The demands of an aging population and a commensurate increase in poor health and multiple morbidities mean that the ongoing provision of social care services to older people with assisted living needs is a key challenge facing the UK. Social care services (SCS) have long identified digital care technologies as a means of supporting the growing number of older people with such needs. Local Authorities (LAs) continue to invest in Technology Enabled Care Services (TECS) to enable older people to ‘age in place.’ With the provision of ongoing domiciliary care packages, including the deployment of assistive technologies, people can remain at home longer by receiving faster discharge from the hospital or delaying entry into residential care services. TECS routinely deploy first generation pendant alarms, which have been an established part of care systems for decades. recently, second generation ‘telecare’ remote monitoring services, which can collect and aggregate a wealth of sensor data to provide smart monitoring of a person’s activities and health states, have become a core element. Technological change in the sector continues to gather pace, and social care services are currently exploring the potential of integrating an increasingly diverse range of digital, ‘smart’ care and consumer technologies (third generation) with the potential to truly transform digital care services. Such technologies include novel ‘off the shelf’ consumer products, such as mobile Global Positioning Systems (GPS), equipped monitoring alarms, smart speakers, smart phones, tablets, and apps. Wider near-future innovations, such as artificial intelligence and machine learning platforms, social robots, and chatbots, are also increasingly seen as potential options for deployment in TECS as an embryonic fourth generation of technology in care ( Hamblin & Lariviere, 2023).

When combined, these four generations of technology have the potential to transform TECS provisions for older adults. However, our best evidence to date regarding social care TECS indicates that traditional TECS do not achieve their therapeutic goals and are not cost-effective ( Gathercole et al., 2021; Gaugler et al., 2021; Henderson et al., 2014; Howard et al., 2021; Steventon et al., 2013; Woolham et al., 2018; Wright, 2020). While TEC services remain largely based on 2nd generation telecare platforms, recent research shows that services are starting to embrace the potential of embedding a more diverse range of digital care technologies based on predominant policy discourses that continue to state that such innovations will inevitably bring about improvements in services ( Hamblin & Lariviere, 2023; Whitfield & Hamblin, 2025; Wright & Hamblin, 2023). If delivered successfully, these novel technologies possess the potential to deliver innovative, effective, cost-effective, and scalable person-centred services that are better suited to an individual's needs. However, evidence from past research on technology adoption suggests that the benefits of technology innovations in care are not inevitable, and any improvements experienced are likely to depend on the technological, policy, and service delivery contexts in which technologies are delivered alongside the local requirements of need ( Greenhalgh et al., 2017).

The highly fragmented nature of adult social care, involving numerous public, private and third sector actors means significant barriers stand in the way of achieving this vision across digital care services ( Freeman et al., 2022). The implementation of digital innovations in social care can be characterized by ‘pilotitis;’ a preponderance of pilot evaluations conducted by individual LAs which are only rarely shared across the sector and are not routinely delivered at scale ( Whitfield & Hamblin, 2022; Woolham et al., 2018). Additionally, LAs frequently lack technical capacity, capability, and resources to assess the effectiveness of these technologies ( Wright, 2020). Finally, local authorities may use different configurations of technologies and services based on local client needs, budgetary constraints, commissioning requirements, or relationships with technology suppliers. Combined with the high number of small-scale pilots and lack of larger-scale evaluations, such disparities may lead to ‘postcode lotteries’ where technologies become unequally distributed, further exacerbating digital, social, and age-based inequalities ( Whitfield & Hamblin, 2022). In sum, there is limited evidence to illustrate exactly what novel digital technologies are being deployed across social care, what their delivery looks like in practice, what their key outcomes are, whether they are achieving them, the factors influencing their implementation in services and adoption by service users, and what might help or hinder their delivery at scale.

The aim of the DIALOGUE project is to begin the process of mapping out a future research agenda regarding the development and implementation of personalized, person-centred TECS within social care. Using a range of research methods, including co-production with older users of social care services, DIALOGUE sought to identify research priorities in relation to novel TECS within social care to develop a future research agenda for the incorporation of novel technologies within social care-based TECS. This agenda includes research applications to major funders to grow the evidence base regarding current and future provision of technologies within social care services, priority areas where such technologies are being or should be deployed, and the local factors that will influence commissioning, delivery, implementation, and adoption of novel technologies and services within social care settings.

WP1 provides an overview of co-production workshops conducted across three sites located in England: Middlesbrough, Bristol, and Central England. For each workshop (x4) per site, participants were presented with a series of scenarios, whereby the character and various technologies were described in various settings. This approach enabled the workshop participants to consider solutions and discuss potential challenges and concerns. Future considerations for expanding this work include more diverse voices from broader ethnic communities, including people who are digitally illiterate and/or who have limited digital literacy skills.

The findings reported in this WP highlight broader concerns and issues that should be considered by local authorities and ICBs that wish to deploy TEC as part of their social care provision. The insights ascertained through these workshops align with and are associated with aging in place ( van Hoof et al., 2017), and age-friendly cities and communities ( van Hoof et al., 2025). The delivery of social care provision within the community via local authorities or ICBs aligns with Fulmer et al.’s (2020) broader discourse and age-friendly environment framework. There is an appetite for using TEC within the community social care context, based on the argument that such technologies will reduce costs and improve efficiency. Workshop participants acknowledged that the use and deployment of technologies in social care would potentially reduce costs for the NHS, despite limited evidence for the effectiveness of related technology delivered at scale ( Woolham et al., 2021). However, such accruals may be generated outside social care, creating an incentive not to provide such services if benefits are not experienced in social care unless such services are integrated – a goal of service reorganizations such as ICBs.

What was clear was the recognition among participants that there were no one-size fits all, or a silver bullet. Our communities and the people that live within them are not homogeneous, and service providers, local authorities, and ICBs must acknowledge their critical role in this discourse and their responsibility to ensure there is adequate and sufficient support in place (e.g., signposting, point of contact). However, local authorities and ICBs need to be mindful to ensure that not everyone in the community will have experience of using technologies and may not have familial support that they can ask to help them. What is needed and is demonstrated here through the findings of the various workshops, is motivation to use technology, knowledge, skills (digital literacy) support, point of contact, and appropriate signposting by service providers and local authorities alike to service users. Digital skills, broader knowledge pertaining to cybersecurity, and staying safe online ( Marston, forthcoming) are vital when there is an intention to centrally place TEC to deliver social care.

