Recent scholarship has generated much optimism about the potential of -omics and precision medicine to transform healthcare, particularly by discovering mechanisms of disease pathology, identifying sources of interindividual variability, and developing personalized therapeutics. To transform clinical care, this innovative research must be translated into practice, a step easy to ignore in discovery phases. Engaging stakeholders throughout the research process to understand patient and provider needs—and how these needs vary by community, setting, and social identity—will not only aid successful implementation of precision medicine technology but also help ensure that the benefits of these innovations are more broadly democratized throughout society.
While the most readily adoptable tool of precision medicine—pharmacogenomics—has been shown to improve the safety and efficacy of many medications, its implementation has been troublingly uneven. Pharmacogenetic testing has been launched with positive effects on patient outcomes and healthcare costs. Yet these efforts have overwhelmingly focused on patients of European ancestry and have been limited to major academic medical centers serving metropolitan communities and large regional health systems. By contrast, rural, community-based health systems, including and especially those serving populations of non-European ancestry, have been slow to implement pharmacogenetic testing. Complexity illuminated through further precision medicine studies will undoubtedly improve the application of pharmacogenomics, but repeated study of the same populations will likely continue to be of limited value for non-Europeans and those without access to major medical centers.
In these ways and no doubt others, advances in precision medicine threaten to exacerbate existing health care disparities in the United States and the wider world. Though nearly half the U.S. population lives outside of major metropolitan areas (counties with greater than 1 million people), rural dwellers have historically been the last to gain access to emerging healthcare technologies, have less convenient access to medical care, and suffer from a higher burden of chronic disease. People of non-European ancestry—who make up the bulk of the world’s population—often face similar challenges. Additionally, populations of non-European ancestry may have different genetic frequencies and environmental exposures that affect treatment and priorities of care. Emerging -omics knowledge and the resulting tools of precision medicine may not translate to communities underrepresented in biomedical research. Including these populations should be a top concern of researchers.
One worthwhile approach is to directly engage rural and other underrepresented communities to increase their participation in and access to medical innovations offered by precision medicine. Our community-academic partnership between the Confederated Salish and Kootenai Tribes (CSKT), residing on the Flathead Reservation in western Montana, and the University of Montana offers one example of how community-based participatory research can engage rural and underserved communities in precision medicine (1). As with our effort with American Indian and Alaska Native (AIAN) people, scientific engagement with other communities might improve local participation, enrich research questions, and inform implementation strategies. Most broadly, similar efforts can help make precision medicine more widely beneficial for all people regardless of geographic location or social identity.
Engaging communities as partners throughout the research process may improve the willingness of participants and clinicians to engage in biomedical research. Community-engaged research emphasizes trust, shared decision-making, education, and mutual benefit and commitment throughout the research process. In Montana, we built partnerships with healthcare providers and Tribal leaders who helped formulate research questions that aligned with their healthcare priorities. Notably, these partnerships began three years prior to securing grant funding. Tribal Health leadership helped us establish a community advisory board comprised of members of the community to ensure that Tribal interests were prioritized and to provide guidance on community interactions (2). The structures of Tribal governance provided a clearer path to identifying the appropriate approval processes for our research and aided our ability to sustain ongoing engagement with the broader CSKT community. Examples of this engagement include organizing precision medicine workshops; presenting the research partnership across the Flathead Reservation to various stakeholders (e.g. CSKT Tribal Council, Tribal Health clinics, Salish Kootenai College, Tribally-owned businesses, and governmental agencies); and working with our advisory board to develop community-appropriate strategies for disseminating research results in local newspapers and on social media. Members of the advisory board have stressed how community engagement has shaped their views on research. As one member noted, “I know how research is often conducted. I know it certainly wouldn’t prompt me to participate in any research. But authentic community-involved research, I definitely would [participate].” Another member drew an even stronger conclusion. “I think that [new research projects] would have to do the same things that [these pharmacogenetics researchers] did.” (2).
