A systematic mapping review was used to identify the published original articles related to intention to use digital technology by older adults. The review protocol was registered with PROSPERO (International Prospective Register of Systematic Reviews; registration number CRD42022329705) and the selection process is outlined in Figure 1.

A search was conducted from inception to November 15, 2022, in 9 databases including CINAHL Complete, MEDLINE Complete, PubMed, Embase, and Scopus, as well as in 3 information systems–focused databases, AIS Electronic Library, IEEE Xplore, and ACM Digital Library, and in the journals of the AIS Senior Scholar Basket. Broad keywords used were the following combination of 4 groups of keywords: older adults (eg, aged, older, senior, elderly) AND technology factor (eg, acceptance, adoption, use, adherence, rejection) AND influence (eg, behaviour, psychology, motivation) AND technology type (eg, technology, computers, eHealth, system, assistive technology, robotics, smart home, gerontechnology, telemonitoring). The full search strategy was updated on November 15, 2022 and is provided in Multimedia Appendix 1.

Original and peer-reviewed journal publications and conference proceedings written in English and using qualitative, quantitative, or mixed methods research were included. Studies had to include participants who were older adults (aged 65 years or older) and had an evaluative component of either demographic, cognitive, physical, or emotional factors that influenced the intention to use or actual use of digital technology.

All titles and abstracts of the identified studies were subsequently screened for eligibility by 3 reviewers (JS, LD, and TS) independently, applying the following inclusion criteria: (1) original and peer-reviewed research written in English with either a qualitative, quantitative, or mixed methods approach; (2) participants were older adults aged 65 years or more; and (3) research was aimed at investigating factors that influence the intention to use or the actual use of digital technologies for older adults. The 3 researchers then conferred to resolve any discrepancies on eligibility, and if an agreement could not be made, a fourth reviewer (AG) was consulted. The full text of the remaining studies was then checked and those that did not meet all inclusion criteria were excluded. The references of the selected studies were hand searched for other potential studies (snowballing method; Figure 1).

Methodological quality was assessed using the Critical Appraisal Skills Program, the Cochrane Risk of Bias, and the Mixed Methods Appraisal Tool (MMAT) [23] which, in addition to specific criteria for qualitative and quantitative research, also contains specific criteria on the relevance of the use of a mixed methods design and the integration of different types of results. The researchers assessed the methodological quality of all included studies independently, followed by a discussion of their findings to determine the final rating for each study. It was decided not to exclude articles based on quality assessment because there is little empirical evidence on which to base exclusion decisions in mixed studies systematic reviews [23-25]. Instead, it was decided to report on the quality of the reviewed articles and to apply independent triangulation: at least two quality criteria had to be present in studies to be included in the results.

As included articles ranged from qualitative to quantitative methods (or a combination of both), data extraction forms for both design types were designed. For qualitative articles and qualitative information from mixed methods articles, acceptance factors were coded and entered. For quantitative articles and quantitative information from mixed methods articles, variable/factor name and level of significance were entered into the data extraction form. Two authors (TS and JS) reached consensus on the data extraction form for each article.

Thematic synthesis [25] was used to synthesize qualitative data on technology use and intention to use factors. Multiple sessions in the research team were then held to group factors derived from articles into descriptive themes for acceptance in post-technology implementation. JS, LD, and TS each created a conceptual model of the relationships between themes, and through iterative discussions developed 1 final model. Factors derived from qualitative articles and mixed methods articles were compared with factors identified in quantitative articles. This allowed us to highlight which factors were statistically tested in quantitative research. Quantitative articles were summarized thematically. For this purpose, all factors examined were first extracted and compiled. We summarized thematically related factors into individual group categories to create a better overview. Cluster analysis was used as an explorative method to establish the group categories and to classify the individual factors by purpose and type. Subsequently, we also added the factors examined in the qualitative articles.