Considering the findings presented here, we propose three key recommendations that we perceive as initial mechanisms for local authority and ICB social care provision.

Adequate signposting, and information seeking. This would include both digital and analogue forms of information sharing and reach residents who are not digitally connected (e.g., social media, email). Thus, they would not see local authority discoveries shared through mailing lists or Facebook pages.

Safety, security, and Trust. This relates to ensuring that data privacy is secure but also provides knowledge and information to users about cybersecurity information (e.g., password protection, 2-step authentication). Trust plays a significant role in the acceptance and effectiveness of technology ( Lie et al., 2016) and building trust can take time. Learning from peers in the community ( Baker-Green, 2013) is imperative, and the work of the Rekindle group, together with evidence-based research ( Baker-Green, 2013; see Marston et al., 2022, Ch6) for more community suggestions.

Cost and finance are important for participants and community residents. The findings show that cost can be a barrier, but some are willing to pay for a service. However, not everyone in has the same level of finance. Many people live in poverty, hand-to-mouth, and this includes not only older adults but also younger vulnerable adults with disabilities ( Earle & Blackburn, 2023; Marston, forthcoming). This theme needs to be unpacked more to understand whether finance is more important ( Wasserman et al., 2024) than other age-friendly domains ( World Health Organization, 2007) being investigated ( van Hoof & Marston, 2025; van Hoof et al., 2025).

WP2 undertakes a literature review and evidence synthesis of academic and practical literature to identify examples of novel digital care technologies in social care and to explore outcomes and effectiveness. We conducted two literature reviews: one focusing on novel digital care technologies to inform the DIALOGUE study and one that specifically supported the development of the NIHR HTA grant under Objective 4. This step was brought forward owing to changes in the work sequence of the DIALOGUE project, which occurred before the broader literature review was completed (after the systematic searches and during the screening process). The review was undertaken as a rapid scoping review following the staged approach developed by Arksey and O’Malley ( Arksey & O’Malley, 2005) and Tricco et al. ( Tricco et al., 2015). Its reporting is based on PRISMA-ScR the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) ( Tricco et al., 2018) guidance with a PRISMA flowchart ( Figure 1). We used Covidence software to support both screening and data extraction processes.

The research questions were developed from the DIALOGUE project objectives:

•   What are examples of novel digital care technologies in social care?

•   What is known about their experiences, outcomes, and effectiveness (including cost-effectiveness)?

•   What recommendations have been identified by existing research that can inform the use and provision of digital care technologies?

•   What information from existing research could inform the development of the HTA application?

We collaborated with the King’s College Library’s Clinical Library Services to develop an effective search strategy. With the support of experienced librarian, a comprehensive systematic search design was established. Several concepts informed the search strategy: ‘Intervention’: synonyms for assistive technology, care technology, telecare, social care devices; ‘Setting’: community social care, own environments, assisted living; ‘Population’: older people, dementia. A systematic search of Ovid MEDLINE was conducted on April 28, 2024, yielding 3699 references. A second systematic search was conducted in Embase on May 20, 2024, yielding 8630 references. To include practice and grey literature, manual searches were undertaken on platforms such as the TECS Association, Digital Health and Care Institute, Social Care Institute for Excellence, and Google’s search engine. Due to changes in the work sequence in the DIALOGUE project, the findings from this search, along with an additional systematic search, informed the development of the HTA bid, but not the wider literature review.

It was decided that while there were duplications, there was not enough time and resources in the study to complete the systematic screening of over 8000 references; therefore, the review concentrated on the references and publications yielded from the MEDLINE database. Although it is not uncommon for a rapid literature review and searches to use only one database ( Klerings et al., 2023), it is a limitation.

After the removal of duplications, 3689 references were included for initial title and abstract screening, and 276 publications were selected for full-text review.

Eligibility criteria

The eligibility criteria for screening publications were developed using the PICOS framework ( Table 2), which defines the inclusion and exclusion keywords under the heading Population, Intervention/Exposure, Comparator/Context, Outcome, and Study Characteristics.

Criteria are summarised in Table 2:

Screening and analysis process

Three researchers (NS, CWN, and GG) engaged in the screening and extraction process. The screening process encompassed two stages: in the first stage, only the title and abstract were screened to determine which publications should be included in the second stage and the full-text screening. Two researchers screened all references. Using Covidence software ensured that researchers came to the same conclusion without knowledge of other researchers’ decisions to reduce the risk of researcher bias.

Following the screening process, 30 publications were included in this review. Data were extracted from these publications by using a systematic template. The Covidence software provides a basic template, which was heavily amended to record additional, qualitative, information. One researcher performed data extraction using this template. A second researcher then checked the recorded data and amended if necessary. Finally, Covidence provides a collection of the extracted data in form of a downloadable file for further analysis. Reflecting use of a rapid review methodology, publications were not systematically assessed for quality.

The analysis used a thematic approach ( Braun & Clarke, 2022) with a focus on experiences, effectiveness, and implications/recommendations. Two key frameworks were identified to inform the analysis. The first is the ‘Evidence Standards Framework’ (ESF), developed by the UK’s National Institute for Health and Care Excellence ( National Institute for Health and Care Excellence, 2025) and the second is the ‘Non-adoption, Abandonment, Scale-up, Spread, and Sustainability’ (NASSS) framework, developed by Greenhalgh et al. (2017). Both frameworks provide structured criteria for evaluating the effectiveness of technologies, acknowledging the complexities involved in the provision of care. However, it is important to note that both frameworks were primarily designed for the healthcare context rather than social care.

Following the data extraction process, it was found that six publications reported on two wider studies, the LIVE@Home.Path trial in Norway ( Puaschitz et al., 2021; Puaschitz et al., 2023) and the ATTILA study in England ( Curnow et al., 2021; Gathercole et al., 2021; Howard et al., 2021; Lariviere et al., 2021). All publications were analysed to ensure that all data were included, but to avoid duplication, study characteristics are presented at study level, while data and findings are reported at publication level, reflecting varied analytical aims.

