Construction of the Digital Health Equity-Focused Implementation Research Conceptual Model - Bridging the Divide Between Equity-focused Digital Health and Implementation Research

Digital health implementations and investments continue to expand. As the reliance on digital health increases, it is imperative to implement technologies with inclusive and accessible approaches. A conceptual model can be used to guide equity-focused digital health implementations to improve suitability and uptake in diverse populations. The objective of this study is expand an implementation model with recommendations on the equitable implementation of new digital health technologies. The Digital Health Equity-Focused Implementation Research (DH-EquIR) conceptual model was developed based on a rigorous review of digital health implementation and health equity literature. The Equity-Focused Implementation Research for Health Programs (EquIR) model was used as a starting point and merged with digital equity and digital health implementation models. Existing theoretical frameworks and models were appraised as well as individual equity-sensitive implementation studies. Patient and program-related concepts related to digital equity, digital health implementation, and assessment of social/digital determinants of health were included. Sixty-two articles were analyzed to inform the adaption of the EquIR model for digital health. These articles included digital health equity models and frameworks, digital health implementation models and frameworks, research articles, guidelines, and concept analyses. Concepts were organized into EquIR conceptual groupings, including population health status, planning the program, designing the program, implementing the program, and equity-focused implementation outcomes. The adapted DH-EquIR conceptual model diagram was created as well as detailed tables displaying related equity concepts, evidence gaps in source articles, and analysis of existing equity-related models and tools. The DH-EquIR model serves to guide digital health developers and implementation specialists to promote the inclusion of health-equity planning in every phase of implementation. In addition, it can assist researchers and product developers to avoid repeating the mistakes that have led to inequities in the implementation of digital health across populations.


