Digital Health Tools in Genomics: Advancing Diversity, Equity, and Inclusion Open Access

Genomic testing is increasingly moving into mainstream clinical and public health practice. Sequencing is routinely ordered in oncology, pediatrics, and neurology to diagnose and treat patients, while genomic newborn screening is being developed as a public health screening test [1]. This has placed a growing demand for genetics services across medicine. Yet, the limited number of genetic specialists available has revealed an urgent need for alternative models of care to ensure the delivery of high-quality genetics services [1, 2].

Patient-facing digital genomic tools are increasingly being used to facilitate the delivery of genetics services including patient intake, phenotyping, education, counseling, and result reporting [3]. A recent systematic review found that digital tools improved the workflow of genetics providers and service efficiencies [4]. Existing data on patient experiences also demonstrate high acceptance of pretest digital tools across various contexts, with most patients endorsing and expressing high levels of satisfaction with these tools [4, 5].

However, a closer examination of digital genomic tools reveals a concerning lack of consideration for diversity, equity, and inclusion (DEI) principles in their design, evaluation, and implementation within health systems [6]. These principles are a focal point because health disparities continue to persist in contemporary society. Digital tools may reify health disparities by widening the digital divide, which refers to the gap in access to and utilization of digital technologies between different socioeconomic groups. Examples of this include differential access to the internet, affordability of technologies and services, and differences in digital literacy, all of which have been identified among socially disadvantaged communities [3]. This warrants a critical reflection on the current state of the science behind digital genomic tools. In this perspective, we review the current consideration of DEI in digital genomic tools and draw on opportunities and priorities to attend to DEI in the development, evaluation, and use of digital tools for genomic medicine (Fig. 1; Table 1).

Digital genomic tools are often developed and tested across non-diverse populations, with limited consideration of various end-users and communities, or their unique cultural, educational, and accessibility needs. These patient-facing tools are primarily developed and evaluated by English-speaking individuals of highly educated, White backgrounds [4, 6‒8]. This lack of diversity among participants in the upfront design of these tools limits the inclusion of varied needs and perspectives of diverse populations and may serve to widen digital health inequities further [9]. Digital solutions must account for differences in decision-making, communication, values, and norms. Further, existing differential health systems capacities, workflows, and workforce availability can impede the implementation of digital genomic tools, which also need to be considered in development to avoid reinforcing structural disparities.

One strategy for developing inclusive digital genomic tools is human-centered design (HCD). HCD and user-centered design are complementary development approaches that involve working with individuals in the design and refinement of innovations to develop technologies that best serve their practices, needs, and preferences [10]. However, HCD strategies move beyond the end-user (e.g., patient or practitioner) to also consider their broader community and the systems within which they live [11], while addressing utilization barriers experienced by marginalized communities. Detailed approaches to HCD involve participatory design, ethnography, lead-user approach, contextual design, co-design, and empathic design [10]. These approaches may differ depending on the stage of digital genomic tool development and the context of deployment. Promoting a co-design HCD approach facilitates building a sense of ownership and trust in the technology by empowering user-group communities and placing them at the heart of the development processes. This allows them to become partners of innovation, actively shaping and leading the final product. Utilizing this approach during the development process leverages their lived experiences, practices, and needs while also fostering a partnership in shaping the digital solution.

Once developed, it is critical to evaluate the effectiveness of digital genomic tools. However, there is a concerning lack of equity-centered frameworks, measures, and tools for evaluating digital health technologies across medical disciplines, including genomic medicine [6]. For instance, the systematic review by Lee et al. [4] examining 70 patient-facing digital tools developed for use in genomic medicine found that “ethnicity/ancestry,” income, and education level were the only dimensions of DEI commonly considered in the evaluation of digital genomic tools, and other crucial aspects of diversity (e.g., age, sex, gender identity/expression, sexual orientation, physical/intellectual ability) were frequently neglected [4]. Among 87 studies reviewed, only 34% reported all three outcomes related to “ethnicity/ancestry,” income level, and education level. There was also a notable absence of diversity in terms of education and income levels; most studies reported fewer than two education levels or three income levels within their participant cohort. Moreover, the literature also reveals a limited consideration for general literacy, health literacy, and digital literacy when evaluating digital genomic tools. Surprisingly, 82% of the studies reviewed provided no information on participants’ literacy levels [4]. Capturing intersectional dimensions of diversity when developing and testing these tools can enhance their effectiveness for the patient communities which they are meant to serve. This may be done through developing and evaluating digital tools across main user groups and tailoring the tool to serve the specific needs of each group in their settings and environment [12].

