Background

Dementia is a complex acquired brain condition characterised by a decline from a previous level of cognitive functioning with impairment in cognitive domains [1]. Worldwide there are an estimated 50 million people who have dementia and nearly 10 million new cases every year [2]. Informal carers (families, friends and neighbours) provide the majority of care for persons with dementia [3]. Dementia can be overwhelming for carers of persons with dementia and can cause stress from physical, emotional and economic pressures [4,5,6]. Stretched health and care resources necessitate alternative and innovative ways to providing care for persons living with dementia [7]. Assistive technology (AT) has been suggested as a means to support someone who has dementia and their carers to stay independent and remain in the community [8,9,10,11]. AT can be defined as: “any item, piece of equipment, product or system that is used to increase, maintain or improve the functional capabilities and independence of people with cognitive, physical or communication difficulties” [12]. The use of AT by persons living with dementia may by extension also benefit the carer, as it could offer the potential to increase the support to carers and alleviate some of the burden of caregiving [13,14,15,16]. AT may assist carers to address the increased level of responsibility whilst caring for a person with dementia [17, 18]. Additionally, carers of someone who has dementia are in the unique position of using their pre-conceived ideas regarding AT to suggest and decide on the access to and use of AT [19], yet very little is known about carers’ experiences of AT use.

Why it is important to do this review

Currently, AT and Artificial Intelligence driven healthcare solutions are being viewed as a panacea for reducing carer burden [20, 21] and multiple studies are investigating how AT can support people with dementia [22,23,24]. Carers could be using the AT together with the person with dementia (such as safety alarms) and/or carers could be looking after someone who has dementia, who uses AT independently of the carer. Yet, little is known about the experiences of carers using AT and what impact AT has on carer health and wellbeing [25]. This review, aims to fill the gap in literature that so far has predominantly looked at AT from the perspective of people living with dementia and its use within institutional settings [7, 26, 27] as well as identifying carer wellbeing when using AT. This information could benefit carers and persons with dementia considering AT solutions for use at home, help healthcare professionals who prescribe and set up AT solutions, as well as developers/inventors of AT.

Review aim and questions

This review aims to:

  1. 1)

    Identify the types and uses of AT in dementia;

  2. 2)

    Describe the effectiveness of AT for outcomes (including burden, well-being and quality of life) of carers of people with dementia living at home;

  3. 3)

    Describe carers’ experiences of AT use in dementia;

  4. 4)

    Determine the aspects of AT that are valued and work well for carers by integrating (2) and (3) as above.

Methods

The review protocol was registered with the international prospective register of systematic reviews PROSPERO (CRD42017082268). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist is included as Additional file 4.

Types of studies

Quantitative, qualitative and mixed method study designs were included. Letters to the editor, abstract and conference proceedings, book reviews, study protocols and theses/dissertations were excluded. We did not include other reviews but checked references within identified existing reviews on dementia, informal carers and AT to ensure that all relevant studies had been located. Due to funding constraints, only studies in English language or those translated to English language were included.

Evaluation of effectiveness

We included all randomised and controlled trials that compared AT for carers of someone who has dementia to those not provided with the AT, and who received usual care. We also included observational and cohort studies.

Evaluation of experience

We included studies that used qualitative methods of data collection and analysis, either as a stand-alone qualitative study or as part of a mixed-method study.

Types of participants

Studies that included carers who provide unpaid care for a person living with dementia at home were included. Providing care is defined for the purposes of this study as ‘supporting a person with dementia physically, emotionally, financially or socially’ and care could be provided by a relative, a friend or a neighbour. There were no restrictions regarding gender, living arrangements or ethnic background. Studies reporting on carers who provide support to a person living with dementia receiving care in hospital and/or long-term institutions and carers younger than 18 years and formal/paid carers were excluded.

Types of assistive technology

For this review, studies that evaluate AT use in dementia involving carers were included. AT was defined as ‘any advanced electronic equipment, which can be used to enhance support and care, act as a prompt for intervention by carers, monitor welfare and assist in communication and leisure activities for a person with dementia’. This AT can be standalone (e.g. Tablet computers) or be part of an integrated system (e.g. GPS and sensor trackers) and can be stationary or mobile. As the focus of most research studies invariably is on the person living with dementia, any study that reported on effects or experiences of AT use on carers were included. Studies that reported only on AT use for people with dementia without including carers were excluded, as were studies that focus only on electronic therapeutic interventions that are not AT (e.g. computer-based education or support for carers).