The study characteristics of the 59 articles analyzed are shown in Table 1. Most articles (36/59, 61%) were written in the last 5 years, published in the United States (25/59, 42%), and conducted in a community-dwelling setting (23/59, 39%), with 39,153 older adults sampled (sample size ranged from 5 to 14,798).

Different types of technology were examined, ranging from mobile technologies and television-based to assistive technologies, with the most common technology type being technology devices. Based on the scope of functionalities, 2 main categories of technology were summarized [85]: (1) everyday technologies (including hardware devices, such as computers, smartwatches, tablets, computers, and services such as gaming, apps, and other technologies used to support daily living); and (2) remote or assistive care technologies, which were those that use information communication technology devices and networks to deliver health and social care technology. Remote or assistive care technologies were most commonly reported (32/59, 54%).

The purpose of technology was also defined. Technology that supports older adults’ independence through activities of daily living or instrumental activities of daily living was the most common (40/59, 68%), followed by technology that aims to enhance safety (eg, such as monitoring or assistive technology; 36/59, 61%).

In a 2008 study [31], a model was used for the first time as a theoretical basis. Subsequently, studies from 2015 onward have used models for their studies. The studies did not show a trend in terms of setting, study design, technology type, and model use. Most study designs used were quantitative (27/59, 46%), followed by qualitative (21/59, 36%) and mixed methods (11/59, 19%).

Only one-third (19/59) of the articles used 1 of the available theoretical models to explain older adults’ intention to use digital technology. The remaining articles provided descriptive evidence to support, formulate, or extend a theoretical model. Table 2 provides an overview of the theories used, the research fields they derive from, and the articles that applied them to their research.

A classification of the 12 models based on academic discipline of their origin and the constructs applied by the corresponding research article is provided in Table 3. Table 4 presents a summary of the used constructs grouped into categories in articles that used and did not use a theoretical basis.

The information systems field developed the TAM (n=6), TAM2 (n=1), TAM3 (n=1), UTAUT (n=3), and UTAUT 2 (n=1). These models focused on core constructs of perceived usefulness and perceived ease of use, attitude toward using, and intention to use the system (eg, TAM). The UTAUT also included the construct social influence and facilitating conditions. Both the TAM and the UTAUT were further developed, including TAM2 (n=1) and TAM3 (n=1), which highlighted trust and perceived risk on system use, and UTAUT 2 (n=1), which incorporated 3 other constructs including hedonistic motivation (eg, the pleasure of using a technology), price-performance ratio, and habits.

Three models had their origin in health-related research. Both the Health Belief Model [86,87] and Hoyman’s Wellness Model [88] explored the multidimensional unit of health and wellness by emphasizing human health needs holistically within their environment by addressing 4 dimensions. Finally, Anderson’s Healthcare Utilization Model [89] aimed to understand how and why people use health care services, assess inequalities in accessing health services, and aid in the creation of policies that will allow for equitable access to care. To predict or explain one’s use of health care services, the model particularly focused on an individual’s predisposition to use acute health care services, enabling factors that facilitate use and one’s perceived or influenced need for care.

From the sociopsychological domain, the Trans-Theoretical Model of Behavioral Change [90] described an integrative theory of therapy that assesses an individual’s readiness to act on a new healthier behavior, providing strategies or processes of change to guide the individual. The model is composed of constructs such as stages of change, processes of change, levels of change, self-efficacy, and decisional balance. The Regulatory Mode Theory [91] described how people approach situations to achieve their goals. Similarly, the Fogg Behavior Model [92] labeled main motivators (motivation), factors of simplicity (ability), and the types of prompts for goal acquisition constructs. The Self-Determination Theory [93] predicted health-related behaviors; however, the Health Belief Model focuses on the behavioral determinants influencing uptake of health-related behaviors, while the Self-Determination Theory explains behavioral motivation as dependent on whether basic psychological needs for competence, social inclusion, and autonomy can be satisfied.

From the research field psychology, the Self-Efficacy Theory [94] explained how well a person that can cope with the particular situation is dependent on the skills they have and the circumstances they face.

