Mentoring health researchers globally: Diverse experiences, programmes, challenges and responses

ABSTRACT Mentoring experiences and programmes are becoming increasingly recognised as important by those engaged in capacity strengthening in global health research. Using a primarily qualitative study design, we studied three experiences of mentorship and eight mentorship programmes for early career global health researchers based in high-income and low- and middle-income countries. For the latter, we drew upon programme materials, existing unpublished data and more formal mixed-method evaluations, supplemented by individual email questionnaire responses. Research team members wrote stories, and the team assembled and analysed them for key themes. Across the diverse experiences and programmes, key emergent themes included: great mentors inspire others in an inter-generational cascade, mentorship is transformative in personal and professional development and involves reciprocity, and finding the right balance in mentoring relationships and programmes includes responding creatively to failure. Among the challenges encountered were: struggling for more level playing fields for new health researchers globally, changing mindsets in institutions that do not have a culture of mentorship and building collaboration not competition. Mentoring networks spanning institutions and countries using multiple virtual and face-to-face methods are a potential avenue for fostering organisational cultures supporting quality mentorship in global health research.


Introduction
Mentoring has been recognised as an important component of programmes aiming to strengthen health research capacity globally (Bennett et al., 2010;Harle, 2011;Lansang & Dennis, 2004). The programmes include those with mentoring in specific areas of global health research such as injury prevention (Hyder, Meddings, & Bachani, 2009) and HIV clinical trials (Mbuagbaw & Thabane, 2013). Shah, Nodell, Montano, Behrens, and Zunt (2011) developed guidelines for transnational mentorship programmes in clinical global health research. Research mentorship is also recognised as important for public health trainees (Tweheyo et al., 2011;Zea & Belgrave, 2009) and has been incorporated into public health research training programmes (Gourevitch et al., 2012). Yet there are challenges in mounting mentorship programmes in lowand middle-income countries (LMICs) (Nundulall & Dorasamy, 2012). Adequate funding and connections internationally have responded to some of these challenges (Bennett, Paina, Ssengooba, Waswa, & M'Imunya, 2013) though structural problems persist (Nakanjako et al., 2014), in keeping with persistent inequities across and within countries.
As a group of academics located in various parts of the globe and engaged in mentoring newer health researchers in both high-income countries (HIC) and LMIC contexts, we wished to share and learn from our diverse experiences of, and approaches to, mentoring in global health research. The questions we posed were: how have mentorship initiatives developed in different contexts? What methods have been used over what periods of time? What challenges and successes were encountered? We engaged in a process of description, story-telling, curation and reflection in light of relevant literatures to provide suggestions for those engaged in health research mentoring globally.

Relevant notions of mentoring
Mentoring, and the associated terms mentor, mentee and mentorship are understood in myriad ways by different disciplines and organisations (Sambunjak & Marušić, 2009). Often, mentorship is equated in nursing and medical student education with supervision (Andrews & Wallis, 1999;Nakanjako et al., 2014;Roy & Linendoll, 2006). Mentoring in higher education has often been seen as part of faculty responsibilities. In surveys of graduates of doctoral programmes (Aanerud, Homer, Nerad, & Cerny, 2006), faculty mentoring was described as too all-encompassing to be effective. Therefore, 'most respondents were not getting the research and publication mentoring critical for their success within academic (research) careers' (Aanerud et al., 2006, p. 128). Among research funders, the NIH National Research Mentoring Network promotes support to biomedical researchers through grantsmanship training, and the Canadian Institutes of Health Research expect mentors, as part of their Strategic Training Initiatives in Health Research, to be directly involved in training. In contrast, the Wellcome Trust expects the applicant's mentor to provide guidance, rather than training. To clarify our meaning of mentoring, we drew on Bozeman and Feeney's definition (2007) (see Box 1). Box 1. Our definition of mentoring.
'Mentoring: a process for the informal transmission of knowledge, social capital, and psychosocial support perceived by the recipient as relevant to work, career, or professional development; mentoring entails informal communication, usually face-to-face and during a sustained period of time, between a person who is perceived to have greater relevant knowledge, wisdom, or experience (the mentor) and a person who is perceived to have less (the protégé).' (Bozeman & Feeney, 2007).

