We adopted a multimethod design including a literature review and a qualitative substudy [24]. Results were graphically synthesized into a logic model and an initial program theory was formulated comprising the theory of change, the outcomes chain, and the theory of action [17].

We applied the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses, Extension for Scoping Reviews) [25], the SRQR (Standards for Reporting Qualitative Research) [26], and the COS-STAR (Core Outcome Set—Standards for Reporting) [27] to structure our report.

We aimed to map the knowledge on technology-assisted counseling interventions for family caregivers of persons with dementia and followed the Joanna Briggs Institute methodological guidance for scoping reviews [28].

With the aim to explore the lived experience of counseling in dementia care, we conducted interviews with interest-holders in caregiving and counseling in the context of family dementia care.

Contrary to the original plan to conduct the data synthesis into the logic model and the formulation of the program theory consecutively, we brought these two working steps together in an iterative process. We switched back and forth between writing memos and graphically synthesizing the data. Following the recommendation made by Funnell and Rogers [17], we combined the inductive and deductive approach with articulating interest-holders’ mental models to develop the program theory. Mental models are diverse interest-holders’ beliefs about how a program achieves its results [17]. We integrated the mental models explicated by caregivers and counselors with data derived from literature and with information on programs operating in practice. We treated data extracted from studies included in the scoping review as qualitative data [20]. Data were charted and categorized [20], and assigned to the elements of the program theory [17].

The starting point for the development of the theory of change was the situation analysis for which we applied the guiding questions formulated by Funnell and Rogers [17]. Key aspects of the situation analysis are depicted as macro-, meso-, and microcontext in the logic model.

We then explicate our assumptions of how the intended changes can be achieved by formulating the theory of change: (hypothesized) mechanisms of effective counseling, as described by interview partners or identified in literature, were incorporated into the theory of change.

Outcomes derived from clinical studies or mentioned by participants were summarized in tables and served as the foundation of the outcomes chain. The development of the outcomes chain followed the steps outlined by Funnell and Rogers [17]: (1) listing of possible outcomes, (2) clustering outcomes and assigning working labels to each cluster, (3) arranging outcomes in a chain of if-then statements, (4) identifying feedback loops, and (5) validating the outcomes chain.

By explicating how the intervention is designed to initiate the theory of change, we developed the theory of action and identified resources (inputs), activities, and outputs [17]. We incorporated intervention components and strategies that were reported in the literature or by study participants as helpful or effective in achieving the intended outcomes in the theory of action.

We established a study advisory board consisting of a person with long-term experience in caring for family members with dementia and engagement in an informal support network for family caregivers, a person with extensive counseling experience, and an experienced researcher in the field of dementia care. Representing different groups of interest-holders, the members of the study advisory board critically reviewed the instruments and approaches for data collection and were involved in the discussion of results and conclusions for the design of a technology-assisted counseling intervention. Three virtual meetings were held in addition to feedback provided via email over the course of the 12-month project.

We applied the strategies of expert consultation and peer debriefing by consulting the members of the study advisory board throughout the research process.

The ethics committee of the Medical Faculty of the Martin Luther University Halle-Wittenberg approved the ProCOS study (no. 2023‐093). Persons interested in participating received written information on procedures prior to the interview. The time and place of the interviews were determined based on the participants’ preferences. We obtained written informed consent from the participants who were informed that the consent to participate can be withdrawn at any time. We maintained the security of data by storing data protected from access by persons who are not involved in the project. Audio recordings of interviews were pseudonymized during the transcription process. Study participants received a small gift (equivalent to US $5), but no financial compensation.

We identified the following themes in counselors’ interviews: work attitude and standards, unpredictability, expectations, working conditions, organizational influence, and tools: techniques and networking.

The common ground in the lifeworld perceptions of the two interest-holder groups forms the constitutive pattern of somebody to count on. While caregivers express the need for having somebody to count on, who guides them empathically through the challenges of family dementia care, counselors aim at being somebody to count on by competently assisting caregivers to master those challenges.

Caregivers and counselors described effects, which could be used to measure effectiveness of counseling. The potential outcomes are listed in Table 3.

Taken together, the statements of the two participant groups on effective counseling include the following characteristics: an empathetic, appreciative, and compassionate attitude; knowledge transfer; provision of materials and help with bureaucracy; providing a view from the “outside”: classifying, weighting, and sorting caregiving issues; recommendations for action (eg, dealing with changes in behavior); support in decision-making and in defining boundaries; and sustainability and continuity. Comments and suggestions of participating caregivers and counselors that have been incorporated into the program theory and that are represented in the logic model are referred to in the section Logic Model and Program Theory.

The logic model of the technology-assisted counseling intervention is provided in Figure 2.

In the following section, we outline the program theory. The detailed version can be obtained from the authors on request.