Articles showed that older adults had varying acceptance levels to different forms of technology. Two different types of digital technologies were found that focused on (1) everyday technologies or (2) remote or assistive care technologies (Table 5). We further assigned these types by purpose using categories identified by Peek et al [11] that explored how the technology type supported the users’ health and safety, social interaction, or independence.

Models were more commonly applied to remote or assistive care technologies (15/19, 79%), with the TAM and UTAUT models identified in more than 1 article [68]. Other models were used less frequently. Similarly, only the TAM, UTAUT, and UTAUT 2 models were found in articles exploring all 3 technology purposes (eg, [59,68,79,80]).

The setting in which the research participants reside is expected to have a major influence on their intention to use digital technology. Thus, we analyzed which technology types were studied in what setting. Table 6 provides the corresponding overview.

The setting most studied was community dwelling (24/59, 41%), followed by primary care (13/59, 22%) and home care (12/59, 20%). Community dwelling refers to older people who live independently within the community. Primary care refers to health services that include a range of preventive, wellness, and treatment measures for common diseases. Primary care providers include doctors, nurses, nurse practitioners, and health care professionals. Home care refers to the nursing and domestic care of people in need of care outside of partial or full inpatient facilities in their home environment. Overall, technology types were relatively distributed between settings, although home care and primary settings tended to have more remote or assistive care technology types, while everyday technologies were more likely to be found in community dwelling and community centers.

In our mapping review, we found that older ’adults’ intention to use technology was driven by 6 different categories: Demographics and Health Status; Emotional Awareness and Needs; Knowledge, Competence, and Perception; Motivation; Social Influencers; and Actual Technology Features. These categories could be mapped to 2 different main technology types (everyday technologies and remote or assistive care technologies) and 3 purposes (health and safety, social interaction, and independence).

Our mapping review provides an overview of the application of existing theoretical models of technology acceptance while identifying additional key factors contributing to the intention to use digital technology among older adults. Most articles did not describe an existing technology adoption model. In the last 4 decades we found insufficient attention on research of technology adoption models specifically in the health sector, and more recent empirical research presenting major gaps in the actual application of theoretical models. Most articles lacked an explicit theoretical approach, which makes it difficult to interpret and compare the results of studies in this area.

New models incorporating constructs such as belief, resilience, and health status to facilitate the intention to use digital technology for older adults were found, alongside previous reports of individual influencing factors (eg, age, gender). Synonymous with previous reviews on the topic of technology acceptance and intention to use, our review suggests that despite 2 models being popularly used (TAM and UTAUT), older adults’ acceptance of technology was influenced by factors beyond the key constructs of these models.

Existing reviews have investigated older adults’ acceptance and intention to use technology across multiple phases of technology implementation [95-98]. Similar to the extrapolations of our mapping review, other reviews have found diversity in the variables influencing intention to use digital technology among older adults, with individual variables also being considered. This includes age (eg, older age); health status factors such as mobility issues (eg, fractured wrists and fingers) and vision discomfort; technology features; and support factors such as ongoing costs and accessibility of instructions and guidance [97]. A recent meta-analysis also highlighted that social influences (ie, conversations with family, friends, and professional caregivers) had a strong impact on the intention to use digital technology, especially when it is new and in the beginning of the adoption stages [96]. Factors including personality, beliefs, and resilience were found in our review, suggesting a dynamic influence of psychosocial traits on technology. These preferences and concerns adjust over the course of time and technology implementation [98,99].

There have been multiple studies exploring how existing technology models contribute to older adults’ intention to adopt technology. However, research on technology adoption and older people should go beyond describing facilitators and barriers [100]. Instead, a better understanding of the environment of older people and their interaction with this environment should be developed [100]. We believe that uncovering the role of communication and interaction between older people and their environment should be a key health research concern, as a means of contributing to improved care for older people in the community.