Methods
We drew on primarily qualitative study design traditions informed by narrative inquiry approaches (Webster & Mertova, 2007). Narrative inquiry works with text and stories as the fundamental data for description, analysis and interpretation. In late 2012, we contacted global health research mentorship programmes known to team members, with the aim of collecting and analysing a set of case studies representing a range of geographies and types of programmes (see categories below). As the project progressed, team members suggested other colleagues whose experiences would complement the spectrum of programmes identified. In total, 11 cases studies were included; 3 mentorship experiences and 8 formal mentoring programmes.

Comparative description of programmes
We drew on D'Abate, Eddy, and Tannenbaum's (2003) classification of 13 developmental interactions. As defined by D' Abate et al. (2003) these are 'interactions between two or more people with the goal of personal or professional development' (p. 363), such as coaching, mentoring, apprenticeship and action learning. We compared the eight programmes in terms of their goals, modalities, skills targeted, institutional sponsorship, participants and funding supports. Four of the eight mentorship programmes had unpublished needs assessment and monitoring data upon which we could draw (e.g. number of participants, feedback on formal sessions, annual self-evaluations by trainees). The other four programmes had conducted formal mixed-method evaluations with appropriate ethics review (Barnoya, Monzon, & Colditz, 2013;Godoy-Paiz et al., 2011;McCullough, 2012;Toranzos & Rutty, 2011) from which research team members extracted relevant information. Organisational website information was supplemented by email exchanges with programme leadership and Microsoft Excel tables facilitated comparisons.
Three programme-affiliated research team members emailed questionnaires to all former participants in their programme. The questionnaire drew upon research on mentorship programme characteristics (D'Abate et al., 2003), evaluation of mentorship programmes (Berk, Berg, Mortimer, Walton-Moss, & Yeo, 2005;Keyser et al., 2008), and themes related to successful and failed mentoring relationships (Straus, Johnson, Marquez, & Feldman, 2013). These were supplemented by competencies in global health  and essential competencies for faculty members (Bland, Taylor, Shollen, Weber-Main, & Mulcahy, 2009, Figure 4.4, pp. 53-54). Ethical approval for this new data collection was obtained from team members' respective research ethics boards (Hospital de Clinicas, University of Buenos Aires; London School of Hygiene and Tropical Medicine; University of Toronto Health Sciences). Selected data were incorporated into the programme stories.

Story development
In order to explore the personal and deeper learning experiences of key actors in each of the programmes, we adopted research story and narrative approaches increasingly being used in social science health research (Frank, 2010), health education research (Greenhalgh & Wengraf, 2008) and knowledge translation (Bell, 2010). Each team member developed a mentorship case story around a theme chosen both for its centrality to the programme experience and for the dialogue that the theme created amongst the stories such that, when read together, these stories would reveal a rich understanding about the particularities of mentorship in the context of global health research.

Analysis
Drawing on narrative inquiry and making use of social media tools, early drafts of the stories were posted on the Canadian Coalition for Global Health Research's intranet. The research team posted comments about the connections between and differences among the various stories as well as emerging themes. Taking these into account along with story-by-story group discussion on periodic teleconferences, team members prepared final story drafts. During a day-long series of teleconferences and face-to-face meetings, the research team generated a set of emerging themes. The themes were subsequently situated within the current literatures on mentorship, global health research and their intersection.