The contextual factors of family dementia care in Germany on the macro-, meso-, and microlevels are illustrated in green in the logic model: Ongoing demographic developments lead to an imbalance between care needs and families’ capacity to provide care. Challenges faced by individuals assuming caregiving responsibilities can be exacerbated by fragmented health care services and complex legal regulations. The unequal distribution of support services across communities with varying resources can result in inadequate support for family caregivers of persons with dementia.

The problem to be addressed is the inadequate use of counseling services by family caregivers of persons with dementia. Study participants indicated that caregivers are rarely made aware of the counseling services available. Moreover, caregivers often do not actively seek contact with services because they feel overwhelmed by caregiving tasks, experience feelings of shame, or hesitate due to the perceived effort involved.

In our theory of change, we outline how a technology-assisted counseling intervention may help to address this issue: We hypothesize that an increased awareness of counseling services among family dementia caregivers leads to an initial use associated with a beneficial and supportive experience.

As both caregivers and counselors emphasize the importance of continuous counseling and support, we propose that counselors adopt unobtrusive strategies to maintain engagement with caregivers. Using diverse types of ICT that caregivers are familiar with facilitates the ongoing contact with counselors. This approach fosters a sustainable, long-term supportive cooperation that enables consistent counseling throughout all stages of the caregiving process.

In the outcomes chain, we describe the immediate and intermediate outcomes, which were derived by clustering outcomes identified in the literature and those mentioned by study participants (refer to Tables 1 and 3). We assume that long-term counseling cooperation leads to improved caregiver knowledge about caregiving aspects (eg, the (potential) course of dementia, how to deal with changed behavior of the person with dementia, and available support services), to an enhanced attitude toward caregiving, and to improved interaction with the care-receiving person. This entails an altered impact of caregiving, in that the caregivers perceive decreased negative effects such as burden, distress, feeling of helplessness, and increased positive effects such as sense of confidence, empowerment, and satisfaction. Caregivers are thereby enabled to actively and purposefully shape the caregiving arrangement, for example, by using support services and to achieve congruousness of the caregiving arrangement with their own preferences and values. Caregivers perceive stability of the caregiving arrangement and safety of care despite unpredictable developments and dynamics and are aware that they may approach counselors any time needed. This leads to an improved living situation for both the caregivers and the persons with dementia. As indicated by the arrows in the logic model, the levels of the outcomes chain can occur in a recursive process.

Within the theory of action, we specify the inputs and actions through which the theory of change can be initiated [17]. The inputs specified in the theory of action are derived from the statements of study participants presented above and from intervention components described in the literature (refer to Multimedia Appendix 3). As counselors emphasized the importance of adequate working conditions, we incorporated aspects such as sufficient staffing, measures to promote a supportive working environment, and access to informational resources, including libraries and databases. Appropriate technological equipment and infrastructure, along with IT support, enable the reliable delivery of technology-assisted counseling. Furthermore, networks for cooperation with complementary services—either within the organization or in collaboration with external providers—are crucial for providing effective support. Program activities are directed at enabling counselors to provide technology-assisted counseling to family dementia caregivers competently and confidently. The activities depicted in the logic model further promote supportive working relationships by fostering mutual support within the team and articulating shared values in their work, for example, through manuals or counseling concepts. The implementation of a public information campaign may help to raise awareness of counseling services and facilitate outreach to family caregivers. In addition, diverse services are integrated to ensure timely, individualized support. Considering limited personnel, financial, and time resources, counselors are empowered to use ICT and artificial intelligence (AI) applications sufficiently to optimize the efficiency of services. Outputs represent the tangible, measurable products of activities [17]. We included the availability of the intervention manual and the information material, as well as the number of persons using counseling or other services and the number of events such as meetings to enhance cooperation within the team, within the organization, and with other organizations, as well as counts of types of ICT.

In this paper, we present the results of a scoping review and a qualitative study synthesized into a logic model and a program theory of a technology-assisted counseling intervention for family caregivers of persons with dementia. Information obtained from literature, empirical findings from the qualitative substudy, and theoretical approaches have been merged to guide the future COS consensus process and the modeling of the intervention.

Presumed mechanisms of the technology-assisted counseling intervention were informed by theoretical approaches underpinning the interventions identified in the scoping review: The model of determinants of subjective burden of caregivers of persons with dementia forms the theoretical framework of the Dementelcoach intervention [68]. Aspects such as caregivers’ personal characteristics, material and social circumstances, and the support they receive are integrated into the program theory. In addition, the concept map of the Internet-Based Caregiver Support Service (ICSS) described by Chiu and Eysenbach [92] to conceptualize usage behavior of family caregivers by integrating 3 theoretical approaches (Anderson’s model of health service utilization, Venkatesh’s theory of technology acceptance, and Chatman’s and Wilson’s information behavior theories) was used to shape our theoretical understanding. Factors such as dynamic and individual caregiver needs, perceived efforts of ICT options, and preferences in using ICT [92] have been taken into account.