Of the 26 included studies, six were conducted in the United States of America ( Arthanat et al., 2024; Bergschöld et al., 2020; Kerssens et al., 2015; Rose et al., 2018; Turner & Berridge, 2023; Warner & Tipping, 2022), five in Norway ( Gedde et al., 2021; Jentoft et al., 2014; Karlsen et al., 2019; Øderud et al., 2015; Puaschitz et al., 2021; Puaschitz et al., 2023). Two, one with four included publications, were conducted in the United Kingdom/England ( Curnow et al., 2021; Gathercole et al., 2021; Gibson et al., 2015; Howard et al., 2021; Lariviere et al., 2021). Another two in the Netherlands ( Bults et al., 2024; Huygens et al., 2021) and two in Sweden ( Borg et al., 2022; Malmgren Fänge et al., 2020). There was one study each from Canada ( Ault et al., 2020), Finland ( Nauha et al., 2018), Germany ( Megges et al., 2018), Hong Kong ( Lau et al., 2019), Italy ( Rossetto et al., 2023), Singapore ( Cao et al., 2022), and South Korea ( Kim & Jung, 2023). Two studies were multi-country studies, with Bergschöld et al.’s (2020) study conducted in both the Netherlands and Norway, and König et al. (2022) involving three countries: Austria, Spain, and Italy.

Aims and focus

As per the inclusion criteria, all studies explored or assessed effectiveness of technology or programmes involving technology, with two explicitly aiming to examine clinical effectiveness ( Howard et al., 2021; Megges et al., 2018). The experience of using and potential acceptance of technology among the older people participants was examined in 14 studies ( Bergschöld et al., 2020; Berridge et al., 2019; Borg et al., 2022; Bults et al., 2024; Cao et al., 2022; Gibson et al., 2015; Karlsen et al., 2019; Kerssens et al., 2015; König et al., 2022; Lariviere et al., 2021; Puaschitz et al., 2023; Rose et al., 2018; Turner & Berridge, 2023; Warner & Tipping, 2022), and three studies ( Bergschöld et al., 2020; Borg et al., 2022; Cao et al., 2022) also aimed to explore the selection of technologies by older people.

Several publications aimed to investigate the effectiveness for people with dementia and/or other cognitive impairments: nine publications ( Arthanat et al., 2024; Borg et al., 2022; Gathercole et al., 2021; Karlsen et al., 2019; Kim & Jung, 2023; Malmgren Fänge et al., 2020; Nauha et al., 2018; Øderud et al., 2015; Warner & Tipping, 2022) focussed on how technologies might be increasing their independence, four ( Ault et al., 2020; Curnow et al., 2021; Megges et al., 2018; 27) how technologies were impacting on certain symptoms, impact on safety was examined by seven studies ( Arthanat et al., 2024; Borg et al., 2022; Curnow et al., 2021; Gathercole et al., 2021; Lau et al., 2019; Malmgren Fänge et al., 2020; Warner & Tipping, 2022), three ( Borg et al., 2022; Jentoft et al., 2014; Lariviere et al., 2021) explored how technologies could improve social relationships and activities, and four ( Bults et al., 2024; König et al., 2022; Lau et al., 2019; Rossetto et al., 2023) looked at effectiveness in more general terms.

This review does not include any publications that solely focus on family or unpaid carers/caregivers, nevertheless, due to involving multiple populations, 12 publications ( Arthanat et al., 2024; Ault et al., 2020; Berridge et al., 2019; Curnow et al., 2021; Gathercole et al., 2021; Gibson et al., 2015; Jentoft et al., 2014; Lau et al., 2019; Megges et al., 2018; Nauha et al., 2018; Puaschitz et al., 2023; Warner & Tipping, 2022) also focussed on impact on carers, that is carer burden and quality of life, and six ( Gedde et al., 2021; Gibson et al., 2015; Lariviere et al., 2021; Puaschitz et al., 2023; Rose et al., 2018; Warner & Tipping, 2022) explored experiences and potential acceptance of technologies among carers. Three publications ( Gedde et al., 2021; Huygens et al., 2021; Puaschitz et al., 2021) aimed to investigate the nationwide access and routine use of technologies. Cost-effectiveness was given as an aim in three publications ( Arthanat et al., 2024; Gathercole et al., 2021; Howard et al., 2021).

Study design

Nine studies ( Bergschöld et al., 2020; Berridge et al., 2019; Bults et al., 2024; Gibson et al., 2015; Jentoft et al., 2014; Karlsen et al., 2019; Lariviere et al., 2021; Malmgren Fänge et al., 2020; 27) used qualitative research methods, six studies ( Arthanat et al., 2024; Ault et al., 2020; Cao et al., 2022; Nauha et al., 2018; Øderud et al., 2015; Turner & Berridge, 2023) used a mixed-method design. Six publications ( Gathercole et al., 2021; Gedde et al., 2021; Howard et al., 2021; Puaschitz et al., 2021; Puaschitz et al., 2023; Rossetto et al., 2023) reported a randomised controlled trial (RCT), with both the wider ATTILA study and the LIVE@Home.Path trial being RCTs. Quantitative designs were reported in additional studies, with two using a quantitative sample study design ( Huygens et al., 2021; Warner & Tipping, 2022), and one a secondary analysis of quantitative data ( Curnow et al., 2021). Two studies described a non-randomised experiment design ( Kim & Jung, 2023; Lau et al., 2019). The other studies described cross-sectional ( Borg et al., 2022), crossover sequence ( Megges et al., 2018), multi-centre field study ( König et al., 2022) designs, and one implementation study ( Kerssens et al., 2015).

