The combination of an aging population and declining birth rate is proving to be a great challenge for many modern health care systems, resulting in rising costs and spending cuts [1]. In particular, policy makers express concerns about the rising costs of dementia care, as there are currently 50 million people with dementia, and this number is set to triple by 2050 [2]. Informal caregivers of people with dementia, such as spouses, friends, and other loved ones, provide a large part of the necessary care for people with dementia at home [3]. However, the informal caregiving process often results in chronic stress, leading to caregiver overburden, depression, and anxiety [4].

Policy makers and governing bodies have expressed enthusiasm for electronic health (eHealth) as a solution to tackle these current health care challenges [5,6]. Various eHealth interventions have shown evidence of effectiveness at improving outcomes for the caregivers of people with dementia, such as self-efficacy and dementia knowledge, as well as reduced depressive and anxious symptoms [7-13]. eHealth interventions are defined as “treatments, typically behaviorally based, that are operationalized and transformed for delivery via the internet” [14]. eHealth interventions provide specific advantages to the caregivers of people with dementia, as they can be personalized and adapted to the stage of dementia and allow caregivers to receive psychoeducation without leaving the person with dementia home alone and to seek help without facing the stigma associated with dementia. For these reasons, eHealth is also mentioned as an important part of the Dutch Deltaplan Dementie [15] and in the council of the European Union’s dementia policies [16]. Of course, there are also specific challenges associated with implementing eHealth for the caregivers of people with dementia, including the advanced age of many caregivers. While many older adults show high digital literacy, impaired motor, cognitive, and perceptive abilities can constitute relevant barriers [17-19].

Unfortunately, the implementation of evidence-based eHealth interventions into routine practice has proven challenging [20-22], and previous research has shown that very few eHealth interventions for dementia are implemented into practice [23]. Here, implementation is defined as “the process of putting to use or integrating evidence-based interventions within a setting” [24]. A lack of insight into eHealth interventions’ contextual determinants and process changes is an important factor in the slow implementation of many eHealth interventions [25]. Additionally, challenges in implementing eHealth include limited evidence of the demonstrable effects on improving health care outcomes, skeptical attitudes from health care professionals, lack of coordination and management of interventions within health care organizations, and the often peripheral position of potential end users in eHealth development [26]. Many of these issues result from problematic atheoretical implementation and insufficient implementation strategies [27], which are “methods or techniques used to enhance the adoption, implementation, and sustainability of a clinical program or practice” [28]. This lack of successful implementation is an important missed opportunity for the health care system, as advantages of eHealth interventions for health care include the potential to widen access to more remote areas, lower thresholds for participation, improve quality through increased opportunities for personalization, improve service efficiency, and reduce costs [22,29]. To facilitate the sustainable success of these promising interventions, it has been argued that the development of a business model is paramount [30]. Here, a business model is defined as “the rationale of how an organization creates, delivers, and captures value” [31]. Business modeling can be seen as part of an effective implementation strategy, primarily through its potential to both aid sustainable financing and identify value drivers to ensure the relevance of the interventions to the target users [32]. Finally, it is important to note that the specific challenges experienced in implementing eHealth interventions for the caregivers of people with dementia can differ across settings. In this study, we explored the at-home setting of Partner in Balance, which is implemented through health care organizations.

One example of an eHealth intervention to support the caregivers of people with dementia is Partner in Balance. Partner in Balance is a web-based tool to support the caregivers of people with dementia at home, which is applied in a “blended” 8-week eHealth intervention. This “blended” aspect entails that Partner in Balance is delivered through a coach. These coaches are part of participating health care organizations (for example, dementia case management organizations), who have agreed to offer the Partner in Balance intervention to their clients. Partner in Balance coaches are required to have experience (1) in healthcare and (2) with dementia. The coaches are required to take part in a 2-hour Partner in Balance training course, were the intervention is presented and the coaches take part in various coaching exercises.

Afterwards, caregivers first meet coaches face-to-face for an intake session, where relevant modules are chosen to help the caregivers adapt to their new role. At home, the caregivers complete the chosen modules, which consist of caregiver tips, video vignettes, self-reflective assignments, and web-based feedback from the coach. Finally, the coaches and caregivers meet for an in-person evaluation session. Partner in Balance is currently available in Dutch, French, German, and English. The development and testing of Partner in Balance made use of the stepwise approach of the Medical Research Council (MRC) framework for complex interventions [33]. Information on the results of the needs assessment [34], pilot study [35], randomized controlled trial [36], and process evaluation [37] has been published previously. These last two studies showed that Partner in Balance increased caregiver self-efficacy, sense of competency, and quality of life, and was positively evaluated by both caregivers and coaches.