Mentorship experiences and programmes
The final set of 11 assembled stories encompassed diverse approaches to mentoring across time and place (see Box 2). Two focused on extended relationships, one of co-mentoring between HIC and LMIC colleagues (Wisdom Shared), and the other of senior mentoring of mid-career colleagues, who in turn mentor (Mentorship Cascade). One experience described initiatives at three Canadian universities, with elements of mentorship and leadership development for global health research (Seize Opportunity). The other eight stories were based in health research capacity strengthening programmes in which mentorship either plays (played) an important role, or is (was) the primary focus of the programme (see Tables 1 and 2). The programmes varied in the extent of planning and structure provided to the mentorship process from a collegial network (Thirst for Mentorship) or set of peers (Finding Success in Group Mentorship), through flexible options (Mentorship and Supervision) to more formalised roles (Evaluating a Mentoring Program). Some employed formal contracts to prompt participation and

Great mentors inspire others to become mentors
A key theme in a number of the stories was the mentoring of mentors. It is described with particular poignancy in Breaking New Ground and Finding Success in Group Mentorship, in which two young research trainees received mentoring in countries with a tradition of mentoring (the USA and the UK, respectively). Partly in response to the demand for mentorship, they later started mentoring programmes in their home countries (Argentina and Kenya), as did two mentees of a senior mentor (Mentorship Cascade). Indeed, modelling of mentorship can be seen as a form of inter-generational learning in which knowledge, values and practice pass from one generation of health researchers to another, often accompanied by explicit training in mentoring (Mentorship and Supervision, Thirst for Mentorship) (Bland et al., 2009). Several stories noted 'paying it forward' as an explicit expectation of the programme, encouraging or requiring participants to mentor the next cohort of mentees (Breaking New Ground, Building a Community of Practice, and Mentorship and Supervision). This inter-generational cascade of mentoring is essential for creating a critical mass of trained researchers, as described by an LMIC researcher in Mentorship Cascade: These meetings … are the occasions when I have floated my wild dreams, heard him [a senior mentor] cutting them to size and sharing his words of wisdom, and helped me sharpen my vision. I am using the same mentorship methods to mentor my junior colleagues.

Mentorship is transformative
All of the stories highlight benefits that accrue to menteesacquisition of new knowledge and skills, making new contacts, becoming part of a community of practice, publishing and opening up career opportunities. Mentorship is described as facilitating 'deep learning', in which seemingly disparate pieces of knowledge suddenly connected in a new and exciting way or in which knowledge of one's personal self is enhanced to motivate change (i.e. traditional gender roles). Both types of learning happen in 'safe spaces' (Fostering Safe Places)environments or personal relationships where mutual respect has been established. In these safe spaces, individuals are exposed to a diversity of ideas, opinions and experiences and are willing to open themselves up and 'admit we know nothing about certain things and even less about others' as do the two protagonists in Wisdom Shared. Online academic writing mentoring also resulted in transformative experiences for both research students and their supervisors (Loneliness of the Long Distance Scholar), consistent with the crucial role of academic writing competence for growth as a researcher (Bazerman, 1988). The longitudinal stories (Mentorship Cascade, Wisdom Shared) show how mentorship can flourish in long-term relationships, encompassing the 'whole person'.

Reciprocity
Mentoring is 'a two-way street' in which mentors themselves have learned and grown professionally from the experience of mentoring. In Breaking New Ground, Raúl Mejía points out how mentoring activities have boosted his career. In Wisdom Shared, readers are offered an intimate portrait of the reciprocity of learning and personal growth that can be found in some relationships: Our co-mentoring took place around crisis points, unexpected opportunities and unpredictable needs. When we were assisting each other to navigate complex waters we had to change roles and operate on faith, trust and respect. The more than twenty-year difference in age and the difference in academic positions became secondary. Co-mentorship focused on momentby-moment coaching on cultural nuances that had to be understood for progress to be made. This experience resonates with Lansang and Dennis' (2004) view that 'co-learning (in which the mentor and the person being mentored learn from each other) … that takes place through community-based organisations and civil society groups is an important way of strengthening … skills … ' (p. 765). Centeno (2002) noted 'both mentor and mentee must profit and learn from the process of mentoring in order to fulfil an essential purpose of the university. Their relationship represents the bond between the past and the future'. Cascades of mentorship, similar to those longstanding in research laboratories and in clinical medicine, can be built into a global health research programme (Shah et al., 2011).