Theory or theoretical models and frameworks

Nine publications ( Arthanat et al., 2024; Bergschöld et al., 2020; Berridge et al., 2019; Bults et al., 2024; Jentoft et al., 2014; Karlsen et al., 2019; Lariviere et al., 2021; Megges et al., 2018; Turner & Berridge, 2023) reported the use theoretical models and frameworks to inform and design their studies or sub-studies or to inform the analysis of data and findings. Three studies ( Arthanat et al., 2024; Bults et al., 2024; Turner & Berridge, 2023) used theories pertaining to technology acceptance models (TAMs), such as the Unified Theory of Acceptance and Use of Technology (UTAUT), another naming the Healthcare technology acceptance model (H-TAM). Another study ( Megges et al., 2018) applied the User Experience (UX) concept, which overlaps in terms of its key aspects with other TAMs. Another two studies considered technology-focussed frameworks, such as the Human Activity Assistive Technology (HAAT) model ( Arthanat et al., 2024) and the concept of innosumption ( Bergschöld et al., 2020), which considers tools and practices. The other four studies described using hermeneutic phenomenology ( Karlsen et al., 2019), acculturation ( Berridge et al., 2019), social-situated learning theory ( Jentoft et al., 2014), and another study was informed by practice theory ( Lariviere et al., 2021).

Population

Sixteen publications ( Arthanat et al., 2024; Ault et al., 2020; Curnow et al., 2021; Gibson et al., 2015; Jentoft et al., 2014; Kerssens et al., 2015; Kim & Jung, 2023; Lariviere et al., 2021; Malmgren Fänge et al., 2020; Megges et al., 2018; Nauha et al., 2018; Puaschitz et al., 2021; Puaschitz et al., 2023; Turner & Berridge, 2023; Warner & Tipping, 2022) included older people with dementia, Alzheimer’s or other cognitive impairments and their careers, including dyads, some of which had to live together to participate. Another eight publications ( Bults et al., 2024; Gathercole et al., 2021; Gedde et al., 2021; Howard et al., 2021; König et al., 2022; Lau et al., 2019; Øderud et al., 2015; Rossetto et al., 2023) included older people with dementia, Alzheimer’s, or other cognitive impairments solely. And six studies ( Bergschöld et al., 2020; Berridge et al., 2019; Borg et al., 2022; Cao et al., 2022; Huygens et al., 2021; Karlsen et al., 2019) comprised the general older population.

Settings

Settings for the use or testing of technologies varied, but in the majority of 19 publications or 17 studies ( Arthanat et al., 2024; Ault et al., 2020; Cao et al., 2022; Curnow et al., 2021; Gedde et al., 2021; Gibson et al., 2015; Jentoft et al., 2014; Karlsen et al., 2019; Kerssens et al., 2015; Kim & Jung, 2023; König et al., 2022; Lariviere et al., 2021; Lau et al., 2019; Megges et al., 2018; Puaschitz et al., 2023; Rose et al., 2018; Rossetto et al., 2023; Turner & Berridge, 2023; Warner & Tipping, 2022) it was the older person’s own home, which could be a house or flat. Five publications, including the ATTILA and the LIVE@Home.Path trials, settings in addition to the home environment were considered, including supported and assisted living accommodation ( Gathercole et al., 2021; Howard et al., 2021; Nauha et al., 2018) and inclusion of participants living in residential care homes ( Øderud et al., 2015; Puaschitz et al., 2021). The other three studies were based solely on wider experiences connected to living in one’s own home ( Borg et al., 2022; Bults et al., 2024; Huygens et al., 2021), Berridge et al. (2019) and Malmgren Fänge et al.’s (2020) studies focused solely on assisted living accommodation and facilities. Bergschöld et al.’s (2020) study focused on technology showrooms, where assistive technologies were demonstrated in older people.

Methods and measures used to explore experiences and assess effectiveness

There were several categories of methods and measures to assess effectiveness: quantitative data from older people ( Curnow et al., 2021; Gathercole et al., 2021; Howard et al., 2021; Kerssens et al., 2015; Kim & Jung, 2023; König et al., 2022; Lau et al., 2019; Megges et al., 2018; Nauha et al., 2018; Rossetto et al., 2023; Warner & Tipping, 2022), quantitative data from caregivers ( Ault et al., 2020; Gathercole et al., 2021; Howard et al., 2021; Kerssens et al., 2015; König et al., 2022; Lau et al., 2019; Megges et al., 2018; Nauha et al., 2018; Warner & Tipping, 2022), qualitative data from older people and caregivers ( Arthanat et al., 2024; Bergschöld et al., 2020; Berridge et al., 2019; Bults et al., 2024; Cao et al., 2022; Gibson et al., 2015; Jentoft et al., 2014; Karlsen et al., 2019; Kerssens et al., 2015; König et al., 2022; Lariviere et al., 2021; Malmgren Fänge et al., 2020; Øderud et al., 2015; Rose et al., 2018; Turner & Berridge, 2023; Warner & Tipping, 2022). There were also data collected from distinct groups of professionals, such as Huygens et al. (2021), but these were not reported directly in this review. Four studies ( Ault et al., 2020; Megges et al., 2018; Nauha et al., 2018; 27) used data collected using the technologies themselves, such as frequency of use. National survey data were used in two studies ( Borg et al., 2022; Huygens et al., 2021).

Studies that investigated participant experiences collected demographic data. Several quantitative studies have used one or several validated measures or scales to assess the effectiveness of the use of technologies. We only reported those measures that were used in at least two different studies, with alternatives and sometimes study-developed measures used in others. Examples of validated quantitative measures administered to older people (or in some cases to caregivers as proxies), with some using translated versions, included: to assess cognitive function and status, the Mini Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA); quality of life was measured with the EuroQol EQ-5D-5L, while autonomy of daily living and ability to complete activities of daily living were measured with different instruments, such as the Modified Barthel Index. All studies that included unpaid carers measured aspects of caregiver burden, with one example of a validated measure repeatedly administered to caregivers being the Zarit Burden Interview (ZBI). Measures focusing on engaging with technologies were collected from 11 studies. The validated scales included the System Usability Scale (SUS), which provides efficiency output measures. Several studies have used various satisfaction measures. As will be reported in the limitations section, there was a lack of measuring experience of using technologies prior to the studies.