Introduction
Digital health interventions can provide benefits including reductions in readmissions and improved physical and emotional quality of life health outcomes [1].In the United States, record amounts of private investment dollars have flowed into digital health startups; in 2020 digital health investments topped $9.2 billion domestically and $14.2 billion globally [2].Meanwhile, there have been large investments in health technology through federal and state governments, including $80 million in American Rescue Plan funding to strengthen U.S. public health information technology [3].In addition, the Centers for Medicare and Medicaid Services expanded reimbursement beyond remote patient monitoring, including monitoring non-physiological outcomes such as adherence to physical therapy [4].In light of this insurance coverage and the evidence that digital health can improve clinical outcomes, reduces costs, and improves patient engagement, it is imperative that research and operational implementors ensure that the interventions are accessible and increase equity rather than widening inequities in outcomes [5].
"Disparity" and "inequity" are distinct concepts-disparity describes a difference, while inequity describes unfairness and injustice [6].In the U.S. the use of the term disparity ranges from suggesting a value-free difference to a value-laden unfair judgment.This paper will focus on inequities, as inequities are differences that are unjust and should be rectified.
Striking health inequities continue to plague the United States healthcare system.Inequities persist across socioeconomic (SES) levels [7].Previous work has shown how implicit bias and systemic racism has impacted clinical outcomes for minoritized patients [8].Implicit bias refers to the unconscious attitudes or stereotypes that affect an individual's understanding, actions, and decisions in an unintentional manner.This impacts patients through the unconscious biases of healthcare providers and the influence of those biases on assessment and treatment decisions [8].Systemic racism is associated with poorer physical health, and race is a social determinant of health (SDOH) [9].SDOH encompass domains that influence health inequities including employment, housing instability, language and literacy, and environmental conditions [10].Limited access to healthcare services and reduced access to healthier food and lifestyle choices also result in populations experiencing poorer health [10].
Digital health implementations can increase health inequities.Termed the "digital divide," lack of access to technology have been reported in racial and ethnic minoritized populations, individuals with disabilities, limited English proficiency, living in rural areas, and those with low income [11].Intervention-generated inequity occurs when an intervention disproportionately benefits advantaged people due to inequalities in access, uptake, adherence, and effectiveness [12].For example, patients may lack internet access, preventing them from using a patient portal.Furthermore, some patients may have limited internet access through cellular data plans.In remote and rural areas, the available internet access options may be of limited bandwidth.Limited access to internet resources can result in intervention-generated inequity, where patients in underserved areas with limited connectivity may encounter challenges scheduling virtual appointments.Digital health tools that require virtual appointments are then benefitting only persons with sufficient bandwidth.On the other hand, when digital technologies are carefully planned and implemented using an inclusive approach to development, inequities may be diminished [13].
Persistent inequities exist in the development of digital health and other technologies.Technology development has typically included predominantly white user research participants, a systematic bias that could lead to potential harms [14][15][16].The lack of a diverse sample during development phases results in flawed products, or products that are difficult or unappealing for minoritized populations to use.Seniors, people with disabilities, and people with limited English proficiency may struggle to use applications without proper inclusive and accessible designs [12,17].There is a paucity of research of socioculturally tailored digital health interventions that incorporate design based on input from minoritized populations [13].Example results of engagement in design include actors who are relatable to the user population in photographs and videos [18].Communication patterns and health literacy levels are important considerations, as persons with low health literacy have higher rates of chronic disease than those with adequate literacy [19].Prior user-centered design work crafted motivational text messages in English and Spanish for persons with low literacy levels to increase physical activity [20].A scoping review of digital health studies found that persons with disabilities were rarely included, leading to non-accessible applications [21].
A limited, but increasing, number of studies investigate technology implementations in disadvantaged populations.Effective interventions must encompass acceptability (through user acceptance testing), practicality, and be designed to address the needs of at-risk populations [22,23].The emergence of equity-focused digital health research models have helped improve the delivery of more inclusive healthcare.The adoption-focused Integrative Model of eHealth Use was updated to include health literacy and computer literacy, and the eHealth Equity Framework incorporates technology access, material circumstances, and societal values [24,25].Health equity factors such as poor engagement with digital health, health policies, and digital determinants of health (e.g.digital literacy, lack of internet access) are examined in the Digital Health Equity Framework, which focuses on access and adoption [26].An equityfocused behavioral science model called the ConNECT Framework integrates context, fosters inclusion, and prioritizes training of healthcare providers and implementors to improve health equity [27].The above models include incomplete equity concepts in relation to implementation, focusing on a particular phase or aspect of an implementation.This paper therefore develops a broader implementation model that increases focus on causes of inequities in digital health implementations from planning to outcome tracking.
Few equity-focused implementation models exist to date though there are many general implementation frameworks available to researchers [28,29].The Equity-Focused Implementation Research for Health Programs (EquIR) model was developed in 2019 and enables researchers and executives to incorporate equity considerations during the entire implementation process from the initial planning phase through to evaluation [30].EquIR's from-planning-to-outcome scope, as well as its focus on providing solutions to equity challenges sets it apart from the Consolidated Framework for Implementation Research (CFIR) [31].The CFIR model is a determinants framework; meaning it was designed to guide the systematic evaluation of factors influencing implementations of programs in various settings in the domains of intervention characteristics, outer setting, inner setting, characteristics of individuals, and process.The CFIR model was updated to include human-equality centeredness, but does not include steps informing how to alleviate said inequalities [30,31].Besides EquIR and CFIR, other equity-focused implementation frameworks include The Transcreation Framework and the Health Equity Implementation Framework [22,28].
EquIR was designed by implementation researchers, health equity experts, and results from a systematic review.It is cyclical in nature; beginning with equity-focused population health status and ending with a new post-implementation population health status.It includes five progressive phases: population assessment, planning, designing, implementing, and equityfocused implementation outcomes [30].Each phase includes direction for incorporating equity issues and approaches to diminish them.As it is a relatively new model, the applications to date are limited.However, the EquIR framework was used to investigate challenges that impact gerontology researcher-community partnerships and to generate potential solutions to obstacles [32].
There is a need for a digital health implementation model that similarly supports equityfocus throughout all phases of digital health planning and implementation.The purpose of this paper is to modify EquIR to include constructs from digital health implementation frameworks, thereby creating recommendations on the equitable implementation of new digital health technologies.