Finding the right balance
Several stories describe how a relationship started out with a particular set of goals, and then stretched, bent and blossomed into something different over time. Stories like Wisdom Shared, Mentorship and Supervision, and Finding Success in Group Mentorship reveal the challenges in navigating disciplinary norms and expectations, as well as institutional (Warner, 2002), national (Isichei, 2007) and international cultures (Tams & Arthur, 2007). Other stories, like The Loneliness of the Long Distance Scholar, offer insight into the challenges of aligning individual and programme goals. Several (particularly Building a Community of Practice) showed the importance of getting right the mix of mentoring modalitiesthat is, one-to-one and one-to-group, peer-to-peer and inter-generational, 'chance' and 'planned', or face-to-face and virtual. Collectively, the stories demonstrate the importance of dialogue and listening to one another, and of flexibility and adaptability in working through problems and 'finding the right balance'.

Responding creatively to failure
Neither does every mentee who starts a programme finish, nor is each mentor as committed as desirable or all mentors-mentee pairings a good fit. The judgement of failure varies by culture and discipline, yet it is a key to learning. The stories speak to how failure can be important to innovation. As an LMIC research trainee described in Group Mentorship: When I attended the sessions of the group in my office it occurred to me that this was the opportunity I had been waiting for all along. The discussions were lively and enlightening to the students and to me as well. There was a free flow of ideas and suggestions from the students themselves that enriched their work and helped those who were stuck move on. It was then that I realized that in the current situation, where few mentors are willing to give their time and support to students, more students would benefit if mentorship was done in a group setting rather than the traditional one-on-one setup between the mentor and the mentee. In a group, one mentor is available to multiple students at the same time, thus saving on time and the mentor does not feel overwhelmed.
Another example was the emphasis on flexibility in using multiple mediums of communication, fundamental to the dialogic communication strategy of the E:SEO (Loneliness of the Long Distance Scholar) for transforming failures of both understanding and expression into new opportunities for learning. Responding based on a team's learning becomes one of the responsibilities of those leading the mentorship programme (e.g. Thirst for Mentorship) (Bland et al., 2009), as evidenced by Makerere colleagues' 'skills training [and] induction courses for doctoral students-mentor teams' (Nakanjako et al., 2014).

Level playing fields
In The Loneliness of the Long Distance Scholar, two mentors observe that 'the unequal production and accumulation of knowledge [is] an aspect of globalisation as worrying as the inequitable distribution of wealth [and disease]'. Disparate access to the validation of knowledge persists for authors of the Global South (Falabella et al., 2007). Recognition of disparities in training, career opportunities and recognition of health researchers globally was a starting point for most programmes. Several stories also speak to struggles faced by global health researchers in Canadian universities 'achieving not only permission for, but recognition of, their global work'. Mentoring involves not only developing skills or bolstering self-esteem, but jointly fostering improvements in the conditions of colleagues in LMIC institutions and facilitating policy change to improve the health of the vulnerable (Breaking New Ground) (Vasquez, Hirsch, Giang, & Parker, 2013). Such mentoring resonates with partnerships taking critical, constructivist stances (Philpott & Batty, 2009).

