Mild cognitive impairment (MCI) affects a person’s memory or how they think, feel, or behave. It is not a normal part of aging, but up to 20% of adults older than the age of 65 years may be affected, and up to 15% of those may progress to dementia [1,2]. Historically, MCI was considered a decline in a person’s memory, but the criteria and definition of MCI have expanded with more research to include other cognitive domain deficits and affective attributes such as language, executive functioning, or visuospatial ability [3-5]. A memory decline is only labeled single-domain amnestic MCI. A decline in 1 nonmemory cognitive domain is labeled single-domain nonamnestic MCI, and a decline in memory and one or more other cognitive domains is labeled multiple-domain amnestic MCI. People with MCI may forget things more frequently, lose items more often, struggle to remember words or have problems with language, have trouble making decisions or following instructions, lose their train of thought, have new difficulties regulating emotions, behave more impulsively, have trouble with visual perception, have trouble sleeping, experience motor coordination issues, or be less able to follow daily routines [2]. Given that visual, perceptual, cognitive, and behavioral functions are needed for human-computer interaction (HCI), the limitations associated with MCI create challenges for HCI, such as difficulties remembering process steps or commands, understanding instructions, interpreting feedback, correcting errors, or maintaining focus and attention. Changes to emotions, feelings, or mood may also make MCI more difficult and thereby create frustration, increase fear of making mistakes, and reduce self-confidence, all of which have been shown to contribute to factors influencing technology adoption (TA). These challenges can be managed through a considered design of technology solutions.

The global prevalence of MCI is increasing and placing additional stress on both health and aged care services [6]. The number of older Australians with MCI is estimated to grow from 884,000 in 2022 to 2.03 million in 2070 [7,8]. The Australian Royal Commission into Aged Care Quality and Safety [9] noted that aged care services are already under pressure and recommended exploring technology options to address the growing needs of older Australians. Australia’s National Dementia Action Plan 2024‐2034 [10] recommended that action be taken to (1) implement evidence-based early interventions to reduce cognitive decline, and (2) provide support and resources for people with MCI.

Research has shown that technology provides opportunities to increase health care service productivity [11], but only if solutions are adopted by the target users. Aging may result in slower cognitive speed, poorer memory, reduced concentration, and negatively impact vision, hearing, speech, dexterity, mobility, and learning abilities. These limitations have been found to discourage the use of technology [12], and studies have shown that the use of digital health solutions is lower among older adults than younger adults [13]. In addition, as people experience MCI uniquely and with different symptoms, older adults with MCI will have different and diverse needs for technology support.

TA has been studied from several perspectives. Early models, such as the technology acceptance model (TAM) [14], TAM2 [15], and TAM3 [16], take a behavioral perspective and consider TA in an organizational context. They describe TA as being dependent on a user’s perception of the technology’s usefulness, ease of use, and the determinants of these. The unified theory of acceptance and use of technology (UTAUT) [17] considers TA from a combined behavioral, psychological, and information management perspective, and UTAUT2 [18] extends UTAUT to consider TA outside of business settings, such as at home, and shows that the context of use has an impact on the relevance and contribution of determinants. However, these models were not developed based on data about older adults with cognitive impairment or the assessment of health care technologies. More recent models, such as the senior technology acceptance model (STAM) [19] and the model for the adoption of technology by older adults (MATOA) [20], focus on older users and the effects of physical and cognitive aging on TA. They note that ease of use becomes more important as users age, and MATOA identifies self-management (the extent to which a person can remain independent and maintain control of their life, emotions, and role) as a determinant for intention to use. The lifespan theory of control [21] describes the importance of control and states that people will attempt to retain control of how they behave and interact with their external environment for as long as possible because control contributes to a sense of well-being. It is therefore relevant to consider user empowerment as a valuable design feature in technology solutions designed for older adults, and this is supported by a literature review of older adults’ intention to use technologies [22]. STAM and MATOA do not specifically consider TA by older adults with cognitive impairments or technology designed for health care or assistance with activities of daily living. The health care technology acceptance model (H-TAM) [23] describes acceptance of health care technologies by older adults with hypertension. It includes many factors that appear in TAM-based TA models and proposes compatibility (how well the technology integrates into a person’s everyday life) as a determinant of facilitating conditions.