Methods
The digital health and equity-focused implementation model (DH-EquIR) builds on the original EquIR model by integrating digital health implementation into the conceptualization of equitable implementations.DH-EquIR was designed by identifying patient and programrelated factors contributing to the successful implementation of digital health technologies.We conducted a narrative review following the Scale for the Assessment of Narrative Review Articles [33] to comprehensively describe and interpret previously published literature on digital health implementation models incorporating concepts of health equity [34].A literature synthesis table was created with each EquIR construct as a header.As each digital health implementation study was reviewed, digital health concepts that related to constructs from the original EquIR model were inserted into the literature synthesis table.These patient-and programrelated implementation factors were then fused into the EquIR model to create an adaptation capable of guiding equity-focused digital health implementations.The lead author conducted the primary synthesis of results.The other authors, who all have extensive knowledge of the material, then gave feedback on the contents.The "D&I Models in Health Research and Practice" website (https://disseminationimplementation.org) was used to explore theories, models, and frameworks that involved both health equity and implementation research.

Inclusion and exclusion criteria
The search included English language articles published from January 2017 through May 2023; international studies were included.Primary research, systematic reviews, and scoping reviews covering health equity, implementation frameworks, and digital health implementations were included.Individual primary research studies were required to involve patients as the focus.Exclusion criteria omitted clinician-focused articles, conference abstracts, editorial and opinion articles, magazine articles, and protocol proposals.

Conceptual model
Equity-focused recommendations and observations were extracted from the available articles and organized by the original EquIR model's constructs, which may be viewed in its original publication [30].Digital health planning-related factors were those related to the identification of disadvantaged groups, the quantification of health inequities, and the identification of equity-sensitive recommendations.Similarly, relevant digital health literature was incorporated into the process of designing an intervention and the identification of key actors, barriers, and facilitators to equity-focused recommendations.The implementing EquIR phase was augmented with concepts from digital health implementations including communication plans, the definition of resources and incentives, strategies to overcome barriers, and monitoring and evaluation strategies for digital health.Equity-focused implementation outcomes included digital health monitoring metrics and strategies.While keeping the original EquIR model's visual structure, the DH-EquIR model revised the constructs to be digital health equity implementation-focused based on the data extracted from the source articles.

Adapted conceptual model
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram is shown in Fig 1 .Sixty-two articles were reviewed including existing digital health equity conceptual models and frameworks, digital health implementation models and frameworks, research articles, guidelines, and concept analyses.Digital health models included CFIR, the Technology Acceptance Model (TAM), unified theory of acceptance and use of technology (UTAUT), and the Integrated Technology Implementation Model (ITIM) [36][37][38].Health equity models included the Digital Health Equity Framework (DHEF), the Framework for Digital Health Equity, and the Equity Checklist for Health Technology Assessment (ECHTA) [26,39,40].Research articles include systematic reviews of barriers and facilitators to digital health uptake, interventions to increase technology usage by vulnerable populations, and articles on inclusive design.Table 1 includes details around implementation models and equity tools that were used to construct the DH-EquIR model.Table 2 depicts concepts taken from each source as they relate to the original EquIR model, as well as implications for future research identified in the articles.Table 3 includes exemplars of studies investigating equity issues within digital health implementation studies.The 17 exemplars were selected based on their inclusion of diverse patients and equity-focused recommendations.Additional specific examples of digital health implementation studies may be referenced in a scoping review of digital health interventions in ethnic-minoritized populations [41], a systematic review of interventions to increase patient portal use [42], and a systematic review of health and wellness technology use by historically underserved health consumers [43].
The DH-EquIR model (

Population health status and determinants of health
Population health status of both the general population and the disadvantaged population should be taken into account.Population health status guides the researcher in broadly considering the inequities in a population before an intervention is designed and implemented, and evaluating how those inequities are improved as a result of a program [30].Relatedly, SDOH such as occupation, gender, poverty, race/ethnicity, and socioeconomic condition are typically included in an assessment of population health status [30].Digital determinants of health (DDOH) refer to access to technology, digital literacy, and community infrastructure (e.g.broadband internet) [26,39].The DH-EquIR model emphasizes the combined importance of SDOH and DDOH in assessing population health.