Types of study outcome measures

The search was not limited to specific types of outcome measures and included carer self-reported outcome measures of burden; quality of life; and well-being; and self-reported or researcher observed experiences of usefulness; benefits and disadvantages of AT and impact on carer /person living with dementia relationship.

Search strategy

The search strategy was developed in collaboration with a Bodleian medical library librarian at the University of Oxford.

Searches were carried out on:

Databases

Including MEDLINE (Ovid) from 1946 to June 2018; EMBASE from 1974 to June 2018; PsycINFO from 1806 to June 2018; AMED 1985 to June 2018; CINAHL from 1981 to June 2018; Database of Abstracts of Reviews of Effects (DARE), OT seeker and The Cochrane Library of Systematic Reviews. The search included studies within ALOIS (from inception to June 2018).

Unpublished literature

The International Standard Randomised Controlled Trials Number (ISRCTN) registry [28] and the National Institutes of Health Clinical Trials Database [29] were searched for information on unpublished ongoing trials. Searches within these databases were used to identify additional studies and authors to contact for full text reports.

Manual searches

We also conducted manual searches of reference lists to identify relevant research studies.

Details of the full search, with search strategies and the number of records identified in each database are included in Additional file 1.

Screening

Electronic search results were downloaded into Covidence software [30] (an online digital platform that streamlines the production of systematic reviews and allows screening and data extraction between collaborating reviewers) as .ris files. Duplicates were removed using the software. Authors VS and MP independently screened all titles and abstracts for eligibility against the inclusion/exclusion criteria. For studies that had insufficient information from the title and abstract, full text articles were retrieved to determine inclusion. Studies marked for possible inclusion underwent a full-text review. At full-text review, when both VS and MP agreed that a study did not meet the full eligibility criteria, the study was excluded. CJ was consulted when VS and MP did not agree on a study. Discrepancies were resolved by mutual discussion.

Data extraction

A bespoke data extraction form (Additional file 2) developed by all the authors was used and initially piloted on a sample of studies to refine the form. Data from the studies were logged using Microsoft Excel 2016. There were no deviations from the published protocol.

Effectiveness

Data extraction items from quantitative studies were based on the recommended items from the Cochrane handbook for systematic reviews of interventions [31]. Information on citation including authors, date of publication, study design, duration, number of participants, participant gender, age, ethnicity, country where the study took place, relationship status to the person living with dementia, types and use of the AT, outcome measures used, time points of data collection, missing participants and key conclusions from the study authors were extracted.

Experience

In addition to collecting information from qualitative studies on citation, author details, study design, duration, and participant information, country and time points when information was collected, VS extracted data based on study authors’ commentaries and conclusions [32, 33]. MP and CJ checked extracted data for accuracy and completeness. Disagreements and clarifications were resolved by discussion among the authors.

Results

The first search was carried out in December 2017 and repeated in June 2018. A check for duplicate records was carried out electronically. To confirm results gained from Covidence [30] an additional screening using reference management software Mendeley [34] was undertaken. From the 11,553 records retrieved from database search 3635 were removed as duplicates. The titles and abstracts of the 7918 retrieved records were independently screened by VS and MP. A total of 7746 records were excluded (including further duplicate records) and full-text articles for the remaining 172 records were independently assessed for inclusion based on full texts by VS and MP. Fifty-six papers met the inclusion criteria and were included in this review for data extraction. Reasons for exclusion of the full-text papers were documented and are listed separately (Fig. 1).