A study of people’s intention to use service robots [24] suggests that TA models may need to be revisited to incorporate the influence of technology innovations such as smart technologies and robots, which may be adaptive and exhibit human-like features. These advances have enabled new uses and new modes of interaction, and users can now choose their device and interaction mode, which is empowering. However, it is not clear how this impacts TA for older users with cognitive impairments, in particular, perceptions of usefulness and ease of use, which are shown to be determinants of TA. Researchers have investigated the effectiveness of unimodal and multimodal interaction for older users and found that modality affects use [12,25-29]; however, these studies did not include older adults with cognitive impairments. Subtle speech differences can be observed in people with MCI [30], and literature reviews have reported on the use of voice-activated technology solutions by older adults [27,31,32]. However, these did not always include older adults with cognitive impairment as participants.

There are gaps in understanding the combined impacts of aging and cognitive impairment on factors affecting TA for older adults with MCI. It is also unclear how MCI impacts HCI and device and interaction mode preferences. This scoping literature review examines published studies about technology solutions developed for older adults with MCI (amnestic or nonamnestic, and single- or multi-domain) to answer the following research questions:

To our knowledge, this is the first study that collates opinions and feedback from older adults with MCI about technology solutions proposed for them.

This scoping review design was guided by previous literature [33,34] and is reported according to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines [35]. It was preregistered on the Open Science Framework Storage (United States) [36].

An electronic database search was performed to identify studies published in English between January 1, 2014, and May 1, 2024. A 10-year period is considered sufficient given the fast evolution of technology and the potential for change in people’s opinions.

The search was performed consistently across nine electronic databases (ACM Digital Library, EBSCOhost CINAHL Plus with Full Text, EBSCOhost Computers and Applied Sciences Complete, Google Scholar, JMIR Publications, IEEE Xplore, EBSCOhost MEDLINE, Scopus, and Web of Science Core Collection).

The search strategy was developed with the assistance of a qualified librarian. The following search query was used in all databases:

Studies that satisfied the search criteria were imported into the referencing software program EndNote (Clarivate Analytics) and then uploaded to Covidence (Veritas Health Innovation) to remove duplicates and for screening [37].

Three reviewers and the primary author screened the titles and abstracts of each study and excluded any that did not meet the following inclusion criteria:

Papers were excluded if:

Where it was unclear if a study met the inclusion criteria, or there was a disagreement among reviewers, the study was carried forward into the full-text review.

The primary author independently reviewed all full-text studies extracted from the previous stage. The remaining authors shared the role of second reviewer. Full-text papers were included if they reported the following:

Any disagreements were resolved by discussion among the authors.

Data were extracted into a structured Microsoft Excel spreadsheet by the first author and evaluated by 2 other authors. Data were extracted about (1) study details (such as author, location of study, year published, journal name, study aim, study type, technology being evaluated, and participant details) and (2) participant feedback. Participant feedback was recorded as any reference to (1) a participant’s opinion about the technology (usefulness, ease of use, usage, and context of use), (2) how the solution made them feel, (3) preferences for device or interaction, (4) issues or challenges with use, and (5) recommendations for changes. Statements were captured as written, relating to the type of technology and solution.

All extracted data were reviewed for familiarization and to start the process of identifying similarities, relationships, and trends. Participant feedback was organized and analyzed using inductive, thematic analysis as described by Thomas and Harden [38], to ensure transparency and traceability from the data to the outcomes. This process has been used effectively to complete other thematic analyses of qualitative data that describe people’s experiences, opinions, and feelings.

Inductive thematic analysis of participant feedback was guided by the 2 research questions and completed in 4 steps:

During the above 4-step process, there was a continuous feedback cycle with coauthors.

A total of 4271 studies were imported for screening, 687 duplicates were identified and removed by Covidence, and 14 duplicates were identified and removed manually. A total of 3570 studies were screened by title and abstract, and 3320 studies were subsequently excluded. The full text of 250 studies was assessed for eligibility, and 167 studies were excluded, leaving 83 studies for data extraction. Figure 1 presents the flow of studies through the PRISMA-ScR process, and Table 1 provides a summary of the studies and extracted data, sorted alphabetically by the author.

Below, we first list 4 broad categorizations of the examined studies in terms of (1) study details and nature of participants, (2) purpose of technological solutions, (3) types of interaction devices, and (4) interaction modalities, and provide a short description of how studies evaluated usability. This is followed by the resulting themes derived from analysis of participant feedback.