Equity-centered evaluation metrics should consider the multifaceted impact of digital genomic health innovations on the overall health and well-being of the particular patient communities and the settings in which they operate. Such measures should comprehensively consider the social determinants of health (SDoH) broadly, beyond education, and socioeconomic status, to account for differences in perceptions toward genetic testing and diagnoses, accessibility of genetic health technologies, and supportive community resources. Interrelated facets include general literacy, genomic literacy, digital literacy, and health literacy, in addition to considering gender, religiosity, region/rurality, attitudes, access and trust in the healthcare system, and factors that influence decision-making for genetic testing [13]. These intersectional dimensions of diversity are often overlooked in the design and evaluation of digital genomic tools. There remains a pressing need to develop equity-centered frameworks and evaluation measures that address multidimensional DEI variables and their intersectionality in digital solutions, especially for genomics [6]. Developers should prioritize evaluating relevant outcome measures that are specific to the target users, communities, and settings of their digital genomic tool while accounting for the bias built into existing measures and adapting these measures and scales to ensure they are assessing impact appropriately.

Digital genomic tools can be co-designed and evaluated with an equity-centered lens but ultimately it is how digital tools are implemented that can create or mitigate disparities in genomic medicine. Evidence about the implementation of digital genomic tools is scarce, and a minority of studies have integrated equity – at least in part – in their design [4]. Only two of the 87 studies reviewed by Lee et al. [4] integrated participants’ sociodemographic characteristics in their outcome analyses, yet still disproportionately enrolled participants who are fluent in English or have a White background [4, 8]. Implementation science frameworks, strategies, and learning health system approaches can advance DEI in genomic medicine.

Mitigating disparities in genetic health services and specifically ensuring that digital genomic tools do not reinforce them requires a deliberate focus on equity during implementation. The Consolidated Framework for Implementation Research (CFIR) [14] and RE-AIM [15] are widely used to study implementation and have recently been expanded to integrate health-equity considerations. The latest CFIR iteration added sub-constructs such as human equality-centeredness of implementing settings, as well as considerations of the priorities, preferences, and needs of end-users [16]. The RE-AIM extension to enhance sustainability [17] is used to assess implementation outcomes such as reach, adoption, or effectiveness in diverse subpopulations. Such frameworks are needed to guide researchers in selecting and reporting relevant disparity indicators such as representativeness, effectiveness, or heterogeneity of health outcomes.

At the preimplementation needs assessment stage, the health equity implementation framework [18] should be used in planning for equitable implementation. It was developed to understand why disparities exist and can guide researchers toward the determinants known to have a major impact on health equity. Considering some of these determinants during the needs assessment phase allows researchers and implementers to select and tailor appropriate implementation strategies to address them. For example, considering complementary strategies to integrating digital technologies with low-tech options may include integrating peer moderators and community peer advisors, having interpreters, language translations, public health kiosks, and community support can help increase reach and effectiveness in specific populations (e.g., underserved patients with limited health literacy, reading capacity and computer experience) [9, 19]. Planning and tracking implementation strategies used to improve accessibility and monitoring their impact on outcomes will help build evidence of their effectiveness that can be transferred to other settings or tools.

Finally, collecting data on both service and implementation outcomes at the population level – and not just the users – is not only more reflective of the real-world impacts of digital genomic tools but is required for continuous monitoring and adaptation of these tools to population needs. A crucial step to equitable implementation of digital genomic tools is knowing which patients are not reached or for whom they are not effective. Gathering these data can be used as a springboard for learning health systems and informing healthcare improvement efforts. Considering the fast pace of genomics advancement, and the high level of coordination necessary to deliver quality genomic services to diverse populations in multiple specialties, digital innovations will require ongoing monitoring and adaptations. Embedding in these tools the capacity to collect and analyze variables relevant to equity will help tackle current and future equity issues, and the need to augment digital approaches with complementary low-tech strategies.