Fig. 1
figure 1

PRISMA flow diagram of study selection

Included studies

Of the 56 included publications, 30 were qualitative, 17 quantitative and 9 mixed-methods reporting on a total of 50 studies from 19 countries. There were 2 Randomised Controlled Trials (RCTs) and 1 Controlled Clinical Trial. The publications were from 2000 to 2018, reporting on findings from 2016 carers (660 men and 1165 women, where gender was reported) and 84 types of AT. Carers’ age ranged from 19 to 91 years, with 13 publications not reporting an age range for participants. Several methods were used for data collection including interviews (32), surveys (14), observations (8), focus groups (7), questionnaires (6), diary/log entries (4) and video recording and email and blog reviews (1 each), with 19 studies using more than one method for data collection. Seven studies [35,36,37,38,39,40,41,42] reported on ethnic backgrounds of participants which were predominantly from white backgrounds alongside Hispanic, African American, Asian and ‘other’ backgrounds. Most studies reported the severity and type of dementia, without specifying a classification system; for ease of use, we have followed the 3-stage classification system of clinical dementia rating assessment [43]. Where reported in the studies, most studies involved people with dementia who had mild or moderate dementia. The carer relationship to the person with dementia ranged from spouses, siblings, children, daughter/son-in-law, nieces and nephews, grandchildren, neighbours and friends. Some studies included paid carers and participants who lived in long-term care facilities alongside carers of someone who has dementia living at home. The results described in this review relate only to family carers and persons with dementia who lived at home from those studies.

Quality of included studies: appraisal of included studies

As this review involved quantitative, qualitative and mixed-method studies, the Mixed Methods Appraisal Tool [18, 44, 45] for assessing quality of included studies was used. MMAT scores are provided (Additional file 3) for the included publications. The score is a subjective appraisal of a study’s methodological quality. For qualitative and quantitative studies, the percentage of criteria met is stated. For mixed-methods studies, the overall score cannot exceed the lowest score of a component, so if one part received 100% but the other 50%, the overall score would be 50%. This means it would be possible for a study to have a strong quantitative section and a weaker qualitative section, or vice-versa, but the overall score would be low, suggesting the study might be less valuable [18]. The included studies were critically evaluated by VS and discussed with MP and CJ with discrepancies resolved through discussion. A majority of the included publications scored over 50% on the MMAT score with six of the qualitative studies [46,47,48,49,50,51] scoring highly for due consideration of results in context and for researchers’ own influence on data collection and interpretation of results. The RCTs [35, 36] and the controlled clinical trial [37] scored poorly on allocation concealment and blinding and the RCT pilot trial [36] also had a high attrition rate. While not ignoring the strengths and weaknesses of the studies, we have included all studies, to answer the questions for this review and add to the richness of our findings. Results are presented in line with the questions that this systematic review set out to explore. Characteristics of the included papers are presented in Table 1 and additional details are available in Additional file 3.

Table 1 Characteristics of included studies

Data synthesis

As the included studies were a mixture of quantitative, qualitative and mixed-methods studies, we completed a narrative synthesis of the evidence [32, 33, 93,94,95]. The narrative synthesis aims to present a descriptive summary of findings across the included studies and themes relevant to the aims of this review, such a synthesis can produce new insights and understanding from different aspects and provides a more informed view of carer experience with AT. Lins et al. describe that when “descriptive qualitative studies that are characterised by ‘thin descriptions’ are available, an aggregative method is more appropriate; if the identified evidence includes a high proportion of in-depth qualitative studies characterised by ‘thick descriptions’, an interpretative method can be applied” [96]. Since all of the qualitative studies in this systematic review had ‘thin descriptions’ available we used an aggregative method for qualitative synthesis.

We followed the method of Timulak [97] for qualitative data synthesis. We transferred data from the extracts of the included studies into data analysis software NVivo Version 12 [98]. The first step of the analysis was to read and get familiarised with the included studies. This was followed by creating a conceptual framework of categories on carers’ experiences as an emerging process using a few studies. Subsequent studies were coded into pre-existing concepts, and new categories were created when deemed necessary. As a third step, from these categories, themes of meaningful units are described and presented as findings. VS extracted and summarised the data for the results. MP and CJ reviewed and highlighted outstanding issues and final themes were subsequently arrived at through discussions. As this is an aggregative synthesis, we maintained reflexivity throughout the review process through discussions and reflections on extracted data and coding.

Question 1: identify the types and uses of AT in dementia

To date, there appears to be no agreed way of classifying AT available for use by people with dementia, and we have classified them by their use as part of this review. A list of AT described in the included studies (Table 2) was created with uses of the AT classified based on work developed by AT dementia [99] and Lorentz et al. [72]. From the included studies, AT is a mixture of active (requires action or interaction from the part of the person living with dementia or carer) and passive devices. Some devices had multiple uses e.g. the electronic medication reminders acted both to orient the person with dementia to time when they needed to take their medication as well as acting as a safety device to ensure they took important medication on time. 64 of the AT products described in the studies were commercially available with 10 studies describing AT that were research prototypes.