Understanding and focusing on the health priorities of a community may ensure that research adds value. Our community-academic pharmacogenetics partnership began because of community research priorities in oncology and a desire to improve outcomes in cancer treatment. Given these community-defined goals, we were the first to evaluate tamoxifen pharmacogenetics in AIAN women and found that women who are cytochrome P450 2D6 poor metabolizers are at an increased risk of having subtherapeutic concentrations of endoxifen, the active metabolite of tamoxifen, and may be at risk for cancer recurrence (3). Our approach allowed us to identify genetic variation affecting cancer treatment in AIAN communities and to contribute to general scientific knowledge by informing tamoxifen pharmacogenetics globally. These early pharmacogenomics research studies have evolved to include gene-environment interactions and qualitative assessments of the acceptability of pharmacogenetic testing as trust and interest in research have grown in the community. Community partnerships and a commitment to local engagement can add breadth to precision medicine knowledge by characterizing biomedical relationships in more diverse populations.
Tailoring research to particular healthcare infrastructures and communities may encourage caregivers and care receivers to adopt and persistently use the precision medicine tools developed in those settings. Like treatments, implementation strategies are not one size fits all. Distinctive challenges regularly facing rural and underserved settings include geographic remoteness, shortage of personnel with -omics expertise, fragmented or antiquated electronic health records, lack of local testing facilities leading to delays in the return of test results, and a general unease of rural dwellers with the “big city” (4). For example, clinical decision support tools that add little value to a specialist may prove invaluable for a generalist, while treatment strategies perfected in a large medical center may be incompatible with workflows elsewhere. Including diverse communities in developing implementation strategies will make these strategies more appropriate and useful and likely improve their adoption. For instance, we currently are working to conduct specific needs assessments in collaboration with the CSKT Tribal Health to identify barriers and facilitators and to develop an implementation framework for pharmacogenetic testing in their clinics. By engaging communities, inclusive biomedical research has the potential to reach a wider range of people who could benefit from its results.
While crucial for improving precision medicine knowledge and interventions, targeted engagement of rural and underserved communities can be challenging and requires a different, often more time-consuming approach than conventional biomedical research conducted in large medical institutions (Table 1). The most important aspect of community engagement is for researchers to avoid the “helicopter” research that has eroded underserved communities’ trust in research. Principal investigators and project leads must commit to regularly traveling to the communities with whom they work, to being present at community events, and to maintaining relationships and partnerships with communities after research findings are published and grants end. Another consideration is how researchers identify who in the community will provide guidance for the research (e.g. who are the right people or groups of people in a community from whom to seek approval and what scale of community engagement is required). To add further complexity, these questions will differ from community to community. To build a strong level of trust between the CSKT community and academic researchers, we laid the groundwork for our partnership six years prior to our first pharmacogenomics publication (5). Community engagement may progress more slowly and require more flexibility than conventional research methods, yet by fostering trust, it ensures that precision medicine will have a wider impact and reach a broader group of people.
Table 1.
