Dementia is a progressive condition with no current cure [1]. The condition is considered a global public health challenge and is the seventh leading cause of death worldwide [2]. It is characterized by impairments such as cognitive, functional, and behavioral [1] affecting the individual’s memory, behavior, and daily functioning [2]. In 2020, the number of people living with dementia worldwide was 55 million, with a projection of 139 million in 2050. In Sweden, 150,000 people lived with dementia in 2019, with a projection of 250,000 by 2050 [3].

Most of the care provided to people with dementia in communities is provided through informal care by unpaid family caregivers (FCs) such as family members and friends [4]. Therefore, FCs play an integral part in the health care system, providing everyday care that is complex, challenging, and prolonged [1]. In 2019, the World Health Organization (WHO) estimated that FCs, on average, provided care and supervision to a person with dementia for 8 hours per day [5]. Most FCs of people with dementia in Sweden are older partners (including spouses and cohabiting partners) and women [6], and a recent study reported that about half of spouses of people with dementia provide care for ≥30 hours per week on average [7].

Positive aspects of informal caregiving include feeling satisfaction and pride in contributing to the quality of life of the person with dementia [8]. However, FCs often face challenges managing the person with dementia’s complex and challenging symptoms of dementia such as restlessness, agitation, and physiological distress [9]. The increased dependency of the person with dementia as the condition deteriorates poses another challenge for FCs, often leaving them overwhelmed [8]. The added responsibilities of FCs occur gradually and tend to increase over time as the condition progresses. This may lead to constraints in their daily lives, such as having to leave work [10]. The challenge of informal caregiving may lead to caregiver stress, which is multifaceted and related to dimensions such as psychological, physical, and social burden due to factors including limited time resources, required vigilance, and supervision [11]. The challenges of informal caregiving can also lead to caregiver depression. The risk factors of depression among FCs of people with dementia include being a woman, being a spouse, and cohabitating with the person with dementia. Caregiver depression among FCs of people with dementia is also related to functional decline and mortality [12].

The formal support provided to FCs in Sweden is most commonly indirect, in that the support is aimed toward the care recipient (eg, home care). Direct support to FCs is often provided by social care professionals in the municipalities. The educational backgrounds of social care professionals vary and include nurses, social workers, and nurse assistants. Social care professionals have an essential role in the development and coordination of support for FCs, including providing counseling through individual and group support and through education (eg, coping strategies) [13,14]. This type of support provided to FCs in Sweden [15] is often divided into three types of resources: (1) instrumental (ie, material aids and help with daily tasks), (2) informational (ie, relevant information, education, and provision of knowledge), and (3) emotional (ie, empathy, care, and validation) [15].

The National Board of Health and Welfare, through the Social Services Act (SOSFS; 2001:453), obligates municipalities to provide support to informal caregivers [16]. Despite this, most FCs are reported to be unaware of the available support services and who to turn to for support [6]. A study including FCs, people with dementia, and professionals conducted in eight European countries, including Sweden, explored the barriers and facilitators to accessing and using formal dementia care. The study found that a lack of knowledge about dementia and available services among FCs of people with dementia contributed to them not seeking needed support [17]. FCs of people with dementia reported retrospectively that they had waited too long to make use of professional support [17]. In line with this, research shows that FCs often delay seeking formal support services and neglect their personal needs [18] until experiencing a crisis [17,18], such as a sudden inability of the FCs to provide care to the person with dementia or behavioral symptoms of the person with dementia [17]. External support is also often sought when FCs no longer perceive they can function effectively due to increased responsibilities [10]. The competencies of social care professionals, including specific knowledge of dementia and having a key contact person, are important for access to formal care services [17]. In addition, a recent systematic review [19] aimed at identifying the needs of FCs of people with dementia living at home found that receiving tailor-made information to meet one’s own needs as FCs was important [19].

The increasing need to support FCs of people with dementia has brought forth the development of innovative interventions delivered through mobile health (mHealth) technology, such as mobile app [20]. mHealth is defined by the WHO [21] as a medical and public health practice supported by mobile devices, such as mobile phones and other wireless devices. The benefits of mHealth technologies are that they are accessible and therefore reported to be easily incorporated into the daily lives of FCs of people with dementia when time permits [22]. The accessibility of mHealth technologies also increases the potential of reaching people who live long distances from health care services and those who are unable to meet in person [23]. Another benefit of mHealth technologies is their potential to provide cost-effective support that is personalized and needs-based by facilitating the connection between the health care provider and the health care recipient in real time [23]. However, a recent meta-analysis found that there was a limited number of interventions using mobile apps targeting FCs of people with dementia living at home [24]. There are also limited available apps that are directed toward the personal needs of FCs, such as stress management [25]. Furthermore, there are limited studies that assess the effectiveness of interventions through mobile apps in terms of health-related outcomes such as caregiver stress and depression of FCs of people with dementia living at home [24].