This paper describes the implementation of Partner in Balance (an evidence-based eHealth intervention) as a use case to inform developers of other evidence-based eHealth interventions for the caregivers of people with dementia. Using insights from real-life pilots and stakeholder interviews, the aim of this study was to shed more light on the implementation context and aid future researchers in the implementation of similar interventions. The specific objectives of this study are to (1) formulate evidence-based implementation strategies, (2) develop a sustainable business model, and (3) integrate these elements into an implementation plan.

Ethical approval for these studies was granted by the Medical Ethical Oversight Commission of Maastricht University (approval number 2018-0489).

The following section describes this study’s findings from explorative implementation (real-life pilots and stakeholder interviews), while the subsequent section describes how these findings are integrated to achieve this study’s objectives concerning sustainable implementation (devising implementation strategies and a business model). The final section integrates these findings into a concrete implementation plan.

This first theme concerns the views stakeholders had on what was good about Partner in Balance and what could be improved in the future. The first category of this theme “good content” showed that all groups of stakeholders had positive attitudes toward the Partner in Balance content and thought many of its components were useful and timely. The second category refers to how stakeholders (especially policy makers and health care professionals) thought that Partner in Balance met caregiver needs, but emphasized that they saw it as complementary to and not as a substitution of face-to-face caregiving services. In the final category of this theme, stakeholders suggested options for additional Partner in Balance functions. These mostly centered around more contact and support, either online (through chat functions and forums) or offline (through meetings, symposia, and a contact person).

The second theme discusses what stakeholders mentioned concerning the broader eHealth context. In the experience of health care professionals, eHealth is rather difficult to implement, especially in older populations. They also felt that the topic of eHealth was important to health insurers, and the implementation often needed to be pushed by the eHealth provider. Several groups of stakeholders mentioned that eHealth is often not very user-friendly and saw this as an important barrier.

In the third theme “caregiving context,” stakeholders sketched the context in which dementia caregiving support usually takes place, as well as the associated challenges. These challenges included health care professionals’ unfamiliarity with the topic, as well as the importance of case management and how it (and dementia caregiver support in general) is organized. In terms of policy, a trend emerged across the different stakeholders. For policy makers, health care professionals, and health insurers, policy tended to focus less on caregiving and more on self-management, personalization, and positive health. These policy trends were in line with the Partner in Balance content, and this match between the intervention and current policy trends was considered a notable intervention selling point.

The fourth theme groups stakeholders’ views on the financial context of Partner in Balance. This included responses from policy makers on whether it was ethical to market an intervention developed with public money, as well as different options and calculations for various subscription and licensing models. Regarding the latter, large variations in the suggested price were observed, with caregiver contributions of €0 (US $0), €1 (US $1), €25 (US $28), or €35 (US $40) for a full course (as a way to ensure adherence) and €200 (US $226) to €700 (US $791) paid by care organizations (including the costs of training, coaching hours, and hosting). However, the majority of stakeholders did not think that informal caregivers should be the ones paying for the intervention, but rather that this should fall to the care organizations, municipalities, or health insurers (no stakeholders suggested the intervention be somehow free for all parties). The health care professionals favored a yearly subscription model, where organizations could buy licenses for the desired number of participants. In the second category “potential financers,” the Dutch national health care insurance system and how it relates to the municipal prevention mandate were the main topics of discussion. In particular, the classification of Partner in Balance as a tool for prevention (as buying a license could then be more suited to a municipality) or treatment (as buying a license could then be more suited to a health care insurer) was important. Other potential financing options were interested parties from the labor market (to combat loss of workforce to caregiver burden) and buy-in care networks (where local dementia care organizations group together in care networks).

The final theme groups stakeholders’ statements on the process for their organizations to potentially adopt, disseminate, and implement new eHealth interventions for the caregivers of people with dementia. Concerning the first category “purchase process,” the policy makers emphasized the need for the intervention to be approved at many levels, including in the budget and policy (especially for municipalities), as well as the added value of testing interventions through pilots with local collaborations. A number of evaluation criteria used by the organizations to decide whether to implement an intervention were discussed (Table 2). Most importantly, health insurers repeatedly mentioned the need for data on effectiveness and cost-efficiency. Interestingly, they emphasized that the data could be speculative and qualitative (and not necessarily longitudinal or randomized controlled). Useful outcomes with which health care organizations (such as dementia case managers) could measure implementation success were waiting list reduction, less case management hours, more referrals, more supported caregivers, more caregivers able to safely live at home, positive real-time evaluations, and more care efficiency.