Planning the program
Target population.Identification of relevant research questions and the target population of interest should be the first step of planning.
Identify disadvantaged groups.Within the target population, identify which disadvantaged populations will potentially be impacted by the digital health implementation.Disadvantaged groups for digital health frequently include financially disadvantaged, older adults, rural populations, racial and/or ethnic minoritized populations, persons with low digital literacy, persons without access to devices and the internet, low levels of education, and persons with disabilities [41,[44][45][46].Quantify current health inequities.The population health status starting phase is the general population; while in this step work is done to more specifically measure health inequities in the target population.Methods of quantifying health inequities from the literature include digital literacy assessments, health literacy and limited English proficiency assessments, estimates of proportion of adults with at least a high school degree, and interviews with engaged community members [40,[47][48][49].Implementations require consideration of subgroups that will use the technology, as well as an assessment of potential biases for or against specific population groups (e.g.identifying under-represented groups not included as engaged partners that should be) [40].The appraisal may include collecting demographics and a SDOH screener.Inequities may be identified by using a framework such as PROGRESS Plus (place of residence, race/ethnicity/culture/language, occupation, gender/sex, religion, education, socioeconomic status, and social capital) [50].Additional DDOH can be assessed by surveying specific populations about smartphone ownership and internet access, or by leveraging community-level data to examine structural barriers such as broadband internet availability.
Collecting this data helps determine potential inequities in an implementation so that strategies may be developed.Note that many other assessment and measurement tools are available, this article mentions those tools found during our literature review.
Equity-sensitive recommendations.Program recommendations for equity-sensitive digital health implementations include focus on software ease of use, training and assistance programs, realistic scenario development, human-centered design principles, and tailoring technologies to target regions [42,44,45,47,[51][52][53][54].In some cases, it may be appropriate to collaborate with social benefit programs to reach patients [55].Cost and payor coverage must also be considered during planning [56].Implementers should also consider community and government resources to support digital health projects such as the United States Health Resources and Services Administration's Telehealth Resource Centers.Digital health design activities should prioritize the inclusion of disadvantaged populations that may benefit from a solution, as opposed to the mid-to-high socioeconomic users who are frequently recruited to provide feedback on initial designs.Models such as ADAPT-ITT tailor existing applications to meet the needs of target populations [57,58].The equity checklist for health technology assessment includes additional considerations when developing a new solution [40].

Designing the program
Identify key actors.The identification of key actors for digital health implementation include engaged personnel who will support the implementation, regulatory approval contacts for some forms of digital health (e.g.telemedicine visits), evaluating and contacting payors for coverage information, considering contacts for government subsidies for those who need devices or access, and securing buy-in from healthcare executives [38,40,51,59].Other roles that facilitate a successful implementation include effective champions and facilitators to guide the project, and responsive design and engineering teams to develop the solution [38].It is recommended to include members from under-represented groups as engaged partners in development.Note that key actors may potentially contribute to either barriers or facilitators, depending on their view of digital health.
Identify barriers and facilitators to equity-focused recommendations.Potential barriers include fear of using digital health, lack of validation of accuracy of digital health, preference for in-person care, lack of care team resources, increased time and effort, and previous experiences of culturally insensitive behaviors and communication barriers [40,41,51,[60][61][62].Digital health may eliminate transportation requirements and increase time savings due to its remote nature and convenience [63].Patients reported they would be more likely to use digital health if it was recommended by their provider; a facilitator of digital health credibility [44].Including community health workers in digital health programs has shown to enable "high tech, high touch" implementation approaches [64], while community advisory boards and patient advisory committees reduce intimidation and facilitate community input [57].In the US, India, and other countries, community health workers are a supportive role filled by persons who are not typically health professionals; they serve as a frontline advocates and educators, bridging the gap between healthcare services and underserved populations to promote and improve overall community health [65,66].Digital health's ability to improve communication between patient and clinician, technology support from family or friends, and potential for saving time are other major facilitators [61].Another facilitator is the creation of digital health tools that respect the cultural attributes of patients; such as making the tool available in languages other than English [45].