In terms of qualitative approaches, 15 studies ( Arthanat et al., 2024; Bergschöld et al., 2020; Berridge et al., 2019; Bults et al., 2024; Cao et al., 2022; Gibson et al., 2015; Jentoft et al., 2014; Karlsen et al., 2019; Kerssens et al., 2015; König et al., 2022; Malmgren Fänge et al., 2020; Øderud et al., 2015; Rose et al., 2018; Turner & Berridge, 2023; Warner & Tipping, 2022) used interviews with older adults, caregivers, or both, with most having both pre- and post-intervention elements. Focus group discussions were conducted in two studies ( König et al., 2022; Øderud et al., 2015), and observations in older people’s homes and home visits in four ( Arthanat et al., 2024; Jentoft et al., 2014; Lariviere et al., 2021; Øderud et al., 2015). Two studies ( Øderud et al., 2015; 27) also used qualitative surveys and two ( König et al., 2022; Nauha et al., 2018) participant diary logs.

Cost-effectiveness and benefits were analysed in two studies: the ATTILA study ( Gathercole et al., 2021; Howard et al., 2021) and to some degree in the study by Arthanat et al. (2024). While the ATTILA study used quality-adjusted life-year (QALY) measures and the EuroQol-5 Dimensions (EQ-5D) index, Arthanat et al. (2024) developed ‘guiding questions’ based on the Unified Theory for Acceptance and Use of Technology (UTAUT) to ask about perceived cost and benefits in interviews with participants.

Types of technologies deployed

There are different ways to summarize and categorize care technologies, whereby there is often an overlap between different categories given the potential complexity of systems and the rapidness of innovation ( Doughty et al., 2007; Greenhalgh et al., 2018). Curnow et al. (2021) provides a list of possible categories of assistive technology installed in homes of people with dementia; the categories and grouped devices and equipment are similar to those used in this review, but also include a category ‘basic AT’ which includes pendant alarms alongside unmonitored sensors, which we have grouped under ‘Emergency response’ given that they require assistance from additional parties such as unpaid carers or response services. An additional categorisation by specific target participant or user group, aside from the general population of older people, can be derived by looking at the participant population categories, which indicates that the majority of technologies examined targeted people with dementia or other memory disorders, another target group were family and other unpaid carers.

Considering technology types, 15 studies ( Arthanat et al., 2024; Ault et al., 2020; Berridge et al., 2019; Bults et al., 2024; Cao et al., 2022; Curnow et al., 2021; Gathercole et al., 2021; Howard et al., 2021; Kim & Jung, 2023; König et al., 2022; Lariviere et al., 2021; Lau et al., 2019; Malmgren Fänge et al., 2020; Nauha et al., 2018; 27) involved systems or packages of interconnected devices and active or passive sensors, reminders, and trackers. The digital platform was integrated into five studies ( Berridge et al., 2019; Bults et al., 2024; Huygens et al., 2021; Kerssens et al., 2015; Rossetto et al., 2023). Four studies focused on individual products or devices: GPS trackers were the focus of the studies by Megges et al. (2018) and Øderud et al. (2015); social/personal alarms in Puaschitz et al.’s (2021) publication; and a basic remote was assessed for its use and effectiveness by Jentoft et al. (2014). The other nine studies ( Bergschöld et al., 2020; Borg et al., 2022; Gedde et al., 2021; Gibson et al., 2015; Huygens et al., 2021; Karlsen et al., 2019; Puaschitz et al., 2021; Turner & Berridge, 2023; Warner & Tipping, 2022) explored general use, whereby participants have used or were planning on using different technologies which in itself were not compared; devices here were mainly personal alarms, reminders, monitoring sensors and GPS trackers, and digital platforms.

Rather than investigating the hardware and/or software systems or products themselves, five publications ( Bergschöld et al., 2020; Huygens et al., 2021; Lau et al., 2019; Turner & Berridge, 2023; Warner & Tipping, 2022) focused on technology education. This includes Turner and Berridge’s (2023) publication, which introduced ‘let’s Talk Tech, described as a way to facilitate discussion among people with mild dementia and care partners around four kinds of technologies: location tracking, in-home sensors, web cameras, and virtual companion robots, leading to a structured technology use plan that aligns with the older person's needs, values, and concerns. In addition, Rossetto et al. (2023) explored a rehabilitation programme and Lau et al. (2019) involved signposting additional services as part of social care.

Findings on overall experiences and effectiveness

Several studies (e.g. Bergschöld et al., 2020; Gibson et al., 2015; Jentoft et al., 2014; Nauha et al., 2018) found that simple devices are often more useful and more aligned with users’ and caregivers’ everyday needs than complex systems, which are viewed as ‘cumbersome’ ( Bergschöld et al., 2020) and carers’ especially seek simplification ( Bergschöld et al., 2020; Jentoft et al., 2014). Similarly, Arthanat et al. (2024), Jentoft et al. (2014), and König et al. (2022) found that traditional methods are often preferred, as they feel more familiar, intuitive, and secure, potentially limiting their willingness to adopt new technologies, regardless of efficiency. König et al. (2022) also found that the more technologies used, the more positive they were perceived. As Cao et al. (2022) indicated, adoption and acceptance rely on both intrinsic and extrinsic factors. Two studies ( Gibson et al., 2015; Lariviere et al., 2021) identified perceived technologies as a threat to replace personal care, with participants in Gibson et al. (2015) calling technologies a ‘necessary evil.’

Findings on design and technical aspects

Connected to potential scale-up and sustainability, design and technical aspects were discussed in seven studies ( Arthanat et al., 2024; Ault et al., 2020; Bults et al., 2024; Jentoft et al., 2014; König et al., 2022; Megges et al., 2018; Nauha et al., 2018), with providing a language choice ( Bults et al., 2024) and appropriate sized devices ( Bults et al., 2024) being some identified positives, while fragility of products ( Ault et al., 2020), poor sound quality ( Nauha et al., 2018), software not working ( Bults et al., 2024), connectivity issues and limitations ( Bults et al., 2024; Megges et al., 2018; Nauha et al., 2018), and missing interoperability with external devices, e.g. own mobile phone ( Bults et al., 2024) being found as factors that needed more development. König et al. (2022) found that poor design could add to the stigmatization associated with care devices.

Use and acceptance among older people, including those with cognitive impairments

This section explores the reasons why older people, including those with cognitive impairments, accepted care technologies and factors influencing technology acceptance.

Two studies ( Bergschöld et al., 2020; Gathercole et al., 2021) reported a limited fit between care needs identified in assessments, technological recommendations, and installation. Gathercole et al. (2021), for an English context, found that 62% of provided devices were not identified during assessment and 53% of recommended devices not being installed.