Taken together, the findings from this mapping review have important repercussions on the validity and applicability of popular theoretical models of technology adoption and intention to use digital technology. Our findings correspond with those of previous research expressing concern about the impact of unexplored factors along with their potential interaction with key components of commonly used technology adoption models and their subsequent reduced predictive ability [95]. We have used the findings of previous research along with the additional constructs found in our mapping review to suggest contributions to the discourse around technology adoption and intention to use digital technology specifically in relation to older adults.

Traditional models of technology adoption have been largely developed from information systems and health behavior models (eg, Table 3, Figure 2). However, factors from different disciplines have not been successfully combined so far. It has been shown that health-related and social-psychological factors also represent a major element in the adoption of technology, especially for older people. Our mapping review highlights that traditional information system models such as the TAM and UTAUT require the implementation of health and psychology constructs. Here, we identified a significant gap in the research. Furthermore, it is essential to note that existing models do not particularly focus on the age of technology users. The importance and awareness of the role of age in information systems were already highlighted by Tams et al [101]. They pointed out that future research needs to focus more extensively on cognitive age rather than on physical age, as older people show difficulties especially in dealing with complex and contemporary technologies. Perceived physical old age as opposed to cognitive old age, as well as perceived health status, constitutes further factors of investigation.

The advantage of this approach is that it emphasizes the relevant beliefs and antecedents for general intention to use digital technology and, consequently, provides more directive insights for the design of intention to use digital technology. Based on results, we underline the nature of personality (eg, attitude, self-efficacy expectation, resistance to change, resilience), emotions and social influences as well as the cognitive age, which are linked to health-promoting or health-damaging behavior and can have an activating effect on the intention to use digital technology [102,103]. Alternative factors including demographic and health data should be incorporated to ensure models can be applied appropriately (ie, at multiple stages of implementation), are flexible (ie, to account for different types of technology and changing older adult preferences), as well as being both individual- and context-dependent to assist with sustained intention to use digital technology (eg, inclusive of health status and disabilities). However, application of our framework should be considered in the context of the research objective and adapted accordingly. For example, while these constructs provide a basis for understanding early users’ intention to use digital technology by specifically focusing on the attitudinal, social, and normative belief structure, functional features of the technology type need to be considered for the design, implementation, and intention to use digital technologies (eg, ease of interoperability, size of screen), as well as the research question.

Our review had a comprehensive search strategy covering social sciences, health care, and technology fields. However, it found a relatively small number of studies covering broad technology types and included limited theory-based studies as well as studies in conference proceedings. However, as ours is a mapping review, we focused mainly on identifying evidence-based gaps and trends across technology adoption in older adults more generally and the work was intended to be broad in scope. This may result in studies being overlooked (eg, from other databases and conference proceedings that are not peer-reviewed). Future studies should expand the scope to identify other factors and outcome variables, with more specific terms to capture wider forms of technology (eg, “digital assistants” or “bots”), as well as more inclusive eligibility criteria.

An additional limitation is that we discuss the stated theoretical models as equivalents without elaborating on their respective quality criteria. Previous research has already examined these models in the context of their research and shown variances, but there are considerably more opportunities to examine different age groups especially and to constantly include newly developed technologies. Future research could focus further on individual models in connection with specific categories of new and innovative technologies. Here, the focus should be on the health sector in particular. Existing studies were dominated by references to the TAM and UTAUT. Studies that pay more attention to interaction dimensions, emotional dimensions, and other resilience factors were lacking. Future research should bring these paradigms closer. While data on factors influencing use in the preadoption phase are extensive, findings on the postadoption phase were limited. To support the independence of community-dwelling older people over long periods, more research is needed to understand what influences the continued or sustained intention to use digital technologies after their introduction. Furthermore, additional quantitative research is required to understand which factors may have a greater impact than others and to examine the moderating or mediating relationships between factors.

Technology acceptance is influenced by numerous factors. Existing models of technology acceptance should be more intensively integrated and revised. However, there is limited research on technology acceptance among older adults. Further research targeting understanding of the complexity and timing of the acceptance process of different types of technology by older adults is warranted.