Considering the aforementioned gaps and the potential use of mHealth technologies in bridging these needs, we developed a tailor-made mobile app named STAV (STöd till AnhörigVårdare: support to FCs) with Karolinska Institutet eHealth Core Facility and byBrick Elevate [26]. The mobile app STAV consisted of (1) a chat feature that enabled FCs and social care professionals to interact, (2) a digital diary that enabled FCs to note any stressful or upsetting events such as behavioral changes in the person with dementia, (3) a mindfulness feature offering access to 3 guided sessions, each averaging 5 minutes in length, (4) a contact feature for registering relevant contact information, and (5) a web links feature with a curated collection of updated and relevant information regarding caregiving and dementia. The app was available on both iOS (Apple Inc) and Android (Google) digital platforms [26]. STAV was developed based on ideas and insights from user perspectives following individual interviews with FCs of people with dementia living at home and a focus group interview with community-based social care professionals specialized in dementia care [27]. The benefits of receiving support through the app for FCs have previously been reported, including offering a foundation for inner calm and relevant information in one place [28]. To gain a more comprehensive understanding of the potential of professional support delivered through the mobile app STAV, a mixed methods research design was used [29]. This approach enabled us to gain insights into how changes in the quantitative outcomes were related to the type of support provided by social care professionals as recorded in chat conversations, an approach that has rarely been used [30]. This study aimed to examine, among FCs of people with dementia living at home, (1) the potential association between professional support provided through STAV and caregiver stress and depressive symptoms, (2) the types of support provided through chat interaction between FCs and social care professionals, and (3) how provision of support through a mobile app relates to changes in caregiver stress and depressive symptoms.

An explanatory sequential mixed methods design was used in this study. This method is useful when applying qualitative data to explain quantitative results [29]. The pilot study adhered to the Good Reporting of A Mixed Methods Study (GRAMMS) checklist [31].

The study was conducted among FCs of people with dementia living at home in municipalities located in three regions of Sweden: Stockholm County, Västerbotten County, and Skåne County. The social care professionals who provided support through the chat feature of STAV were purposively sampled and worked in a municipality in Sweden, providing community-based support to FCs of people with dementia living at home. The social care professionals who participated in the study consisted of registered nurses specialized in dementia care, dementia coordinators, social workers, social aid officers, and nurse assistants specialized in dementia care, and worked in the municipalities or senior day care centers for people with dementia. The FCs of people with dementia were purposively recruited through the networks of social care professionals. Other channels of recruitment were local health care centers, websites, newsletters, and social media channels of the municipalities, dementia-related associations, and FCs’ networks. The eligibility criteria for the FCs included being an adult aged older than 18 years, who provided care for a person with dementia living at home for a minimum of 6 months, could read and write Swedish, possessed a smartphone or tablet or both, and had access to the internet at their own expense [26]. STAV was available on both iOS and Android devices. The participants received a brief introduction to STAV and, when needed, technical support by a member of the research team either via the telephone, in person, or both. Forty-six FCs were recruited for the intervention, all of whom completed the intervention, including the preintervention assessments. Of the 46 FCs, 11 (23.9%) FCs were lost to follow-up. Therefore, a total of 35 FCs were included in this study. This sample size is considered sufficient for a pilot study [32,33].

The 8-week intervention in terms of professional support was delivered by community-based social care professionals through the chat feature of STAV. The support was delivered through synchronous and asynchronous chat conversations tailored to the individual needs of the FCs [26]. The other support features of STAV—the contact, the diary, web links, and mindfulness feature—were also available for the FCs to use at any time. Before the study commenced, social care professionals received an information package containing details about the intervention and the intervention tool STAV [26]. The social care professionals also participated in an introductory meeting, either in person or online, with members of the research team, to become familiar with the project. The intervention was implemented within the participants’ existing social care context and was tailored and delivered based on the FCs’ individual needs. As such, no additional training or retraining on the support content was provided beyond the initial introduction. However, regular contact was maintained between a member of the research team and the social care professionals throughout the intervention period. FCs were also given a brief introduction to the intervention tool STAV. In addition, the backend server where the logged chat was stored was monitored regularly. Members of the research team were available to provide technical support to both the social care professionals and the FCs when needed.