Table 2 Types and uses of Assistive Technology

The most commonly used AT was for safety and security (n = 38) including tracking devices and home safety devices. Followed by devices used for supporting memory and orientation for the person living with dementia (n = 23) and for social interaction and leisure activities (n = 16). In this review, very few studies (n = 3) considered AT which supported basic Activities of Daily Living activities such as feeding, washing, grooming or dressing. The AT used (including some research prototypes) are adapted from aids/devices that many people, with and without cognitive impairment, already use. None of the AT were for advanced instrumental Activities of Daily Living, such as managing finances, shopping or preparing meals and none of the AT addressed behavioural issues such as aggression or disinhibition, which is quite common in someone who has dementia.

Question 2: describe the effectiveness of AT for carers

The included studies reported on a wide range of carer-oriented measures (Zarit Burden interview, satisfaction with AT, carer well-being score), many of which were created for a specific study. A list of outcome measures used is presented in Additional file 3. Not all included studies reported on the effectiveness of AT for carers and due to the wide range of outcome measures and uses of AT, a descriptive summary of reported changes is provided (Table 3). From the 16 quantitative studies (17 publications), AT were reported as ‘somewhat’ or ‘very useful’ and AT is viewed as an adjunct to caregiving. There were no significant changes in carer reported well-being or burden. Surprisingly none of the studies considered or reported adverse events from AT use. Generally, carers reported they would recommend use of AT to others in similar situations, especially AT that supported safety and security for people with dementia. Where this was specifically asked, carers reported wanting to continue to use the AT, after the trial period. AT devices for safety, including tracking devices were the most used and appreciated by carers.

Table 3 Reported changes in informal carers

Question 3: describe carers' experiences of AT use in dementia

Thematic synthesis from the qualitative data generated 4 themes and 15 sub-themes. Quotations from studies to support themes and sub-themes are listed in Table 4.

Table 4 Sample quotes for Themes and Sub-themes

Positive aspects

All the studies reported that the experience of cares using AT was generally positive.

Relationships

The use of AT for leisure and social interaction, memory support; orientation; safety and security seemed to help strengthen relationships between the person living with dementia and their carers. The AT was perceived as helping the carer function better in their caregiving role and became a ‘member’ of the wider social network of the person with dementia. For example, the use of a picture button telephone assisted a person with dementia in longer instances of interaction and maintaining social contacts with neighbours, friends and family.

Freedom and autonomy

Some of the studies reported carers having to use controlling methods such as locking and restricting access and the AT seemed to offer an alternative solution of enabling the person living with dementia to become independent and participate in meaningful activities. This in turn had a positive effect on the carers. The AT also provided carers with additional personal time which was highly valued and, in many instances, helped create the balance between their own personal space and independence with that of staying connected with the person with dementia.

Safety

Carers viewed someone who has dementia’s ability to stay in the community and their physical safety as more important than privacy and autonomy. Tracking devices that supported safety were enthusiastically received and AT provided carer reassurance and enhanced independence for both the carer and the person with dementia.

Quality of life

Whether the person living with dementia used the AT independently or the carer assisted them, AT was perceived as removing worries and burden and generally improved mental well-being, especially when the carer was living away from the person with dementia.

Competence

AT was perceived as improving independence for someone who has dementia, this had a positive effect on the carer, with some carers also reporting benefitting from using the AT themselves, such as the simple remote control for TV and memory aids.

Negative aspects

While the overall experience of AT use was perceived as positive by carers, some important negative aspects were also raised.

Relationships

When AT failed or the person living with dementia was no longer able to use the AT, this invariably caused constraints in the relationship, as an outcome of the presence of the AT. Some carers also perceived that the AT would replace the ‘person’ component of caring.

Freedom and autonomy

There were perceptions that the person living with dementia’s declining abilities could be further worsened using AT as they would no longer be actively challenged cognitively. Carers also believed that with the people with dementia who did not have adequate social care could be left alone with the technology without additional support for autonomy or social contact.