Acceptance and ongoing use was associated with usefulness and achieving intended outcomes ( Arthanat et al., 2024; Bergschöld et al., 2020; Karlsen et al., 2019; Kim & Jung, 2023; Lau et al., 2019; Warner & Tipping, 2022), for example increased independence and living at home for longer ( Huygens et al., 2021; Karlsen et al., 2019; Malmgren Fänge et al., 2020; Øderud et al., 2015), improved quality of life ( Huygens et al., 2021; Øderud et al., 2015), perceptions of increased safety ( Cao et al., 2022; Karlsen et al., 2019), or by preventing potentially harmful incidences such as wandering in people with dementia ( Lau et al., 2019; Øderud et al., 2015), and leisure activities ( Bults et al., 2024; Jentoft et al., 2014; Kerssens et al., 2015; Øderud et al., 2015), though many people with dementia needed assistance in this context. Huygens et al. (2021) found that it was sometimes difficult for professionals to estimate whether an older person would accept technologies or not, which points towards the importance of thorough initial assessments and reviews ( Bergschöld et al., 2020; 05). However, Bergschöld et al. (2020) also found that access to formal provision of care technologies could be a complex bureaucratic process, and people with dementia need more professional support to successfully participate in the selection of assistive devices ( Borg et al., 2022).

Several studies have indicated that care technologies are perceived as most useful if it is possible to integrate them into everyday practices, ensuring that use becomes habitual ( Arthanat et al., 2024; Bergschöld et al., 2020; Jentoft et al., 2014; Øderud et al., 2015). Some studies ( Jentoft et al., 2014; Øderud et al., 2015) have found that the introduction of devices and systems at the onset of disease aids learning and integration in daily life. Kerssens et al. (2015), König et al. (2022), and Nauha et al. (2018) found that when devices did not achieve their intended outcomes, they were not considered useful.

When compared with older people without dementia, Borg et al. (2022) found in their nationwide study on technology provision by municipalities that older people with dementia used fewer assistive devices and products and benefitted less from them than older people without cognitive impairments. Gibson et al. (2015) found that users with dementia found at least some devices to be intrusive and irritating, Kerssens et al. (2015) reported that in their study people with dementia were not irritated, and Malmgren Fänge et al. (2020) found that most participants with dementia were not aware of the technologies.

Non-adoption and abandonment

Barriers to adoption and reasons for abandonment among older people include deterioration of cognitive impairment ( Øderud et al., 2015) or physical limitations ( Kerssens et al., 2015) that make continuous use of technologies unviable. Concerns around affordability include the cost of initially buying and installing devices and ongoing costs such as fees or increased electricity use were found in three studies ( Bults et al., 2024; Cao et al., 2022; Gibson et al., 2015). Lack of trusting devices ( Berridge et al., 2019; Cao et al., 2022) and ethical concerns about lack of privacy and monitoring have been raised in some studies ( Bults et al., 2024; Cao et al., 2022; Malmgren Fänge et al., 2020), whereas Øderud et al. (2015) found that this was not an issue for the majority of participants using GPS trackers in their study. Puaschitz et al. (2023) reported that about half of the participants with dementia in their study around personal alarms were unaware of having social alarms installed, explaining their non-usage. They also found that compared to their caregivers, people with dementia were more likely to say that personal alarms gave them a false sense of security (28% vs. 9.9%), potentially due to their former experience of falls and increased anxiety.

Experiences and outcomes for family carers and unpaid caregivers

One of the main aims of providing and using technology is to decrease caregiver stress and/or caregiver burden. Overall, the effectiveness of technologies for carers depends on several factors, including knowledge, timing, ongoing professional support, and the quality and usability of the technology itself. The studies examined in this review have shown some positive impacts of technologies on unpaid carers, with others finding no differences and some reporting notable challenges. As with older people, studies found that simplicity and technology that achieved the intended use and outcomes were positive ( Arthanat et al., 2024; Lau et al., 2019; Nauha et al., 2018; Øderud et al., 2015; Rossetto et al., 2023; Warner & Tipping, 2022), although Lau et al. (2019) reported that the intended outcomes had no lasting effect at one year. Several studies ( Gibson et al., 2015; Kerssens et al., 2015; Lau et al., 2019; Rossetto et al., 2023; Warner & Tipping, 2022) have indicated that technologies can decrease caregiver stress and provide emotional relief by offering peace of mind. Karlsen et al. (2019) and Malmgren Fänge et al. (2020) identified emotional ambivalence among carers, who appreciated that support devices and systems could offer in enabling older people to stay living at their own homes, but also experienced stress from responsibilities connected to using care technologies.

König et al. (2022), Gathercole et al. (2021), and Kerssens et al. (2015) found no significant decrease in caregiver burden, and Bergschöld et al.’s (2020) study reported an increased burden due to the requirement of caregivers to function as enablers. Technical issues such as delayed alarms and reliability problems have also been reported to have a negative impact on caregivers ( Malmgren Fänge et al., 2020). Many caregivers struggle with a lack of formal support ( Gibson et al., 2015; Karlsen et al., 2019). Lariviere et al. (2021) found that, while most carers were able to adapt care technology to their routines, success varied. Bults et al. (2024) and Jentoft et al. (2014) reported decreases in caregiver burden but only after an initial increase during adoption and implementation. Warner and Tipping (2022) further noted that caregivers experienced improved self-assurance, greater knowledge in managing challenging behaviours, and a reduced need for constant attention to daily tasks. Technologies have also been reported to reduce feelings of being housebound among caregivers ( Karlsen et al., 2019; Øderud et al., 2015).

