Introduction—Mentoring Context

Integrating novel technologies and approaches in HIV research will require a capable and diverse scientific workforce. Such a research workforce would follow “Team Science” (i.e., disciplinary integration on methodologic and theoretic aspects) as the predominant research philosophy, model, and approach. This approach is consistent with one of the research priorities of the Training, Infrastructure, and Capacity Building area of emphasis for the FY 2016 Trans-NIH Plan for HIV-Related Research [1]. A team approach is timely because comprehensive multidisciplinary HIV treatment and management is needed in the changing landscape of HIV science. As we have entered a new era in HIV prevention, where priorities have expanded from biomedical discovery to include implementation, effectiveness, and effective combination prevention at the population level, multidisciplinary research will be critical. These emerging ‘high-impact’ prevention strategies, such as treatment-based prevention, rely on the integration of care, prevention, and innovative transdisciplinary team approaches. This approach will also be fruitful for training investigators for HIV health disparities research through the creation of new hybrid basic/clinical science disciplines. For example, a training strategy that fosters research teams that advance neuroscience theory and cultural/contextual research, such as cultural neuroscience, as it relates to population disparities could be one such hybrid [2]. This type of approach would embrace an integrated conceptual framework to address both upstream determinants of health (e.g., social-environmental factors) and downstream determinants (e.g., neurobiological features). In fact, the transdisciplinary approach has been proposed as a key element in the understanding of health disparities to inform evidence-based practice, social action, and effective policy change to improve health in disparate communities [3].

Translational research, a subset of transdisciplinary research, is also an essential component of HIV research, bridging gaps across the basic, clinical, and services continuum. In addition to the establishment of best practices for creating partnerships and collaborations (with researchers, public health agencies, communities, institutions, and other stakeholders), there is a critical need to translate research findings into policy and practice. This may be addressed through HIV implementation science to optimize the incorporation and use of established procedures, enhancing the adoption of best practices in community settings. The recent launch of the National Center for Advancing Translational Sciences was specifically targeted to advance this purpose—to offer translational education and training for the support of a pipeline of investigators who can accelerate or rapidly transfer basic research findings into the clinic to improve human health. Expediting this translation will better enable the responsiveness to scientific advances, regardless of the specific HIV scientific domains.

Unfortunately, relatively little is known about the unique characteristics of transdisciplinary training or the elements that effect specific training outcomes [4]. It may be assumed that scientists tend to be more responsive to transdisciplinary and translational research during their early pipeline career stages and are more fixed in their scientific approaches and possibly less open to incorporating new perspectives and models at later career stages. However, no confirmatory data is currently available to verify this assumption. Transdisciplinary research often requires multiple mentors—as often the scientists who serve as mentors were trained in a single disciplinary framework. There may be advantages to a single discipline for relatively discrete research questions, but the majority of HIV research topics, especially those focused on the reduction of HIV disparities, lend themselves much more to the transdisciplinary approach in focusing on mutable mechanisms and targets for treatment and interventions. This may include studies to: (1) advance combination behavioral-biomedical intervention approaches; (2) examine the HIV care continuum; (3) incorporate context and social determinants into the development and testing of interventions; and (4) scale-up of evidence based interventions across diverse subpopulations. Central to conducting research on these and other HIV research topics are the collaborative research partnerships and training opportunities that capitalize on resources, technology, data registries, cost-effectiveness analyses, biologic outcomes, appropriate use of controls and measures, and ethnographic/other qualitative methods.

Regardless of the approach selected, the obvious question is how many diverse investigators should be mentored and trained to achieve an “adequate” number that select an independent research career path in the HIV research field. A number of possibilities have been considered to identify this target number: (1) using population representation from the population in which the study is conducted; (2) reflecting the level of HIV risk; (3) matching the HIV incidence in groups currently disproportionately impacted; or (4) reaching a critical mass of investigators. While the absolute numbers of scholars should not be the sole criterion alone, it is clear that the significant gap between the numbers of majority and underrepresented minority investigators needs to be closed. This can be justified on a number of levels including: (1) enhancing the breadth of perspective that diversity brings; (2) broadening the creative and scientific thought process; and (3) broadening science to include the voices of those impacted and studied but underrepresented at the investigator level. In so doing, scientific inquiry and its targeted populations become accessible to all and not a privileged few [5].