Competence

Carers seemed to be more willing to use AT in the future rather than currently. Elderly carers also worried about their competence and familiarity with AT, especially when there were technical failings with the AT or when the devices required to be replaced with new AT, as the illness progressed.

Quality of life

Occasionally, the use of AT seemed to create more dependence of the person with dementia on the carer, which led to increased stress for the carer, and the attitude of the person living with dementia towards the AT (from hostility to indifference) also led to additional carer burden, while choosing and using the AT.

Use of the AT

Ethical issues

Carers weighed the needs of personal reassurance and sense of security with that of autonomy of someone who has dementia while deciding on use of AT. Often there was no perceived ethical dilemma where the safety of the person with dementia was concerned. There was a consensus among carers that people with dementia must be involved as much as possible to select and use AT. Ethical issues around who held the power of choice of usage and discontinuance of AT and whether the needs of the person living with dementia were altered to match the potential of the currently available AT also seem to arise from the studies with no definitive conclusions.

Help and support from carers

Carers continuous engagement and willingness to provide support with the use of AT for the person with dementia was key in the use of AT in most of the studies. The carers’ attitude, commitment and willingness to learn about the AT were vital if the equipment was to be useful and functional.

Raising issues of using AT

Carers used different methods to convince people with dementia to accept and use AT, especially when the person living with dementia was hostile towards or did not understand the need to use the AT. Carers especially had difficulty convincing someone who has dementia where monitoring and safety devices were to be used compared to using AT for leisure and social interaction.

Acceptance and knowledge of AT

Costs and resource

Carers noted that AT was generally expensive, however most of the studies included in this review either provided the technology to the participants or participants did not mind spending the extra costs for AT that could support the person with dementia to stay for longer, in their own home.

Acceptance of AT

Many of the carers accepted AT as useful and their adoption depended on the perceived usefulness of the AT. They would also recommend its use to other carers and people with dementia. Carers also saw technological innovations as inevitable and expected the use of AT to increase and future generations of carers would have better skills and motivation to adopt them.

Knowledge of technology

There was a general feeling among carers that information regarding AT should be provided early in the process of diagnosis and support available to the person living with dementia, especially as the progress of dementia was unpredictable. The main need of information was on simple and practical AT solutions with most carers unaware of new AT devices and solutions available.

Discussion

The aim of this systematic review is to identify the types and uses of assistive technology in dementia and describe the effectiveness and experience of its use for carers. The studies included cover the last 18 years and give a broad picture of AT use in dementia care. Caregiving for people with dementia in the community is usually unplanned, unpaid work carried out by the relative of the person living with dementia. The role of carer can be rewarding, but it can also be detrimental to a person’s well-being and can put them under a lot of stress [100, 101], especially for a carer who has little experience. AT is one way for supporting people with dementia and their carers to stay for longer in the community.

The symptoms which have the highest impact on carers of persons with dementia are repetitive questions, apathy, getting lost, aggression and incontinence [37, 40, 66] but the AT solutions from studies included in this review did not effectively address behavioural problems except safety/alert devices for wandering and getting lost. Fuhrer et al. [102] argue that effectiveness, efficiency, device satisfaction, psychological functioning and subjective wellbeing are essential outcomes for continued short-term and long-term use for AT. Findings from this review highlight that carers of people with dementia may prefer a specific type of AT, such as a GPS tracker, movement sensor or medication reminder and perceive it as useful but it may not have any real effect on outcomes of burden, satisfaction or wellbeing, similar to findings from other reviews on AT [27, 103]. One reason for this could be that existing outcome measures that are being used in AT studies may not be sensitive enough to measure change when using AT or are not valid in this context, perhaps as most measures were developed before AT was introduced.

This review highlights the continued lack of consistency in describing or classifying AT [104]. Other studies and reviews [7, 102, 105,106,107] have highlighted different ways of classifying AT used in dementia care. Having a classification system based on use (with more than one use per AT) from the perspective of the person with dementia and carer, as described in this review, may improve consistency of reporting and enhance synthesis of findings from trials and reviews. We have classified AT based on (i) Name of AT (ii) Type of AT (iii) function assisted (use or intended impact) and (iv) availability (commercial/prototype).