Intervention effectiveness and treatment effects

Several studies have investigated the effectiveness of interventions for people with dementia and other cognitive impairments with mixed findings. Lau et al. (2019) found improvements in reducing wandering, and Rose et al. (2018) reported that sensors were useful for determining agitation, sleep, and incontinence. Rossetto et al. (2023) reported that the ABILITY intervention led to improvements in global cognition, particularly in language, executive function, and memory, with some evidence of long-term effects. Kim and Jung (2023) found statistically significant positive effects on cognitive function in individuals with mild cognitive impairment, but not in those with dementia, and improvements in criteria such as quality of life, daily living, mood (depression and anxiety), and satisfaction in both disease groups. However, other study outcomes were less positive. Kim and Jung (2023) noted no significant effects on instrumental activities of daily living (IADLs). More critically, publications on the ATTILA study ( Gathercole et al., 2021; Howard et al., 2021) found that comprehensive assistive technology and telecare packages did not yield improvements in the time spent living at home, safety, or quality-adjusted life years. In some cases, participants receiving a full ATT package reported poorer quality of life than those receiving only basic support.

Cost-effectiveness

An economic analysis of assistive technology and telecare interventions for people with dementia as part of the ATTILA study found no significant cost-related benefits. Gathercole et al. (2021) reported that after 104 weeks, there were no significant differences between the intervention and control groups in terms of health and social care resource use or overall societal costs. Howard et al. (2021) reported no evidence of cost-effectiveness when assessing days lived in the community, health-related quality of life, or quality-adjusted life years (QALYs) based on proxy-reported EQ-5D.

Turner and Berridge (2023) found that adoption, continuous use, or discontinuation of one intervention or technology does not predict using or not using others, and Puaschitz et al. (2023) discussed that access to and acceptance of technology is not the same as actual successful and sustained usage. For digital technologies to work effectively and be integrated successfully into social care for older people, particularly those with cognitive impairments such as dementia, and their family or unpaid carers, the publications discussed key preconditions and provided useful recommendations.

Technologies need to be user-friendly, intuitive, and simple to use, especially if they are to be used by people with cognitive impairments ( Jentoft et al., 2014; Megges et al., 2018; Nauha et al., 2018). This does not exclude multidimensional solutions, but they need to be easily approachable or well supported by professionals ( Lau et al., 2019), and they should preferably allow for some form of control. Technologies need to be adaptable and tailored to changing and evolving individual needs, and their supply should be person-centred rather than technology- or provision-led. Social care providers must also ensure that devices and systems work in real-world settings ( Gathercole et al., 2021; Gibson et al., 2015; Nauha et al., 2018).

Assessments as part of care technology provision must be comprehensive so that the most suitable technologies are implemented, with regular follow-up and review to adapt technology provision when there are changes in needs and abilities or better solutions become available ( Borg et al., 2022; Cao et al., 2022; Curnow et al., 2021; Huygens et al., 2021; Øderud et al., 2015; Puaschitz et al., 2023). Sometimes, this can be the least intrusive intervention rather than a full package of devices. Assessments also need to consider unpaid carers’ abilities, capacity, and the relationship with the person drawing on care. It is important that assessors are sufficiently trained and experienced to make appropriate decisions on the provision of devices and services so that technologies are used in an optimized and sustainable manner ( Curnow et al., 2021; Gibson et al., 2015; Huygens et al., 2021; Øderud et al., 2015).

Carers are essential in creating technologies for older people. Although devices and/or applications can reduce caregiver stress and burden, they often add new tasks to caregivers’ responsibilities. Comprehensive information, support, and training on devices and systems are needed to ensure that technologies do not become a burden on themselves. Several studies have found that technologies do not necessarily reduce the carer burden (e.g. Bults et al., 2024; Karlsen et al., 2019; Kerssens et al., 2015; Megges et al., 2018). Kerssens et al. (2015) noted that expectation management might be needed to avoid or reduce carers’ abandonment because of unfulfilled expectations. The importance of involving caregivers in the assessment process has been mentioned above.

Ethical concerns, particularly regarding privacy, data protection, autonomy, and personal dignity, should be addressed during design, development, and provision ( Malmgren Fänge et al., 2020; Megges et al., 2018; Øderud et al., 2015). In addition to ethical challenges, cultural sensitivities and concerns need to be considered to ensure that cultural values and expectations around caregiving are addressed. Berridge et al. (2019) highlight the pivotal role that professionals (social workers) can play as ‘cultural brokers’ helping to overcome challenges related to language, trust, and access to technologies.

Early introduction of technologies in the care process enhances effectiveness, as their use can become familiar and habitual, particularly for people with a diagnosis of dementia and other cognitive impairments, as their capabilities might decrease ( Arthanat et al., 2024; Jentoft et al., 2014; Malmgren Fänge et al., 2020; Øderud et al., 2015). Additionally, a general positive attitude towards technology is vital for the adoption of devices and systems in everyday life, regardless of age ( Bults et al., 2024; Jentoft et al., 2014; König et al., 2022). Prior positive experiences using technologies can help, whereas negative experiences may create challenges ( Cao et al., 2022).

Costs associated with implementation and ongoing costs can be barriers to adoption and continued use ( Bults et al., 2024; Cao et al., 2022; Gibson et al., 2015), public sector provision and involvement can help here. Do-it-yourself (DIY) solutions can be cost-effective alternatives to specialized care technologies and high-tech systems, as Bergschöld et al. (2020) reported. They are often perceived as easier to adopt and more pragmatic to use and are available without the additional task of receiving them through formal social care providers. Bergschöld et al. (2020) advocated that greater attention should be paid to low-tech solutions.

Information and ongoing training and support are essential for the adoption and continuous, successful use of technologies. Older people have greater interest and understanding than is sometimes expected. However, the lack of digital literacy among older people and their caregivers was seen as a vital barrier to interest in and access to novel digital devices and applications, alongside traditional devices such as personal alarms, as Puaschitz et al. (2021) found. This can be overcome with better information and updates on technological developments ( Kerssens et al., 2015; Puaschitz et al., 2021).