It could be asked if diversity and diverse representation matter. Certainly, important contributions have been made by committed non-minority investigators—to deny this would be short-sighted and inaccurate. Rather than focus on race-based restrictions, the focus should switch to outcomes based on experiences of working to overcome disadvantage. Appropriate training with positive outcomes leading to a strong workforce is important. Given the nature of research training, measurable outcomes often take 10 years or longer. Therefore, programs must exist long enough to allow for an evaluation which includes follow up over a sustained time period. It may be more accurate to examine the quality and the quantity of disparities research as outcomes (measured against certain projected goals) rather than actual investigator number.

A practical approach to determining the size of the research pipeline addresses the extent of training in a fiscal context, also referred to as balancing the pipeline with the funding breakpoint [6]. The funding breakpoint is directly related to the pipeline, i.e., the higher the funding breakpoint, the more the pipeline can be fed and the lower the funding breakpoint, the less we can feed the pipeline. However, it is challenging to find the best balance between building a pipeline to train new investigators and funding of sufficient numbers of highly meritorious grant applications among those who have been trained. In order to maintain a sufficiently sized pipeline pool of investigators, in the context of the fiscal climate, several funding and training factors must be taken into account. These include the success rate (i.e., the number of funded applications divided by the number of unique applications in a given fiscal year), the requirements for different levels of training (from predoctoral to early career), and the length of training needed for satisfactory evaluation.

As described in the preceding papers of this supplement, academic health centers and their departments play a larger role in encouraging formal research mentorship in public health (e.g., HIV/AIDS) where mentor–mentee interactions are deliberate, structured, and goal-oriented. The NIH has implemented a number of initiatives targeting investigators from diverse backgrounds at the individual (e.g., K mentored career awards) and institutional levels (e.g., T32 training awards, R25 research education awards) (Table 1). While research mentoring is an aspect of all these initiatives, few have focused exclusively on formal mentoring programs. Subsequent to the Ginther et al. report of racial disparities in successfully-funded NIH grant applications, the NIH has demonstrated renewed emphasis to prime the research pipeline and enhance the diversity of the pool of highly trained biomedical researchers [7]. These efforts have been detailed in a report with the recommendations of the NIH Advisory Committee to the Director [8]. This report highlighted mentoring as a key component of biomedical researcher development and, more specifically: (1) identified mentorship as the second most important issue, based upon respondent feedback to an NIH Request for Information, (2) identified transition points in the biomedical research career as places where mentoring was particularly important, and (3) recommended funding initiatives for mentoring-related programs (e.g., National Mentoring Research Network) [9–11]. To achieve the maximum benefits from these mentoring efforts, the key principles of mentoring and models of effective mentoring need to be identified.

Table 1 Some HIV mentoring programs building research workforce of diverse investigators (2015–present)
Full size table

There is wide variation in the types of mentoring models employed, as well as mentoring formats [12, 13]. Some models rely mainly on practical considerations such as available resources, while others may focus more on the unique relationship between the mentor and mentee. In contemporary HIV research, the multiple co-mentor model has been preferred as the benefits include: (1) immersion in a wrap-around mentoring climate; and (2) institutional structure conducive to cooperation and coordination of an apprenticeship approach [14]. Transdisciplinary mentoring, the most frequently used mentoring model, utilizes a framework that cuts across organizational silos, fosters an integrated approach across multiple disciplines, and requires strong collaborations between researchers and community organizations, service providers, and systems [4]. The advantage of this model is that contextually appropriate and relevant research is conducted, with findings that can be utilized and translated at many levels—from the individual to the systemic—hopefully leading to sustained improvements in population health. Although infrequently used, the peer mentoring model is one of the more promising mentoring models. Peer mentoring focuses on collaboration and team building, conferring many benefits to underrepresented minority investigators including, but not limited to, social support, socialization within academia, and visible role models [15]. Peer mentoring can indirectly address the problem of the scarcity of mentors by developing a new group of potential mentors—an important addition to the biomedical workforce.