Further findings of the scoping review illustrate how technologies used for delivering counseling have evolved over time from technology-assisted counseling provided exclusively via telephone to counseling via chats or SMS text messaging. In addition, multiple technologies are used in some interventions for establishing contact or delivering counseling. This diversification has led to a partial overlap in the categories formulated to group interventions in this paper.

The appropriateness of technological options for administering counseling varies with respect to availability, flexibility, and requirements for technological equipment and skills [3,8]. In addition, the suitability of the various technological approaches for discussing personal issues in depth or addressing specific requests directly is perceived differently [3,8]. Therefore, we will integrate multiple technological ways to access counseling in the development of the intervention to accommodate caregivers’ individual needs and resources and to enhance technology acceptance.

The combination of various ICT systems poses higher demands on counselors’ technological literacy and counseling competencies. To effectively switch between various modes of delivery, counselors require specific qualifications and skills, for example, to compensate for the lack of visual clues during telephone counseling, to facilitate openness in conversations via videoconferencing software, or to address counselees’ concerns in written asynchronous communication such as email or chat [8]. To enable counselors to use ICT confidently, inputs of the intervention aim at promoting an effective working environment by including adequate technological equipment and infrastructure as well as available support by IT specialists for training counselors and troubleshooting.

Caregivers and counselors participating in the qualitative study largely expressed a preference for in-person counseling. The reluctance to use ICT that we observed despite the shift toward technology-assisted services during the SARS-CoV-2 pandemic [110] could have been partly influenced by the older age of the participating caregivers. Nevertheless, Gonzalez-Fraile et al [3] state in their systematic review that remotely delivered training and support interventions appeared to be less acceptable than control interventions, as assessed by attrition rates. Therefore, we integrated face-to-face services offered at counseling centers or in caregivers’ home environments into the program theory of the technology-assisted counseling intervention.

We hypothesize that the combination of diverse technologies in addition to in-person counseling contributes to overcoming barriers to utilization of counseling services. The different characteristics of technological options allow for the targeting of distinct user groups. Findings of the qualitative study indicate that SMS text messaging is rather highly accepted among middle-aged or older persons and can be used to maintain contact and emotional involvement by sharing photos or invitations to events. Results from the literature also show that SMS text messaging is suitable for addressing disparities in access to caregiving support for persons belonging to minorities [98,106]. The needs of ethnic groups may also be met by counseling via email, offering an alternative service model [99,100]. Participants of the qualitative substudy stated that nonnative speaking persons may benefit from conversations via email as the asynchronous communication gives them more time to consider counselors’ suggestions.

As outlined above, caregivers described in detail the consequences arising from the caregiving responsibility summarized in the theme “being affected.” Feeling overwhelmed and stressed can prevent caregivers from actively seeking information and support [111]. In addition, the lack of awareness as well as the regional lack of availability contributes to the underuse of support services such as counseling [111]. We therefore included a campaign in the program theory to increase awareness among public and professional communities about counseling opportunities. By reaching out through an enhanced digital and analog presence, we aim to encourage family dementia caregivers to initiate contact to counseling services.

Inquiries about the caregivers’ mental models during the interviews indicate that family dementia caregivers often do not distinguish between the different objectives of various support services. While professionals intend to give room for sharing caregiving experiences in support groups, for example, some of the participants also expected information to be conveyed at these meetings. Embedding counseling in a variety of support services may contribute to supplementing the supportive effect of individual interventions, to maintaining contact, and to actively offering support when need occurs. Interventions offering counseling as part of a comprehensive program with non-technology-assisted components provide examples of such an approach: One example combines telephone counseling with a group activity program for community-dwelling individuals with dementia, aiming to empower persons with dementia and their caregivers [97]. Another combines telephone coaching (Dementelcoach) with respite care, thereby enhancing the effectiveness of the intervention [68]. A third example links two eHealth caregiver interventions to existing Meeting Centers for persons with dementia and their family caregivers [63,96].

With care needs increasing due to the demographic development, the shortage of skilled health care workers results in higher workloads and accelerated turnover rates [112]. Various strategies are described for retaining health care staff [112]. Authors of the Dementelcoach study reported that professionals delivering the intervention were part-time workers in the psychogeriatric sector. These professionals were encouraged to expand their work hours to provide counseling in conjunction with their existing jobs [68]. Offering a complementary activity may help to meet existing support needs in the face of the shortage of skilled staff [68].