The six components of the implementation plan are presented in this section. The operationalizations of the targeted CFIR subdomains were combined into components 1 to 5, while the business model canvas led to component 6. Based on these inputs, the components of the integrated plan for the sustainable implementation of Partner in Balance were as follows: (1) a ready-to-use Partner in Balance inspiration and intervision session format (live and online version and online coach training) to stimulate inner setting enthusiasm, (2) a guide for the implementing organization, in which Partner in Balance is included as a fixed offer in the first phase after diagnosis, (3) the writing of a report on financing options and cost-effectiveness, (4) efficient communication of project results through different channels, (5) a coordination plan and division of responsibility, including risk management, and (6) a sustainability plan including a licensing model. This licensing model is currently structured for an organization to pay for coaching one client per payment. The coach’s organization and municipality chooses how many coach licenses they wish to buy. Additionally, implementation costs are charged depending on the type of package desired by the implementing organization.

In previous research, the Partner in Balance process evaluation [37] reported that initial implementation challenges were related to a lack of financing and time necessary to implement the intervention. The findings from this study’s real-life pilots and stakeholder interviews made it possible to more precisely describe the previously identified issues and devise solutions by constructing a preliminary business model. In this study, stakeholders reported an unwillingness to commit resources to an intervention that might not be available in the future or that they might not be able to afford. This is in line with previous research that also advocated the application of business models to evidence-based interventions to facilitate long-term implementation [32,47]. Thus, this study contributes to literature on the implementation context by providing insights into this important implementation barrier, namely the lack of reliable pricing information for implementing evidence-based eHealth interventions to support the caregivers of people with dementia [48]. Additionally, the implementation strategies developed in this study using the CFIR helped ensure that the different components of successful implementation were considered in the business model.

There has been little research on the perspectives of the parties involved in the real-life implementation of evidence-based eHealth interventions for the caregivers of people with dementia [23]. For this reason, it was necessary to formulate a targeted implementation plan for Partner in Balance, which would help tailor Partner in Balance to this relatively underexplored dementia health care context. The implementation strategies and proposed business model resulted in an implementation plan that aimed to facilitate the integration of this evidence-based intervention into the organizational structures found in clinical practice. In this study, stakeholders in the domains of eHealth and dementia care perceived eHealth as difficult to implement and the usage data showed that 84% of SFC municipalities chose to implement other dementia-related projects instead of Partner in Balance, underscoring the role of organizations as gatekeepers in the implementation of evidence-based eHealth interventions for caregivers of people with dementia. This is in line with previous eHealth research that has cited the unfamiliarity of both implementing staff and the target population with web-based support tools as an important barrier to implementation [17,19]. Additionally, the fact that eHealth circumvents traditional health care delivery structures contributes to the difficulty many care organizations and governing bodies experience in implementing the interventions and adapting existing structures and norms to integrate them [49]. However, in the context of eHealth for dementia, the stakeholders did see Partner in Balance as needed and timely, particularly as it fits into current trending policy targets of self-management, personalization, and positive health, which have been advocated by the literature [50-52]. These findings confirm that it is important to continue to investigate and accommodate the evolving role of dementia care professionals in the context of emerging eHealth innovations and consider embedding eHealth care education into training programs for health care professionals [53,54], as proposed by the strategies integrated into this study’s implementation plan for Partner in Balance.

It is the authors’ aim that the findings presented in this study also inform future eHealth interventions for the caregivers of people with dementia and facilitate more efficient development and implementation. We present the following recommendations based on the lessons learned throughout the various phases of the Partner in Balance implementation:

(1) Health care organizations are often willing to pay for eHealth for their caregivers of people with dementia as long as the price of implementation is set, the evidence base is reliable, and the benefits to the organization are clear.

(2) It is recommended to form an “innovation cluster” with dementia health care institutions (the implementers, such as dementia case management organizations) together with parties who can buy licenses (such as municipalities), while other organizations (such as health insurers) reimburse the health care organization’s staff hours.