Mentoring Principles

These principles are not fixed, but serve as a framework to guide mentees, mentors, mentoring programs, and institutions [16–22]. These principles can help guide the approach for the establishment and development of collaborations that foster an environment that promotes and rewards quality, effective research mentoring—ultimately moving the mentee toward research independence. Moreover, the mentoring process is influenced by a number of overlapping and interacting domains/variables (i.e., mentor, mentoring program, infrastructure, institutional leadership) and is in the service of generating a body of research knowledge to ultimately impact the disparities affecting the vulnerable population targeted (Fig. 1). When addressing multilevel and multifaceted issues in public health, a multimodal training approach is required, that cuts across organizational silos and fosters integration across multiple disciplines [23].

Fig. 1
figure 1

Main interacting mentorship components contributing to diverse research workforce

Full size image
  1. 1.

    Mentoring programs should be evidence-based and theoretically-guided. The knowledge base of social science research must be incorporated into an evidence-based approach for effective mentoring with conceptual models of factors that influence the development of a research career (e.g., mentee background characteristics, mentee social/ecological and cognitive factors and competencies) [24].

  2. 2.

    Mentoring programs can leverage collaborations and partnerships, capitalizing on shared interests. More immediate proximal outcomes of mentoring programs (e.g., publications) are enhanced by taking advantage of past investments in research capacity building, as well as existing infrastructure (e.g., research centers, training grants) that can also help with more distal outcomes such as sustaining programs.

  3. 3.

    Mentoring programs can be both dynamic and iterative. The mentoring process involves an institutional team and partnership, each with different roles. These range from conceptualizing projects to implementation and subsequent publication of study results with many activities in between. Team collaboration can and should create a program that drives the development of new ideas and identifies priority areas of study while nurturing individual creativity.

  4. 4.

    Mentoring programs can advance the highest priorities in HIV/AIDS science. Areas of mentored research can be aligned with priorities of the trans-NIH Plan for HIV-Related Research and with NIMH’s Division of AIDS Research (DAR), for mental health/central nervous system (CNS)-related mentored research. Research education mentoring programs need a thematic concentration in the proposed targeted HIV/AIDS research area. Training is needed not only in HIV priority research areas, but also in novel methodologies to assess in real time new field and/or laboratory observations, as well as to better utilize novel resources (e.g., big datasets) for complex, multifactoral research inquiries. Mentoring programs must be responsive to the guidelines for the recently published NIH overarching HIV/AIDS research priorities [25].

  5. 5.

    Mentoring programs can prime the HIV/AIDS pipeline. Priming the career pipeline can be accomplished by targeting a broad range of career levels (high school to early career), a broad spectrum of academic and scholastic achievement, and specific stages of the research career ladder (e.g., transitions from one career stage to the next).

  6. 6.

    Mentoring programs can be personalized. Tailored mentoring and targeted career development activities are especially useful for individuals from diverse backgrounds. These activities may involve developing and improving science process skills such as analytical skills, hypothesis generation, research independence, and motivation to learn.

  7. 7.

    Mentoring programs can be implemented within the context of institutional infrastructure, leadership, and accountability. The benefits of institutional support are unmistakable, and in conjunction with a plan to enhance diversity an institutional climate in which mentoring can thrive can be created.

  8. 8.

    Mentoring programs can integrate the trainee into the research culture. It is important to educate the trainee about the academic/research culture, to teach networking skills and establish professional contacts, while modeling how to become part of the research culture. This allows the bidirectional flow of intellectual and cultural capital between mentee and mentoring team.

  9. 9.

    Mentoring programs can be culturally based and contextually focused. In order to be sensitive to the range of sociocultural and economic perspectives found among diverse communities, it will be necessary to employ methods and measures that are responsive to these differences. In addition to employing these methods, the regular incorporation and involvement of community members and stakeholders in the research generated by such programs will be essential.