Though some research, involving robotic technology in institutional and simulation/lab based settings is looking into this [108, 109], this review identified the lack of sufficient number of AT to support basic and instrumental Activities of Daily Living for people living with dementia at home. This could be because it is difficult to develop and deploy potentially bulky/expensive AT in a non-institutional setting or perhaps human/assisted care is seen as easier and less expensive way of providing this care [110]. It is also possible that technological advances in miniaturisation and artificial intelligence have not yet caught up with this area of need.

It is also clear from this review that installation of AT at home for use by someone who has dementia was often wrongly seen as a one-off event, rather than an ongoing process for getting the best out of AT. Similar to other findings [111,112,113,114], this review found that carers as users of AT often struggle to understand and engage with the technology in their homes as a result of poor understanding, a lack of knowledge of available AT and lack of on-going support from professionals and design flaws in the AT itself.

The review also highlights the perceived fear among some carers that use of AT could lead to social isolation. However available AT solutions such as tablet computers and monitoring devices to alert carers gives them a sense of participating in the life of a person living with dementia even when the carer is not physically present, this led to AT being viewed as a positive addition. There was no evidence within the included studies that multiple AT solutions were being harnessed to bring them together for an integrated solution that could assist both people with dementia and carers. AT devices were used in isolation for specific functions rather than a combined use of the devices. With the rise of internet of things [115, 116] and connected AT devices combining multiple AT for use with a person with dementia or carer is feasible and in most instances more desirable [117].

Interestingly all the studies considered the introduction of AT after a diagnosis of dementia, the timing of introducing devices may be important. Safety/tracking devices were introduced pre-emptively to prevent secondary problems [7, 27] such as falls and wandering, which in turn could potentially reduce admissions into long-term care [118] but equal consideration and further research may be needed for the use of AT as a preventative measure especially in areas of orientation, memory and leisure.

Many of the installed AT did not meet the needs of the user. Despite a surprising lack of reporting on adverse events, some of the negative reactions to AT were because they were ‘Off the shelf’ devices and were rarely useful, especially with a progressive condition like dementia. The AT needed to be adapted or customised for the carers and people with dementia’s individual needs and when this was not the case, led to abandonment of the AT [117, 119, 120]. Co-creating AT with users has steadily improved over time. Carers need to be involved in the design and testing of AT solutions and in prioritising the problems that need to be addressed to allow AT to be accepted as a solution for caring for people living with dementia in the community [121, 122].

Implications and recommendations from this review

  1. 1.

    The function assisted domain (e.g. Memory device, GPS tracker) as a way of naming the AT is usually defined by the manufacturer/developer of the AT. We recommend a shift towards considering naming the use of the AT from the perspective of the person with dementia and their carer to ensure that device is appropriately used and can provide the intended benefits of that AT [123] for both the carer and the person living with dementia.

  2. 2.

    Further research should be carried out on how multiple AT devices could work together or be combined to better support someone who has dementia and their carers rather than how individual AT devices can support them.

  3. 3.

    Future research should focus on AT solutions which are co-designed by those with lived experience of the challenges of dementia at home and should include carers, who live with and away from a person with dementia.

  4. 4.

    Ability of a carer to ‘problem solve’ should be a consideration in AT prescription and use. Technology should match the needs of the person requiring the use of the AT, rather than the person being ‘moulded’ to match what technology is available for them.

Limitations

Due to the variety of AT devices and outcome measures used, we could not pool results from the quantitative studies and have provided a narrative review instead. Due to financial constraints we did not include studies in languages other than English within this review and this could have potentially led to some suitable studies being missed. However, we did scan for reference lists of all studies that were included for full text review and are confident that this review captures all suitable studies that met our inclusion criteria.

Conclusions

Technology is advancing at an extremely rapid pace, especially within the fields of artificial intelligence and machine learning with their resultant healthcare applications. It is likely that AT powered by AI may become ubiquitous soon. The quality of research focussing on AT use in dementia continues to be low. AT solutions helps improve carers’ experience of providing care to a person living with dementia. AT would support people with dementia and carers in the community but researchers, healthcare professionals and technology developers should adopt a family centred model for use of AT than pursuing only an individual/person centred model of care.