Implementing the program
Design an equity-focused communication strategy.Digital health programs were found to be most effectively communicated with social reference and advertisements [67].Social reference describes the referral of a product or service by someone an individual trusts.Advertisements may help to establish a digital health tool's credibility.Placing adverts in patient newsletters, patient-centered social media groups, community groups, and disability groups are other communication strategies [46,55].
Define resources and incentives.Resources and incentives for digital health include possible health benefits for patients, reassurance from a connected care team, increased ability to self-manage a condition, and reduced travel [44,60,62].The incentives will differ based on the target population (e.g diagnoses, medical specialties) and intervention.One example of an incentive is better control of hemoglobin a1c and long-term comorbidities in patients with diabetes.
Define strategies to overcome the identified barriers.Implementation teams should design strategies to overcome the barriers identified in the previous phase.Strategies include the development of visually intuitive solutions, provision of technology support, subsidization of smartphone plans and technology renting programs, inclusion of translation features, and development of accessible software interfaces [42,[45][46][47]51,59,[68][69][70].Other strategies include using video instead of written materials to increase comprehension among persons with low literacy skills, simplifying text, and creating tools that respect the cultural attributes of users [43,45,47].The transcreation process may be used to culturally tailor interventions [71].New methodologies for precision cultural tailoring have been introduced, while others promote a better goal as usability and accessibility across multiple cultural groups [72].To improve accessibility of applications, compliance with technical accessibility guidelines (such as appropriate font size and color contrast) is encouraged [57,73].Issues around device and internet access may be addressed with device loaner programs and the use of cellular devices which do not require individual data plans [52].
Define monitoring and evaluation strategies.Engaging targeted end-users in developing outcomes relevant to them is an important best practice to utilize when defining evaluation and monitoring strategies [74].The impact of digital health is indirect compared to interventions such as pharmaceuticals [53].Clinical pathway mapping defines the clinical contexts in which new technology will be used; this may serve as a foundation for health economic evaluation [53].Other outcomes include user experience and patient satisfaction surveys, the estimation of cost effectiveness, and completion of an unmet needs analysis to support redesign of the intervention [36,53].Digital health should be routinely evaluated for data errors as a result of human error [51].Finally, web and mobile analytics software can be used to track engagement [42].

Equity-focused implementation outcomes
Sustainability and interoperability.Digital health sustainability is enabled with sustainable financial models and interoperable platforms [38,51,59,62].Interoperable platforms improve sustainability by removing workflow inefficiencies that drive abandonment of digital health initiatives and costly integrations that prevent many practices from sustaining digital health tools.Interoperability also impacts other implementation outcomes.
Coverage and payor reimbursement.Coverage involves the degree of reach, access, or effective coverage on the disadvantaged population [30].For example, the proportion of eligible target population members who participated in the digital health program.Separately, reimbursement coverage could be impacted by Medicare, Medicaid, and private insurance coverage for particular technologies in the US.Even if not directly reimbursed by payors, implementation cost can be offset by cost savings and efficiency from digital health [62].
Implementation costs, cost savings and efficiency.Implementation costs should be tracked, while outcomes such as improved efficiency, avoided emergency room visits, and decreased hospital readmissions can indicate cost savings.It is also important to monitor for associated costs such as additional workforce needed to maintain the program [62].
Fidelity, personalization, and engagement.Fidelity is improved with patient engagement features such as easy reminders, personalized components, and improved communication with healthcare providers [42,51,56,62].Key performance indicators may include utilization rates of features such as personalized education and clinician messaging.
Feasibility, device management, and technical support.Feasibility is impacted by complexity of the application, the support of friends and family, and financial factors [42,43,56,61].The complexity of the application can be measured with usability reviews.
Appropriateness, satisfaction surveys, and usability studies.Appropriateness can be measured with digital literacy assessments, satisfaction surveys, and ease of use surveys [62,67,75].
Adoption, utilization, and workflow integration.Adoption is improved with personalized education, ease of integration into daily routines, and two-way messaging with care teams [61,62,5,76].Acceptability includes perceived usefulness, behavioral intention to use, compatibility, perceived privacy, and perceived cost [36,37,67,77].Integration of self-monitoring data into the care team's existing workflows enables personalized feedback to be given to the patients [64].The act of monitoring and responding to clinical data requires a dedicated team member at the clinic, or provisions to incorporate the extra required work into a provider's limited time.
Acceptability, performance, and social influence.Acceptability or effectiveness is another important outcome.Web and mobile analytics software as well as qualitative inquiry can help to measure performance, social influence, and usefulness of a digital health intervention [42].