All participating counselors reported a high workload, and the growing scarcity of skilled professionals may further exacerbate work-related stress [113]. To reduce time spent on documentation and administrative tasks, AI technology will be integrated into the intervention at the level of partial automation by augmenting human performance [113]. In using large language models for transcribing counseling sessions and composing summaries, counselors are relieved of documentation tasks and allowed to focus solely on counselees’ concerns during sessions. AI can also be used for scheduling tasks and to compile lists of regionally available support services for caregivers based on an AI-augmented database [114]. AI-drafted content will be evaluated by counselors to ensure accuracy and quality of information. This may help to alleviate concerns associated with the use of AI in the context of vulnerable groups—in particular, concerns about the reliability of information, privacy, and data security [113].

In sum, technology can make an important contribution to overcoming problems arising from the increasing need for support due to the growing number of persons with dementia [6], the growing shortage of skilled health care professionals [115], and social and infrastructural inequalities [116]. It is essential that technology-assisted counseling interventions are designed with consideration of individual and situational requirements in order to enhance their acceptability and feasibility from the perspectives of both recipients and providers [117].

We established a profound database for formulating an initial program theory and preparing for the development of a COS by integrating knowledge obtained from literature and the lived experience of two interest-holder groups. Participants in the qualitative study were recruited all over Germany. Caregivers living in urban and rural regions shared their experiences with counseling services, but we did not succeed in recruiting a heterogeneous sample in terms of age and gender. This may impose limitations on findings, as perspectives of younger and diverse caregivers, who might be more technologically savvy, are underrepresented. Younger family caregivers who are also employed, particularly long-distance caregivers, may be more inclined to use ICT to access counseling services. By combining various technological access options with in-person counseling, we advocate for an inclusive approach that addresses the needs of diverse target groups.

We were able to identify commonalities and differences in the interest-holders’ expectations and mental models, which were incorporated into the program theory.

Another limitation is the inclusion of only two groups of interest-holders. Due to the limited duration of the project, we were not able to expand the study population of the qualitative inquiry. Representatives of providing organizations and policymakers will be included in the consensus process of the COS.

We followed accepted methodological guidance [13,17] and integrated methodological approaches in an innovative way. This proved to be beneficial: When asked about potential outcomes for measuring effectiveness of counseling, interviewees reflected on and critically questioned their own ideas and expectations. We found it advantageous to consider adequate outcomes from the beginning of the development process, to debate underlying (hypothesized) causal relationships, and to visualize causal assumptions in the logic model. This theory-led approach to the development, implementation, and evaluation of a technology-assisted counseling intervention for family caregivers of persons with dementia is consistent with the recommendations of the updated UK Medical Research Council guidance [9].

We will perform a theory-led approach to modeling, implementing, and evaluating the technology-assisted counseling intervention for family dementia caregivers. In a first step, the “long list” of outcomes will be created in accordance with the COS development methodology [13]. The participants of the qualitative study found it surprisingly difficult to name potential outcomes for assessing the effectiveness of counseling interventions, which led to a high number of (partially) overlapping outcomes. In a follow-up study, we will involve representatives of interest-holder groups such as family caregivers, counselors, managers of provider organizations, and researchers to integrate outcomes reported in literature (Table 1) and outcomes identified by interview partners (Table 3): Redundant outcomes will be removed; for example, outcomes such as “burden,” “guilt,” and “stress or distress” were identified in both the literature and interviews. The remaining outcomes will then be grouped into outcome domains applying ontologies, as proposed by Williamson et al [13]. The resulting “long list” of clustered outcomes forms the basis for the consensus process of the COS [13]. We will apply the Delphi approach involving groups of relevant interest-holders to determine important clinical outcomes [13]. To establish a shared understanding among participating interest-holders, elements of the program theory will be gradually integrated into the consensus process. At the outset of this process, we will use the logic model to visualize and critically discuss the theory of change. Subsequently, the importance of individual outcomes will be rated in Delphi rounds and finally consented in a consensus conference [13].

The refined logic model and program theory will then be used for modeling the intervention in collaboration with practice partners in a future research project. The logic model will serve to foster a shared understanding among the individuals involved and to systematically document the adaptation of key components to the specific context [9,18].

We developed an initial program theory of a technology-assisted counseling intervention for family caregivers of persons with dementia by introducing a methodological innovation. Findings obtained from interest-holder groups and literature are synthesized into a program theory and visualized by a logic model. We have also compiled a comprehensive list of potential outcomes, which includes the outcomes examined in clinical studies and those that are relevant from the perspective of interest-holders. This enables the consensus process for finalizing the COS for technology-assisted counseling interventions.

These results will inform the theory-led modeling, implementation, and evaluation of the intervention, which will include a customized ICT package. This package has the potential to improve accessibility to counseling for caregivers by overcoming disparities in access to health care services. In addition, the design of the intervention can positively impact work conditions for health care professionals delivering support and improve the efficiency of services.