The 8-week intervention had 2 data collection time points for the quantitative element (preintervention and immediately after the intervention). Data were collected through self‐administered questionnaires. Preintervention data were collected from the FCs through the mobile app STAV after obtaining informed consent and after they had downloaded the app. Follow-up data were collected through postal questionnaires.

Data for the qualitative element were collected using logged chat data based on chat conversations between the FCs and the social care professionals during the 8-week intervention. The logged chat data were stored in a back-end server of STAV and hosted by Karolinska Institutet.

The study was approved by the Swedish Ethical Review Authority (Dnr: 2019‐01632 and Dnr: 2020‐06882) and conducted according to the Declaration of Helsinki [42]. Informed written consent was obtained from the participants before the study. The participants were informed that their participation was voluntary, including consent to retrieve the logged chat data for research analysis. All the data were pseudonymized and kept confidential. No identifiable information was linked to the participants in the presentation of the results. The research data collected through the mobile app were securely stored on a backend server at Karolinska Institutet.

Thirty-five FCs of people with dementia living at home completed the pre- and post-intervention evaluations. The FCs were aged 42-85 years (mean 69.4, SD 11.9 years) and most (28/35, 80%) were women and partners, including spouses and cohabitating partners (24/35, 68.6%), who lived together with the person with dementia (26/35, 74%). Almost 86% (30/35) of the people with dementia received respite care (data not shown). The baseline characteristics are summarized in Table 1.

The mean score of FCs’ caregiver stress was marginally higher postintervention (mean 24.1, SD 9.3) compared to preintervention (mean 23.9, SD 9.2), although this difference was not statistically significant. The mean score of FCs’ depressive symptoms declined between preintervention (mean 6.5, SD 5.1) and postintervention (mean 6.2, SD 5.2), but was also not found to be statistically significant.

Participation in the intervention did not have a significant association with caregiver stress and depressive symptoms among the FCs of people with dementia in this study. However, the effectiveness of internet-based or mobile app interventions for FCs of older adults has previously been shown for psychological outcomes, including reduction in depression and caregiver stress [43]. The results showed that informational support, followed by emotional support, was the most mentioned support type. A review [44] reports that some RCTs based on multicomponent interventions, including the provision of informational support, have shown positive impacts on outcomes such as perceived stress and depression. However, to be considered useful, the information provided must be tailor-made to the needs of the FCs [44]. This impact was potentially indicated among the subgroup of FCs with decreased stress who received support in the chat by the social care professionals, tailored to their needs (ie, reassurance, validation, and information) to a greater extent than those FCs with no change or increase in caregiver stress postintervention. However, it is important to note that the impact of informational support is difficult to assess since it is often integrated as a component of a broader intervention [44], as for the mobile app STAV, where web links to available resources are included. Although these results are not generalizable, the mixed methods analysis gave important insights into how the changes in the primary outcome, that is, caregivers’ stress, were related to specific types of support provided by the social care professionals through chat conversations with FCs.

The heterogeneity of FCs of people with dementia also needs consideration when designing online-based interventions. One of the main implementation challenges includes the condition of the care recipient (ie, the variation in the severity, type, and stage of dementia) [45], which was not considered in this study. However, the FCs with no change in stress between pre- and post-intervention generally had low engagement in the chat, which might be indicative of a low support need related to caring for a family member with early-stage dementia, as reported in a previous qualitative evaluation of the intervention [46]. Another important factor that needs consideration is the relationship between the FCs and the person with dementia. High caregiver stress is more prevalent among FCs who cohabitate with the person with dementia than among noncohabitating FCs [47]. In this study, a significant association was found, as partners, including spouses and cohabitating partners, were more likely to report higher levels of depressive symptoms than others (ie, adult children). These results are in line with a recent Swedish study where spouses reported significantly higher mean scores in negative outcomes of informal caregiving, including “physical and psychological stress” and “trouble finding time to spend with friends,” than nonspouse FCs [7].