Considerations for engaging rural and underserved communities in precision medicine
| Challenges | Recommendations |
|---|---|
| Disinterest or opposition to research questions or study methods proposed by researchers. | Start small to build trust. It is important for principal investigators and project leads to be engaged and represent the research. Formulate research questions with guidance from community advisory board and community leaders to address important questions that result in visible value for the community. Adapt study methods to accommodate community needs and concerns. |
| Mistrust and lack of receptivity to participating in studies with researchers not known to the community. | Engage community as co-researchers and active members of the team. The most successful research activities occur when members of the community propose important questions, act as representatives for the research, share information, and discuss research with potential participants. Include community members in manuscripts and solicit feedback prior to submitting study results to peer review journals or external presentations. |
| Distance between communities and research centers; geographical remoteness of communities. | Travel regularly as a research team to the community for recruitment events and meetings with key stakeholders and community advisors throughout the year. Be flexible—long travel distances, sometimes in inclement weather, force delayed or canceled events. Consider resources for community members who may lack access to reliable transportation and for whom the cost of travel to the recruitment site is a barrier. |
| Lack of laboratory infrastructure and difficulties in maintaining integrity of biological samples collected away from biomedical research labs. | Equip “mobile labs” (e.g. vehicles or planes) with portable freezers, centrifuges, and phlebotomy supplies capable of processing biological samples in remote locations. Train or recruit community members who can lead data collection and processing. |
| Recruitment settings are variable and subject to change. | Recruit in locations that are available, even if atypical of conventional biomedical research, including conferences rooms, gymnasiums, schools, dispatch centers, and cultural events (e.g. powwow celebrations). |
| Hesitation of communities to relinquish control of data. | Develop and maintain plans for data ownership, stewardship, and security with the community. Train community members to analyze and interpret data. Community oversees future secondary use. |
| Shortage of expert personnel to deliver precision medicine. | Provide training and education for local providers through distance-learning strategies. Utilize telehealth to deliver precision medicine consultations on testing services and translating -omics information into clinical actions. |
| Insufficient infrastructure to integrate precision medicine technologies into clinical care. | Tailor clinical decision support tools to integrate within the available infrastructure and workflows. Centralize testing and interpretation at a single center. Leverage resources at large medical centers to lessen the burden on local care centers. |
Community engagement, of course, will not guarantee that precision medicine is developed and delivered equitably, and it comes with design challenges of which researchers, practitioners, and participants should be aware. Most immediately, communities and medical systems are highly diverse. Defining communities—where to draw geographic or socio-cultural boundaries—is highly subjective, often politicized, and risks either overinclusion or inappropriate exclusions. Determining membership—who belongs and who doesn’t in a given community—has similar problems. In particular, the use of social identity, especially race and ethnicity, in classifying communities and their members (and in –omics research more generally) may lead researchers to rely on participants’ self-identification or to themselves make assumptions about participants’ identities. Both approaches raise concerns, and especially if researchers attempt to scientifically categorize findings rooted in constructions of identity that are inherently social and political (i.e. not biological). Furthermore, how the results and models generated in specific communities are shared and translated to the global patient population is challenging, as it requires combining results across research studies conducted in different settings and with different patient populations, and somehow applying the collective knowledge to individual patients.
Nevertheless, conducting research in collaboration with communities and healthcare systems will ensure that precision medicine tools capture a fuller scope of human variability and needs and that the resulting interventions are beneficial and accessible for greater numbers of people. We have built a strong level of trust between a rural, underserved community and academic researchers. Our community-focused approach can serve as a model for developing community partnerships and promoting inclusive precision medicine and -omics research.
Acknowledgments
Funding: The work is supported by NIH grants R01HG009500, P01GM116691, and U54GM115371.
Footnotes
Conflict of Interest: The authors declared no competing interests for this work.
REFERENCES
- 1.Woodahl EL, Lesko LJ, Hopkins S, Robinson RF, Thummel KE, Burke W Pharmacogenetic research in partnership with American Indian and Alaska Native communities. Pharmacogenomics. 15, 1235–1241 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Morales CT, Muzquiz LI, Howlett K, Azure B, Bodnar B, Finley V, et al. Partnership with the Confederated Salish and Kootenai Tribes: Establishing an Advisory Committee for Pharmacogenetic Research. Prog Community Health Partnersh. 10, 173–183 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Khan BA, Robinson R, Fohner AE, Muzquiz LI, Schilling BD, Beans JA, et al. Cytochrome P450 Genetic Variation Associated with Tamoxifen Biotransformation in American Indian and Alaska Native People. Clin Transl Sci. 11, 312–321 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dorfman EH, Brown Trinidad S, Morales CT, Howlett K, Burke W, Woodahl EL Pharmacogenomics in diverse practice settings: implementation beyond major metropolitan areas. Pharmacogenomics. 16, 227–237 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fohner A, Muzquiz LI, Austin MA, Gaedigk A, Gordon A, Thornton T, et al. Pharmacogenetics in American Indian populations: analysis of CYP2D6, CYP3A4, CYP3A5, and CYP2C9 in the Confederated Salish and Kootenai Tribes. Pharmacogenet Genomics. 23, 403–414 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