Inductive thematic analysis of extracted data identified five themes: (1) purpose and need, (2) solution design and ease of use, (3) self-impression (how the solution makes the user feel), (4) lifestyle (how well the solution fits with the user’s life and routines), and (5) interaction modality.

Participants across studies reported similar needs but often in a different priority order, and this scoping review cannot (1) provide a definitive priority order for needs; (2) identify if needs are relatable to technologies or devices; or (3) attribute differences to the context of use, cultural differences, participant demographics, or the way MCI is affecting participants’ lives. Mondellini et al [85] found that user needs appear to be location-specific, but further research is required to understand which needs are location-specific versus those that are generalized.

Fadzil et al [139] completed a review of existing technology solutions for older adults with cognitive impairments and found that current solutions are focused on (1) caregiver support, (2) strengthening cognition, (3) cognitive rehabilitation, and (4) using technology for disease prevention or self-management. Fadzil et al [139] do not comment on reported needs, technology preferences, or opinions about existing solutions for older adults with cognitive impairments. Neither does it describe how well existing technology supports user needs, nor how well technology is aligned with users’ preferred ways of doing things. This gap contributes to the lack of understanding of TA by older adults with MCI.

Our review identified that security and privacy are important for older adults with MCI; nonetheless, they are prepared to compromise these to feel safe. However, they do not want to be continuously watched or feel like they are under surveillance. Nastjuk et al [140] used the term “techno-invasion” to describe the stress caused by constant observation and monitoring. Solutions must balance the need for users’ privacy and security with their need for safety and well-being.

A positive UX is a significant enabler for technology acceptance, and ease of use contributes greatly to UX, which suggests that designers must give both priority. User attitude has been identified in TAMs as a determinant of use, and several studies commented that MCI has an impact on a person’s attitude toward technology use [48,74]. This observation has also been made by other researchers [43], who noted that we need a greater understanding of how sociodemographic factors and cognition affect attitude.

Some older adults with MCI are not willing to connect with a robot because they prefer to interact with a human, while others are comfortable engaging with anthropomorphic solutions. Our review identified that participants who created an emotional connection with an anthropomorphic solution were vulnerable to becoming depressed once the solution was no longer available. This raises important issues for designers about how to support users if technology solutions fail, and ethical issues about how to manage emotional connection and attachment to more anthropomorphic solutions. It illustrates that technology represents part of the solution but not the entire solution [75], and that service providers must balance the need for human connection with the need to increase efficiency and reduce demand on health services.

Our review found that participants experience a variety of usability and usage issues, and that the flexibility of use and the ability to configure a solution increase ease of use. Older adults with MCI feel more comfortable, confident, and in control when solutions are consistent, predictable, and integrate into their existing lifestyles and routines. People often use multiple toolsets, and technology providers should prioritize consistency and seamless integration of their products. Overall, our analysis reveals that technology solutions should support how people behave and live, not require them to change and adapt to how the technology operates.

Some studies recommend that technology solutions for older adults with MCI should be designed together with this population [48,73,74,95,102,107,116] to ensure that all needs are identified and appropriately prioritized. This recommendation is supported by our findings, which show that while there is consistency in the top-priority need, there are differences in how participants in different locations prioritize the order of needs. Existing research shows that there is cultural and contextual influence on TA, and differences in the needs and priorities for independent living of older adults in Sweden and Poland, and the technologies they would consider using [141]. Given the evidence for the importance of perceived usefulness and ease of use as factors for TA, it is critical to understand the needs and characteristics of the target user base so we can design to maximize TA. User-centered design (UCD), first proposed by Kling [142], is a design methodology that focuses on usability and UX. UCD promotes the inclusion of end users in every stage of solution creation, starting with initial data gathering to understand users’ characteristics, environment of use, and requirements. UCD is iterative and incorporates regular artifact evaluation and refinement based on user feedback. UCD processes and techniques have been applied successfully to design assistive technologies for people with dementia and shown to be an effective way to accurately capture requirements and ensure solution usability and acceptance [143].

This review revealed that older adults with MCI have opinions about how and when they want to use technology, and it is imperative that future solutions are designed collaboratively with this population. Research is needed to accurately identify the needs of older adults with MCI so we understand what would be considered useful and how best to position technology in their lives. It would also be beneficial to identify needs that are common and which needs vary depending on demographics, location, or other factors.