The interaction between the FCs and the social care professionals through the chat feature of STAV could also have impacted the outcome of the intervention on caregiver stress and depressive symptoms. A systematic review [44] reports that interacting directly with professionals (ie, nurses, social workers, or occupational therapists) through internet-based interventions provides FCs of people with dementia living at home with a positive experience. Furthermore, receiving tailor-made emotional and practical advice can decrease feelings of isolation for FCs [44]. Research [45] further suggests that higher engagement and lower intervention dropout rates occur when human contact (eg, health care professionals, research teams, and coaches) is incorporated in the intervention, as opposed to an intervention with no human contact, depending on external sources of interaction (ie, providing a social media page or a telephone number) [45]. Nurses described in a Norwegian study the importance of offering FCs of people with dementia the opportunity to vent their frustrations remotely through telecare using a web camera and a web forum [48]. Similarly, the intent of the chat feature of STAV in this study was to provide support remotely. The results potentially indicate that receiving individualized emotional support in terms of validation and self-care prompts, as opposed to generic validation, had a greater impact on caregiver stress. This was identified for the group of FCs with decreased caregiver stress postintervention, who also mostly showed improvement in depressive symptoms.

There were varied levels of engagement between the FCs and the social care professional through the chat. The FCs with decreased and increased stress had relatively frequent interaction, while FCs with no change generally had low engagement. A previous qualitative evaluation of the intervention [46] reported that some of the FCs and the social care professionals were unfamiliar with each other before partaking in the intervention. Some of the FCs also received supplementary forms of caregiver support during the intervention period (ie, face-to-face support groups). This variance between the interplay of interacting through the chat and meeting in person needs consideration, since text-based platforms, including the mobile app STAV, lack nonverbal cues, which can be expressed through in-person conversations (ie, touch and facial expressions) [49]. This could have contributed to the engagement levels of the FCs and the social care professionals.

Recent research has shown that digital literacy skills and education needs of FCs of people with dementia are significantly associated with the use of mHealth apps [50]. The fact that most of the FCs in this study were older partners, including spouses and cohabitating partners, of people with dementia could have contributed to some challenges with digital literacy. The motivation, privacy concerns, training, and education of the FC of people with dementia further determine eHealth technology implementation [51]. Therefore, it is suggested [51] to support FCs in the use of technology before implementation. It is also important to provide the FCs with adequate time for practice and provide in-person assistance in case of technical issues [51]. Most of the FCs received a brief introduction to the mobile app STAV used in this intervention. Members of the research team also assisted with technical issues when needed. However, this was mostly conducted via telephone conversations due to the COVID-19 pandemic and at the wish of the FCs. Thus, conducting in-person training sessions with the FCs could have resulted in increased uptake of the intervention for the FCs who had challenges with digital literacy. The duration of the intervention, consisting of 8 weeks, could also have impacted the intervention outcomes because a lack of digital literacy can increase the need for additional effort and time to adapt to the new technology. Subsequently, some of the FCs in this study might have needed an onboarding process where enough time was provided to familiarize themselves with the new technology at their own pace [52]. Finally, other factors that could have affected the associations of the intervention were not considered, such as the education and financial level of the FCs and whether the caregivers lived with or had support from other family members.

The participants of this study consisted mostly of women. Even though the sex distribution is representative of FCs of people with dementia in Sweden [6], a larger proportion of male FCs is needed to explain sex differences in the association between the intervention and the outcome measures. The 8-week duration of the intervention might be too short to bring about significant changes in caregiver stress and depressive symptoms. Digital interventions such as STAV have the potential to reduce selection bias by reaching individuals who are typically underrepresented, such as those living in remote settings [53]. It is possible that the study primarily attracted FCs who were motivated to receive the intervention and were digitally literate. As a result, FCs who needed support but found it challenging to use a mobile app as a mode of communication could have been excluded, potentially leading to a selection bias among FCs of people with dementia. Therefore, the use of the mobile app should serve as a complement, rather than a replacement for traditional means of support. Future studies should also consider the heterogeneity of FCs to the person with dementia, such as the stage of dementia of the care recipient. Future studies could integrate outcome measures with logged analytics data to further examine associations between the digital intervention and the outcome measures.

Patterns observed in the mixed methods analysis at the subgroup level suggest that tailored support provided through asynchronous and synchronous chat delivered through a mobile app may be associated with a potential decline in caregiver stress among FCs of people with dementia living at home. However, a full-scale study with a larger sample size is necessary to confirm these findings and generalize the results of this pilot study to a broader population. These findings highlight the importance of providing frequent and tailor-made informational, emotional, and instrumental support to meet the individual needs of FCs of people with dementia. The findings from this study can help community-based social care providers plan and organize the content of digital support provided for FCs of people with dementia living at home.