two participants, they described gaining communication skills in general, but this was not associated with an increase in con dence in any communication domain (with youth, families, and staff who are in a supervisory or care provider role of the youth).

Instances of Silence
There were 18 instances (by a total of 11 participants) in which there was an increase in con dence in communication skills with youth, families, or with staff who are in a supervisory or care provider role of the youth, but without describing this increase in communication skills during qualitative measures.

Discussion
AMI aims to provide medical students with advocacy training through a unique cascading mentorship program. Medical students engaged with youth raised in at-risk environments and had the opportunity to learn about the broader implications of social determinants of health outside of a clinical setting. Medical students were in turn mentored by resident physicians who provided support in their relations with at-risk youth, as well as career mentorship. The knowledge and skills gained, as well as the relationships built with mentees, generated improved understanding and ability to advocate for others.
Previous research demonstrated that medical students who mentored youth from an underserved community developed con dence in skills such as establishing relationships, self-re ection, compassion, teaching, and communication, along with a better understanding of health inequity [16]. The focus of AMI is to prepare medical students to ful l their role as Health Advocate. As health advocates, physicians are expected to work with and advocate for vulnerable populations, including youth. AMI aided mentors in gaining con dence in skills such as working with vulnerable populations and advocating for the needs of youth. Medical students also described gaining skills such as relationship building with youth, nding community resources, cultural competency, and understanding of social determinants of health -all skills necessary to advocating for and with youth [4,22]. Several medical students described gaining "advocacy skills" explicitly.
CBSL allows medical students to directly observe health inequities [11,12] and learn about social determinants of health [10]. AMI is an example of CBSL, and as such, mentors suitably identi ed such bene ts from involvement in the program. Mentors discussed building a relationship with their mentee, learning about real-life challenges rst-hand (including those related to living in poverty and other social determinants of health), and learning to advocate for youth. Additionally, mentors discussed having a once mentors re ected on their experience with AMI, many mentors described gaining communication skills generally, with youth or with youth's family and other supports.
Physician advocates have been recognized as a valuable, yet under-utilized resource for developing advocacy skills in medical students [18]. The availability of physicians as mentors in AMI is especially important considering that mentors who feel unsupported leads to termination of mentorship relationships [25]. Suitably, medical student mentors identi ed the relationships they built with resident mentors as a bene t of participating in AMI. Furthermore, the medical students outlined various bene ts of having resident mentors. From the bene ts described, it is evident that the bi-weekly meetings with resident mentors were an opportunity for the medical students to discuss their mentorship experiences among peers, as well as receive advice and support from resident mentors, especially when faced with di cult situations. These meetings were also an opportunity for career mentorship, as many programs providing mentorship to medical students tend to initiate [17].
Moreover, students learned about social determinants of health, child development, attachment theory, chronic illness in youth, and autism and other developmental disabilities. AMI provided directed teaching with mentors surrounding these ve domains, as it is foundational knowledge held by various professionals that are best suited to not only understand and work with youth, but also to advocate for youth [22].
Furthermore, the majority of medical student mentors indicated that their experiences would inform their future practice. AMI promoted interest in working with youth, in the eld of psychiatry, or in advocacy. Students also recognized that AMI offered an experience that taught them to be more aware of social determinants of health and gain a better understanding of the advocacy role.
The most commonly reported goals of participating in AMI was to gain a better understanding of psychiatric and developmental disorders in youth and child development more generally; improve communication skills with youth and families; build relationships with youth; gain a better understanding of social determinants of health; and learn how to advocate. Another commonly reported goal was for career development and exploration. It is evident that these goals were aligned with the purpose of AMI and were realized throughout the program.

Limitations
Some mentors did not complete the questionnaires, which may bias the impacts of this program. Not all participants completed both the pre-and post-exposure questionnaires, thus limiting the ability to draw comparisons across the entire group.

Future Directions
To our knowledge, AMI represents the only cascading mentorship program that positions medical students as both mentors to at-risk youth and mentees to resident physicians. As such, future directions for research include expanding our understanding of cascading mentorship in medical education. This will include examining how resident physicians are impacted by the program, including development of skills related to mentoring, teaching, and professionalism.

Conclusions
AMI offers a novel approach to developing advocacy in medical students by integrating experiential community-based learning with physician engagement through a cascading mentorship model. Medical students gained con dence in their advocacy skills as well as gained a better understanding of social determinants of health, children's health and development, and youth advocacy. AMI also led to furthered interests in careers related to working with youth, psychiatry, and advocacy. It is through both mentoring at-risk youth and being mentored by residents that medical students were provided the opportunity to develop as our future physician advocates. Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing Interests
The authors declare that they have no competing interests.
Funding None Authors' Contributions MP is the founder of the Advocacy Mentorship Initiative. MP participated in study design, wrote the rst draft of the manuscript, and contributed to the revising of the manuscript. DA and YX analyzed and interpreted the data, and contributed to writing the manuscript. SS and AS provided further direction and feedback regarding the manuscript and study. All authors read and approved the nal manuscript.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. AdditionalFile1.pdf