(3) eHealth interventions to support the caregivers of people with dementia cannot be implemented as a ready-to-go one-size-fits-all project. Offline guidance and tailoring will always be necessary. Therefore, it is important to budget for this and identify which partners will be a part of the so-called “innovation cluster” to ensure a realistic implementation plan.

(4) Finding a balance between these differing prioritizations and identifying which of the involved parties should be the financer and which should be the implementer in the dementia health care context are challenges best addressed early in the development process (preferably even before the effectiveness trial).

(5) It is important to emphasize to potential eHealth buyers that eHealth should always be complementary to other offline dementia caregiving services and not a replacement of existing face-to-face services.

(6) It is recommended to construct a preliminary business model canvas at the start of implementation (before the effectiveness trial) in order to identify all relevant partnerships, customer relationships, and revenue streams in the local dementia health care context. Doing this will allow researchers to create a product that is attuned to its specific market. If possible, it is also suggested to work with a commercial partner from the start. Using the business model canvas to inform our pricing and implementation plan was very helpful.

(7) Future developers should incorporate an explorative implementation phase after the trial context. It is necessary to flexibly explore different pricing models and iteratively address real-world implementation challenges prior to actually charging organizations.

(8) Using the CFIR helped to formulate implementation strategies targeted at many different aspects of implementation. It was particularly helpful in structuring thinking on project management, as well as the engagement of the implementing organization’s staff and management.

This study has helped fill the knowledge gap concerning the implementation context for eHealth interventions for the caregivers of people with dementia in two important ways. It has added to the existing literature by providing an example of a business model to aid the implementation of an evidence-based eHealth intervention for the caregivers of people with dementia, as well as specific implementation strategies to facilitate integration into the dementia health care context. Future research should evaluate which types of implementation strategies are most successful at achieving long-term implementation. In particular, as concluded from the stakeholder interviews, a more in-depth cost-effectiveness study is needed to encourage more active participation from health insurers and health care organizations.

This study has unique and important strengths. First, this study made use of well-established theoretical frameworks to guide implementation, using the MRC framework for the development and evaluation of intervention effectiveness, as well as the CFIR and business model canvas. Second, despite its theory-driven approach, this study illustrated a practical and real-word representation of the implementation of an evidence-based eHealth intervention. By iteratively adapting the intervention and being able to adapt with more agility to implementation issues than is normally possible in a strict trial context, this study provided a realistic view of the implementation process and context.

This study also had several weaknesses. First, though it was intended as a “real-world” illustration of bringing an evidence-based eHealth intervention to the market, the actual implementation was still very much dependent on the researchers guiding and facilitating this implementation through the research project. However, this study still provides a useful overview of the steps necessary to construct a realistic implementation plan. Second, several of the interviewed stakeholder had been involved in the Partner in Balance development in the past (four out of 14 stakeholders). This could have resulted in some positive bias to look favorably at the intervention’s future implementation. However, the authors believe it was necessary to include some interview participants who had real knowledge of the working of Partner in Balance. Finally, people with dementia were not included as stakeholders in this study. This is because the intervention was developed together with the caregivers of people with dementia, and it exists in its current form as a result of their needs and wishes. The focus of this study’s stakeholder interviews was on the surrounding implementation context and organizational determinants. Furthermore, the use experience of the intervention from the perspective of caregivers was explored in depth in the Partner in Balance process evaluation [37].

Future research will include an evaluation of the proposed implementation plan. In particular, as concluded from the stakeholder interviews, a more in-depth cost-effectiveness study is needed to encourage more active participation from health insurers and health care organizations.

Stakeholders saw eHealth as difficult to implement, but as an approach that is needed and timely, particularly as it fits into the current trends of self-management, personalization, and positive health. Applying the CFIR to devise theory-driven implementation strategies was primarily useful for targeting overlooked implementation aspects, such as ensuring effective and sustained engagement of coaches, streamlining project management, expanding and disseminating the intervention, and enhancing insights into pricing and long-term business modeling, in order to ensure sustainability. Insights from business modeling resulted in two different kinds of licensing agreements (one for municipalities and one for organizations). Finally, the authors recommend thoroughly exploring the organizational and health care contexts of the intervention and then forming “innovation clusters” (consisting, for example, of a technology developer, research team, intervention provider, and health insurer or other funder) from the start of eHealth development. This will help ensure that the intervention meets the needs of its target users (both the end users and the implementing staff).