Malmgren Fänge et al. (2020), Øderud et al. (2015), and Warner and Tipping (2022) stressed that technologies should complement and be integrated with other social care provisions in place. It should not be seen or used to replace personal care without comprehensive assessment or regular review. Integration with other available community care provisions also relieves family carers and allows them to ‘maintain the role of a relative’ rather than a constant caregiver, as Malmgren Fänge et al. (2020 p. 654) noted. When complex multi-device and-dimensional systems and programs are considered, they need support from professionals who understand both the technologies and personal needs. Reliable access to professionals and collaboration between carers and professionals allows for the successful adoption and adjustment of technologies ( Jentoft et al., 2014; Øderud et al., 2015). A trustful relationship with care professionals and services can have an impact on initial openness to technology adoption ( Bults et al., 2024; Cao et al., 2022). Karlsen et al. (2019) pointed out that engaging with telecare services is often an additional responsibility for unpaid carers. Finally, technologies need to be designed in collaboration with professional technical designers and engineers and older people, including those with cognitive disabilities, unpaid carers, community services, and formal social care providers ( Cao et al., 2022; Gathercole et al., 2021). This should also include representatives from diverse ethnic and cultural backgrounds as recommended by Bults et al. (2024).

WP3 explored the current processes and practices involved in localized TECS at all three study sites, with a particular focus on emerging or established service delivery models supporting post hospital discharge and reablement services for older adults in the project’s three regional sites (Southwest – Bristol, North Somerset, and South Gloucestershire, North East – Middlesbrough, and Central – Oxfordshire).

Digitally enabled, post hospital discharge and reablement services (PHDRS) were selected as a priority to focus on for this work package, as identified through previous consultations with councils in Middlesbrough and South Gloucestershire and the Bristol, North Somerset, and South Gloucestershire Integrated Care Board (BNSSG ICB). At the point of application for our NIHR Health Technology Assessment Accelerator Award, the Middlesbrough Council and BNSSG ICB had also begun local pilot projects to assess the potential cost savings and effectiveness of PHDRS to reduce the length of hospital stay for older patients. As such, the PHDRS evolved into a priority service within participating fieldwork sites, justifying their inclusion as a series of focal interventions for process evaluation.

To fulfil this work package’s aim, the team attempted to answer the following main research question:

What are the current processes and practices for digitally enabled, post hospital discharge and reablement services (PHDRS)?

To generate the granular data required to answer this research question, the following sub questions were developed:

The work package has identified two service models and the rationales for their use that are currently used to offer PHDRS in England. The distinction between these service models reflects concerns about key outcomes (timely discharge from hospital), ownership of services and local revenue generation, and perceptions of equity in TEC provision interpreted in individual organizations and across regional partnerships. The full range of partners involved across the PHDRS nexus, including local authorities, NHS trusts, integrated care boards, and technology suppliers identified in DIALOGUE, requires more granular analysis that can ‘follow’ the processes of provision within and across these actor-sites as part of a larger evaluation of the effectiveness of PHDRS.

WP4 used a combination of an e-Delphi study and accompanying consensus workshops to collate complex information from a group of experts in social care ( Black et al., 1999; Hsu & Sandford, 2007). In this case, we asked local authority (LA) employees across the three fieldwork sites who worked with technology-enabled care for older people to participate in the e-Delphi exercise. The aim was to reach a consensus on the facilitators of and barriers to TEC adoption. As the LAs involved were geographically diverse (e.g., Middlesborough, North Somerset, Oxfordshire, and South Gloucestershire), we adopted an e-Delphi process ( Donohoe et al., 2012; Toronto, 2017). The survey instruments and consensus workshops were conducted online using an online survey platform (JISC Online Surveys) and a digital collaboration platform (Microsoft Teams).

Similar to previous e-Delphi studies on social care for older people, we planned three rounds of surveys to reach a consensus ( Shepherd et al., 2017). However, after two rounds, we established a consensus at the 80% agreement level, which is the most widely used measure in e-Delphi studies ( Diamond et al., 2014). Therefore, we stopped after two rounds of surveys and held a consensus workshop in the third round to gather further qualitative insights. In survey two, we selected four themes to explore in more depth (e.g., Q1, Q3, Q4, and Q7) ( Figure 2). Respondents were asked to provide agreement on the results of survey one, rank answers, and qualitatively comment on how TEC alters the fundamentals of care work, why people with dementia benefit the most and why family caregivers benefit the least from TEC, how to measure TEC priorities, and how to reduce barriers to TEC. Findings from surveys one and two were then presented at an online consensus workshop with TEC leads and staff involved in TECS delivery within the three fieldwork sites.

Fifteen employees from four LAs participated in the e-Delphi exercise. Table 3 shows the characteristics of these respondents, including their age, gender, job role, and survey completion. There were 12 responses to the first survey and eight responses to the second survey, with five respondents who completed both surveys. There were three new respondents for the second survey, but they represented LAs from the previous round and provided insights into TEC provision for their LA. There were four males, and 11 females aged between 27 and 56 years, with an average age of 41. There were six from Middlesbrough Council and Oxfordshire County Council, and from within BNSSG ICB, two were from North Somerset Council, and one from South Gloucestershire Council. Job roles include social care, technology and innovation, service providers, managers, and occupational therapists.

The findings from the e-Delphi exercise highlight the acceptance of traditional TECS by LAs, including telecare and fall prevention, mostly benefitting older people aged 65–84 years and those living with dementia. Family caregivers benefit the least (mostly indirectly) because TEC are not generally designed for them and are difficult to access. These technologies are used in services such as reablement, hospital-to-home monitoring, and activity monitoring. While all LAs agree on what TEC is currently offered, newer technologies are sometimes provided on a smaller scale but face challenges in terms of consistent use, integration, and scale-up. TEC is used to promote independence, enhance the quality of life of older people, delay further care needs, and improve care efficiency. However, barriers still emerge, including gaps in the digital skills of LA workforce, limited user awareness and support, and resource constraints.

These findings indicate several key recommendations for TEC to move forward. First, social care services should adopt a more inclusive and person-centred approach by expanding TEC offerings to benefit all groups, including those with sensory impairments. In doing so, family caregivers should be involved in the design and decision-making process to ensure that users and caregivers fully benefit from the technology. LAs should invest in support staff who specialize in installation and maintenance, provide accessible resources, and ensure that technical issues do not fall solely on service users. Second, LAs should upskill their employees by including digital training in inductions, giving them hands-on experiences, and ensuring continuous support. Third, raising awareness among older people can be achieved by communicating through trusted channels such as GPs, third-sector organizations, and community engagement. Finally, to scale up services and ensure cost efficiency, there should be more academic research into the effectiveness of TEC and newer, more cost-effective iterations of TEC.