Solutions designed for older adults with MCI must generate a positive experience because this affects a user’s comfort, confidence, and overall well-being. This review collated suggestions from participants, but more research is required to fully understand the factors to be considered.

Participant feedback suggests that personalization and control, and enablement are important for older adults with MCI, but it is not clear how choice with respect to technology, device, and interaction modality affects TA. This review identified that older adults with MCI do have preferred devices and interaction modalities; however, further work is needed to fully understand the effectiveness of different combinations of interaction modalities and how these are influenced by the effects of MCI, context of use, device being used, and task being performed.

Based on the findings, older adults with MCI want to have autonomy, independence, and feel safe. They are accepting of technology if it positively contributes to their lifestyle and provides a sense of well-being. Future research should aim to provide clarity about how to effectively meet safety, privacy, and security needs and establish technology as an enabler for independence and autonomy in older age.

Finally, the collection of longitudinal data about technology use by older adults with MCI would capture changes to usage patterns, preferences, and adoption factors as MCI advances and enable us to understand how solution designs should allow for adaptation to support MCI progression.

This review comprehensively searched 9 databases for studies over a 10-year period. It considered research about a broad set of technology solutions proposed for older adults with MCI in the pre– and post–COVID-19 environments.

However, the study is not without limitations. This scoping review only included studies in English, so it is possible that untranslated relevant studies were overlooked. Most studies published their outcomes promptly after data collection; however, [56] began data gathering in 2016 but did not publish the results until 2022. This is a long time in terms of technology evolution and potential for change in people’s opinions and attitudes. Some studies did not analyze nonverbal feedback, so it is possible that important information was missed [50,68].

Most studies established an MCI diagnosis using the Montreal Cognitive Assessment [144], the Mini-Mental State Examination [145], or Petersen’s criteria [146]. However, the cutoff criteria for each assessment tool were not applied consistently across studies.

The definition of “older adult” is not consistent across the 83 studies. Most studies considered older than 65 years as “older;” however, some studies included participants aged between 48‐59 years. Some studies did not have an equal representation of genders.

Many studies listed small participant numbers or a lack of cultural diversity as limitations, and most studies included participants from only one location, meaning it may not be possible to generalize outcomes. Less than half of the studies were completed in a real-life setting, so outcomes may not be fully representative of real-life results.

Studies did not collect a consistent set of participant data (such as IT literacy, existing devices and technology being used, education level, or profession), so it is not possible to comment on how these may have impacted participants’ opinions, although age, sex, education level, or experience are listed as moderators in existing TAMs such as TAM, UTAUT, and MATOA.

Some studies did not have people with MCI represented in all phases or excluded people with physical limitations, so the opinions of these people have not been considered. In addition, the perspectives of other important stakeholders, such as medical and allied health professionals who treat older adults with MCI, are not considered in this paper. A study exploring the perspectives of 20 medical and allied health professionals from France, Italy, Japan, and Germany, who treat older adults with cognitive impairments [147], identified that physical, social, and cognitive interventions are considered important, and technology is an effective way to deliver and support these. This scoping literature review is the first part of a multistage research project, and future studies will include other stakeholders, such as medical and health professionals, family, and caregivers.

This scoping review identified that improving or maintaining their current level of cognition is the top-priority need for older adults with MCI. Proposed technology solutions are perceived as useful by this population, even though gaps exist where functional support is missing. However, they are not perceived as completely easy to use, due to a variety of usability issues. In addition, technology solutions may have a positive or negative effect on the user’s comfort, confidence, self-esteem, self-image, and overall well-being, depending on the feelings and emotions triggered by the solution design. Older adults with MCI have preferences for how they use and interact with technology, and which devices they use. They prefer multimodal interaction—especially speech, visual or text, and touch—because it is effective and provides options for use, and devices that are comfortable, convenient, lightweight, readily available, cost-effective, and have large screens. Future work should focus on (1) gathering information about the needs of older adults with MCI and clarifying how these are dependent on location, demographics, or other factors; (2) improving ease of use and UX; (3) enabling customization and personalization; (4) further exploring interaction preferences and the effectiveness of different interaction modes; (5) providing options for multimodal interaction; and (6) integrating technology solutions more seamlessly into people’s existing lifestyle and routines.