Digital genomic tools are only one of many interventions that exist in the genetic health service delivery ecosystem. It is crucial to understand the broader context within which digital health and digital genomic tools operate.

Health disparity research has long shown that the root causes of disparities in healthcare are not biological in nature but rather are due to social and structural factors that result in the social disadvantage of individuals or communities, negatively impacting their health outcomes and experience with healthcare. SDoH, the conditions in which people are born, live, work, and age, influence these disparities [20, 21]. Some examples of SDoH include education, food access and security, safe housing, exposure to pollution, literacy skills, racism, discrimination, and structural violence. Existing social policies, structures, and practices often cause these factors to be unevenly distributed among individuals or communities, creating health gradients from social gradients [21]. These disparities exist within genetics services as well, where racialized, low-income, uneducated patients from deprived neighborhoods often carry a greater burden of disease while simultaneously having the lowest and/or delayed referral and care rates [22, 23].

The interaction between digital determinants and SDoH is often complex and multifactorial, and the digital divide presents significant challenges in leveraging digital tools effectively [24]. In fact, 4.4% of Americans lack access to broadband internet, and approximately 17% of Americans living in rural communities and 21% of Americans in Tribal lands do not have internet access [25]. These values may be even higher in less resourced areas. Efforts to bridge the digital divide should involve improving digital infrastructure, expanding broadband internet access, providing affordable or subsidized devices and data plans, and promoting digital literacy programs [6]. In this digital age, where the fulfillment of many basic needs can be met online, access to technology and digital competency is vital to minimize the digital divide and address broader SDoH without exacerbating existing disparities [9].

Digital genomic tools are just one of several interventions within the health service delivery ecosystem. We must be cautious not to over-conflate the power and reach of digital genomic tools as the sole or best solution for equitable health service delivery. For communities with limited digital access, digital literacy, or are tech-averse, community tailored and multilingual printed materials, community outreach programs, and public advertisements remain valuable strategies for engagement. While digital genomic tools may enhance access to genetics services, they are not sufficient on their own, and low-tech solutions should be developed to augment digital genomic tools. Digital tools are not a one-size-fits-all solution, and continuing to develop digital tools without continuously and concurrently examining which populations are left behind will only widen the digital divide and limit the adoption, usefulness, and effectiveness of future health innovations, and their potential to improve population health [6].

Digital genomic tools have become critical to ease workforce shortages and increase the efficiency, access to, and quality of genetics services. However, existing health inequities indicate that the development of new health innovations does not automatically equate to health gain. There is a clear lack of standardized DEI metrics that are used to design, and evaluate these digital tools, limiting their effectiveness, adoption, and reach to already underserved communities. Challenges remain in the pragmatic implementation of health equity-centered goals, processes, and measures in digital genomics. Without the harmonization of approaches and equitable development, evaluation, and implementation of health innovations, in both digital and non-digital forms, we will continue to implement systems that inherently discriminate against our patients. This requires dedicated funding to resource co-designed, community-empowered, and driven initiatives, diverse representation in research teams, and the tools they develop. Governance, trust, and partnerships between researchers and communities are needed to build, evaluate, and implement tools with an intersectional, DEI lens, to ensure that digital health innovations serve the needs of all, to advance precision genomic medicine.

Y.B. is the co-founder and CEO of Genetics Adviser, Inc. MC is the co-founder and Creative Director of Genetics Adviser, Inc. The remaining authors have no conflicts of interest to declare.

C.B. and S.G. hold a CGS-M scholarship and S.M. holds a CGS-D from the Canadian Institutes of Health Research (CIHR). G.D. holds a CIHR Fellowship Award and a training award for specialty medicine residents from the Fonds de Recherche du Québec – Santé. Y.B. holds the Canada Research Chair in Genomics Health Services and Policy.

D.A., S.M., and V.A. contributed equally to this work. D.A., S.M., V.A., S.G., C.B., M.C., G.D., and Y.B. contributed to the writing and have read and approved the manuscript.

Daniel Assamad, Safa Majeed, and Vernie Aguda contributed equally to this work.