  10. 10.

    Mentoring programs must promote cultural competency training. When mentoring is embedded in a cultural competency framework, it may address issues including cultural customs, language, humility, power, privilege, and social justice in order to facilitate inclusiveness and diversity.

Mentoring Outcomes

Evaluation is an essential component of mentoring programs, for program assessment and modification, as well as the evaluation of the outcomes and tracking of participant progress. Mentoring program evaluation can include four tiers of evaluation: (1) program evaluation by the mentees, (2) mentoring process evaluation by the mentees and mentors, (3) effectiveness evaluation based upon the tracking of mentee progress, and (4) examination of the career trajectory of the mentees as a measure of program impact. While metrics for these tiers have utility, some have suggested that it is necessary to go beyond traditional program evaluation (e.g., research productivity). This can be achieved through empirically and theoretically based frameworks that allow for positing research questions about program effectiveness [26]. A major benefit of such a theoretically guided approach is the ability to operationalize short (e.g., obtain HIV grant support), intermediate (e.g., demonstrate leadership in HIV research field), and long-term (e.g., strengthen HIV research workforce) outcomes that result from meeting the program core competencies.

There is a clear need for evaluation strategies to determine the effectiveness of targeted programs in achieving the goal of enhancing the diversity of the biomedical and behavioral research workforce [8]. Retrospective program evaluation to measure the outcomes of a specific mentoring program after completion is important, but of limited utility from a more analytical perspective. Logic models can serve as an analytical tool for program evaluation, and strengthen the evaluation process by focusing on what the program is supposed to achieve. By focusing on how program activities and processes lead to outcomes, it analyzes a mentoring program in several ways: (1) monitoring program activities and ensuring that programs remain focused; (2) defining measures of success and program impact; (3) determining which mentoring program components are effective, which yield less than the desired result, and which areas could be improved with additional resources; (4) providing conceptual models for framing potential research questions and planning for more focused mentoring programs. Logic models achieve these functions diagrammatically through presenting the changes the program intends to initiate by: (1) inputs (resources dedicated to or used by the program); (2) activities (what the program does with the inputs to achieve its objectives); and (3) outputs (the direct products of the program’s activities) associated with the benefits it aims to generate [26]. Recognizing that mentoring is an intervention, it would be appropriate to take advantage of logic models regarding training effects, much like they have been used successfully in HIV prevention [27]. In summary, utilizing innovative logic models is an effective way to monitor program activities, as well as to ensure that programs stay focused and offer a way to suggest new innovative mentoring plans for the future.

Summary

Within a transdisciplinary research context, this paper presented some principles as a framework to guide quality and effective mentoring approaches, and logic models to serve in the development and analysis of new mentoring programs. There is a need to look beyond individual mentoring programs in isolation and to pursue a systems-based approach to the study of scientific workforce dynamics. Despite a considerable body of data on outcomes, policy makers still lack a solid body of research to guide decisions on intervention training strategies. Census data clearly show that racial and ethnic minorities will become the majority population by 2060, with non-Hispanic whites representing approximately 44 % of the U.S. population; yet despite this growth, minority populations remain grossly underrepresented in the sciences [28]. It remains unclear what interventions are successful in promoting the diversity, health, and stability of the research workforce. Advances in systems science and computational methods will make it possible to build models of the scientific workforce. This will inform our understanding of workforce dynamics, support the development and management of interventions, and guide the collection and analysis of the necessary for program design. Prospectively, much needs to be understood regarding the dynamics that produce successful scientists; the economic, cultural, and social influences that impact scientific training; and how individuals make career decisions. It is important to learn what happens during the different training stages to ensure that minority trainees develop the skills, knowledge, and competencies essential to successful careers in the biomedical research workforce. However, mentoring is not enough. Research mentoring programs must be viewed as just one component of a more comprehensive career development strategy that includes infrastructure support, capacity building, collaborations, partnerships, and inclusiveness training. These activities must be coordinated to collectively build the next generation of diverse and successful HIV scientists.