Discussion
To the authors' knowledge, the DH-EquIR framework is the first to integrate digital health implementation concepts into the EquIR model.This is significant because conceptual models and frameworks to date are very limited in the digital health equity domain.The DH-EquIR framework serves to pragmatically guide the planning, design, implementation, and monitoring of implementation studies that deploy digital health tools into practice within underserved populations.The inclusion of digital determinants of health is an example of an adjustment to the original model.While it is still critical to evaluate SDOH throughout the implementation process, DH-EquIR encourages researchers to further consider elements such as digital literacy and access to internet.Additional elements for DDOH may include accessible content needed by persons with disabilities, or assessments of smartphone addiction in which strategies to maintain reduced screen time may be required.The inclusion of human-centered design principles in the development of equity-sensitive recommendations promotes co-design with members of underserved populations.In the identify key actors stage, the DH-EquIR model highlights the importance of including underserved population members as organizational leaders to balance power dynamics [40].Additionally, user experience surveys and patient engagement concepts broaden the scope beyond EquIR's original bounds.The DH-EquIR framework serves to drive implementation planning and execution questions.Its practical nature allows the researcher to reference the model for ideating around potential barriers as well as solutions.The model is intentionally broad in nature, which presents challenges around specificity.Digital health applications vary in scope and complexity; the model may need to be adjusted for specific use cases.
Given the proliferation of digital health applications and the increasing reliance on internet access in daily life, it is critical to address health equity.Due to SDOH and DDOH, there are many barriers to universal digital health coverage.Without intentional consideration of health equity, digital health implementations skew towards usage in younger, wealthier populations that can afford to pay a premium on the latest devices and applications.Women, minority ethnic groups, and older people are frequently underrepresented in research, resulting in nongeneralizable findings across populations [78].Underrepresentation of these groups in digital health research studies also means that they were not included in the design of the application or device.The introduction of generative artificial intelligence into digital health increases the need for equity vigilance, as AI can perpetuate bias at scale [79].Digital health developers and implementation specialists can use DH-EquIR to avoid repeating the same mistakes that have previously led to inequities.Planning for digital health equity improves the overall quality, usability, and allows the needs of diverse communities to be met.
Further, digital health tool development is a participatory process and involves continuous evaluation cycles intertwined with implementations that change the organization of healthcare [54].The DH-EquIR model incorporates these tenets, as it visually depicts the cyclical nature of the development and implementation lifecycles.The DH-EquIR model can be used to identify key dimensions and factors to consider in their research study, as well as serve to remind researchers to include health-equity planning in every phase of implementation.

Limitations
The DH-EquIR model has several limitations.First, the model requires empirical testing before being used in operational contexts.The added concepts require real world testing in deliberate studies designed to test the model's feasibility.Only patient-focused (technologies designed for patients as opposed to clinicians) studies were included in this analysis.In addition, a single author (LG) completed the literature search, article screening, and determination of article inclusion.Due to resource restraints a higher-level review such as systematic review could not be undertaken.Other gaps include the need for further methods of measuring inequities in digital health and enforcement of standards to enable interoperable programs.

Conclusion
Historically, digital health has not been inclusive of the needs of diverse populations.The digital divide continues to disproportionally benefit individuals not impacted by barriers such as low income, poor internet access, and lack of necessary devices.Health equity research has produced a number of helpful models and frameworks to address inequities, yet a comprehensive model incorporating digital health implementation needs from planning to measuring outcomes was lacking.The DH-EquIR model synthesized concepts from health equity models and combined them with concepts from equity-focused digital health implementation research.The model drives equitable technology implementations beginning with assessing the target population, and progressing through planning the program, designing the program, implementing the program, and measuring equity-focused implementation outcomes.

Fig 2 )
was created by incorporating the digital health concepts from the literature review into the corresponding location on the original EquIR model figure.In

Table 1 . Analysis of models, reviews, and equity tools: Relevant aspects included in the development of the DH-EquIR model. Lead Author, Year Type / Model / Tool Aspects of equity Key Concepts for DH-EquIR Development
line with the original EquIR model, the DH-EquIR model contains five phases: assessing population health status and SDOH/DDOH as a starting point, planning the program, designing the program, implementing the program, and equity-focused implementation outcomes.The model culminates with a reassessment of the population health status post implementation.The text below describes each phase of DH-EquIR to give additional context.

Table 3 . Equity issues in exemplar digital health implementation studies. Lead Author, Year Equity consideration Disadvantaged population
Alwashmi 2020 Software ease of use, training/assistance, credibility of digital health Low digital literacy, older age, financial barriers, lack of internet