Changing mindsets
Hand-in-hand with empowering a next generation of global health research leaders goes the need to develop leadership in mentoring. Several stories describe their protagonists' journeys from mentee to mentor to leader. Certainly research mentorship, from the health professional student to the visiting mid-career fellow level, has been woven increasingly into HIC-LMIC collaborations (Bennett et al., 2013). Yet, as one LMIC researcher noted, 'Until recently, mentoring was not formally recognised as a capacity building tool for research in Latin American countries. The unusual cases that happen are mostly a matter of chance rather than part of an organized, planned process' (Breaking New Ground). LMIC academics have written about the challenges of hierarchy, gender and ethnicity within their universities (Isichei, 2007;Palomar Verea, 2005;Warner, 2002) and the implications of these, coupled with workload demands, for establishing research mentorship programmes (Geber, 2009). Women were generally the minority both of mentors and mentees: about 40% of mentors (both HIC and LMIC) in Building a Community of Practice; about 25% of mentors (all HIC) in Mentorship & Supervision; and 30% of mentees in Breaking New Ground, consistent with ongoing concerns about women in science. Changing mindsets (Ghaffar, IJsselmuiden, & Zicker, 2008)individual and institutionalis key to creating a 'culture of mentorship' (Bland et al., 2009), with senior mentors facilitating access of junior researchers of both genders to a network of mentors to support them (DeCastro, Sambuco, Ubel, Stewart, & Jagsi, 2013). As has been seen in other resource-constrained contexts (Oni et al., 2011), the stories celebrate collaborations which can ensue through mentorship, embracing interdisciplinary approaches, team work and creativity in global health research.

Implications
Although our mentorship examples are promising, research gaps persist on mentoring in global health research, similar to those identified by human resource colleagues in their review (Hezlett & Gibson, 2005) in which they argued for clearer conceptual frameworks and judicious application of qualitative and quantitative methods guided by theory rooted in practice. Similarly, vocational researchers noted concerns with 'over reliance on crosssectional designs and self-reported data, a failure to differentiate between different forms of mentoring (e.g. formal versus informal)' (Allen, Eby, O'Brien, & Lentz, 2008). In the midst of the current flurry of mentoring activity, some energy should be devoted to research development including: tracking trans-national mentoring relationships longitudinally, analysing how online capacity building in academic writing functions as a mentoring process, developing adequate cross-cultural measures to tap into deep learning through mentoring, exploring personal as well as mentoring programme outcomes, and assessing fuller programme and policy impacts of mentoring in gendered and inequitable global contexts, all embedded within more rigorous designs (Sambunjak, Straus, & Marušić, 2006).
Advice for mentors abounds (Lee, Dennis, & Campbell, 2007), some more evidencebased (Straus & Sackett, 2014) than others. Although we approach the making of additional suggestions with trepidation, we set out promising directions in Box 3. In relation to explicit recognition (#1), we concur with Vasquez et al. (2013) that global public health research mentorship programmes would do well to address HIC-LMIC power differentials and promote simultaneous integration of more local and more global contexts. We argue for intentionality but flexibility in mentorship programmes (#2) with greater leadership in tackling institutional barriers to mentorship (#3), both resource lacks and rigid, gendered hierarchies, similar to Vasquez et al. (2013). In terms of mentorship programmes in global health research, the creation of 'safe spaces' (#4) is essential. The use of mixed modalities (#5) shows promise, particularly for connecting globally. Finally, we endorse Young, Alvermann, Kaste, Henderson, and Many's (2004) view that mentoring relationships should be more about co-learning (#6) and interdependency which 'encourage[s] individual growth while simultaneously facilitating a sense of friendship, collegiality, connectedness and caring between the mentors and mentees'. Fostering supportive interdependency across divides, both historical and globalisationrelated, remains a key goal for mentorship in health research globally, perhaps through mentoring networks which span academic centres and countries using multiple virtual and face-to-face approaches.
Box 3. Directions for mentoring in global health research.
(1) Explicit recognition of the global differences in conditions, resources, and access is required for multi-national mentorship programmes (2) Strongly planned and structured mentorship programmes are not always appropriate. Some flexibility is preferable to respond to different contexts and emerging needs of the mentees (3) Leadership in mentorship is required for institutional change and greater prioritisation of mentorship in global health research (4) By creating 'safe spaces', individuals can be exposed to a diversity of ideas and encouraged to open themselves up for holistic development as health researchers (5) Mixing mentoring and communication modalities (e.g. one-to-one and one-to-group, peer-to-peer and intergenerational, 'chance' and 'planned') may be preferable for responding to different mentor capacities and mentee needs (6) A co-learning approach between mentors and mentees may be a useful way to promote the co-development of mentorship programmes and mentees, across hierarchies