1 Introduction

Telemonitoring or remote monitoring between a health care provider and a patient in a home setting is a promising technological innovation to enable older persons with disabling disease such as chronic heart failure to remain at home while receiving improved care and reducing health care costs1. During the past decade, numerous randomized trials have demonstrated the efficacy of telemonitoring interventions in reducing hospital readmission rates and emergency department visits, and therefore health care costs, along with lower numbers of deaths and increased quality of life [1–6]. Telemonitoring care also comes with other benefits valued by patients. Telemonitoring has the ability to minimize the barriers to care among individuals with common co-morbid conditions—like arthritis, vision impairment, and cognitive decline—that makes transportation to doctor’s appointments difficult without extensive support from formal or informal caregivers. Telemonitoring also offers the opportunity to provide patients with frequent and timely advice and instructions without the need for face-to-face, in-person interaction, thus supporting patients’ capacity of self care and needs for autonomy in their home environments [7]. Telemonitoring may even be preferable for some older adults, who would choose the familiar, low pressure home setting to direct communication with healthcare providers in their busy offices [8]. Telemonitoring, in theory, encourages more widespread, equalized access to more frail and less mobile individuals, as well as those living in rural areas, by allowing them the same access to providers as their urban counterparts.

The problem of concern here is about the other side of the coin – those elderly persons who cannot benefit from telemonitoring opportunities. Evidence has begun to emerge, indicating an alarming phenomenon of digital divide—i.e., the inequality of access to, participation in, and benefit from the telemonitoring care programs for older adults, as shown below:

Older adults with high income levels are more likely to have access to equipment, such as computers and internet access that can be used for telemonitoring and telecommunication [9].

Older adults with higher levels of education are more likely to accept telemonitoring, adhere to medication regimes and taking responsibility for healthcare, due to generally increased levels of health literacy and better access to care [10].

When access to telecommunication equipment is not an issue, refusal to participating in a telemonitoring intervention program can be based on a combination of any of the following reasons, including: (a) feeling overly burdened to acquire the skills to engage in new tasks such as registration and routine data entry (low digital literacy); (b) feeling uncomfortable to take on responsibility for one’s own health, given they have grown up being a part of a cohort that was brought up to believe that health problems should be taken care by health care professionals (lower self-efficacy in self care); and/or (c) fear that telemonitoring will leave their medical data unsecure (technological phobia and distrust) [11–13].

Some refuse to adopt or terminate the telemonitoring service not because of the complexity of the medical condition itself, but because of co-morbidities—such as cognitive or sensory impairments—that could inhibit the use of telemonitoring technologies [14].

Those rural elders and their caregivers who originally accepted and participated in a telemonitoring program for heart failure terminated prematurely, often because the burden experienced by caregivers, heightened by the worsening or complications of the health problem of the elder patient, led caregivers to stop collecting and entering and monitoring the data [15].

The above problems are by no means easy to solve. As a result, a large number of older adults may not be able to participate in telemonitoring care. The exiting literature seems to point us to at least three different conceptual frameworks or perspectives about key factors contributing to the phenomena of digital divide. Below we briefly summarize these perspectives in order to provide a basic analysis of the problem of unequal access to, use of, and benefit from the telemonitoring care for elders.

2 Understanding Psychological Factors in Technological Acceptance

One of the most influential lines of research seeking to elucidate specific barriers to the acceptance and use of technology has to do with the Technology Acceptance Model (TAM) and its use in health care. TAM was originally developed in 1980’s and gained increasing recognition in the health care field during the last decade [16, 17]. By emphasizing the key constructs such as perceived usefulness and perceived ease of use, subjective social norm, and behavioral intention, TAM reveals the cognitive processes through which the psychological, social, and behavioral factors contribute to the decision or choice of acceptance or adaption of new technology.

The key limitations of the TAM approach are two fold: one is its overall trend of expansion of the model by adding more variables into the original TAM in order to accommodate the complexity of issues facing the constantly changing IT world. Various independent attempts to expand TAM have been made [18–20], with repeatedly adding new variables that some critiques regard as causing “a state of theoretical chaos and confusion” [21]. The other limitation is the TAM’s focus on the inner psychological process while relatively neglecting external contextual factors. Although newer versions of TAM-like modeling are able to incorporate contextual or moderating variables as shown in the Unified Theory of Acceptance and Use of Technology (UTAUT) that includes “facilitating conditions” [22] as a determinant of behavioral intention to accept technology, the critiques of TAM believe that this approach ignores the critical processes or stages of health information technology (HIT) development and implementation, and is not able to advise which technology is better for a particular social group and clinical setting, and what kind of social and financial consequences of HIT use might be [23].

According to the TAM perspective, the success of a telecommunication program, particularly in terms of reaching and engaging digitally disadvantaged groups, will very much depend on the thoughtful design of the telemonitoring device that should be not only affordable to the target population but also able to take into account the possible differences in their ways of perceiving whether the equipment is useful and easy to operate. An elderly patient with a higher level of health literacy, for instance, may agree to enroll in a telemonitoring program primarily based on her own perception of usefulness when she sees a diagram showing how the relevant symptoms can be recorded by the equipment and transmitted to her doctor for central monitoring, whereas her counterpart with a lower level of health literacy, at least some of them, may be more likely to be persuaded by the endorsement of her familiar doctor than the explanation of how the equipment is able to pick up a list of clinical indicators.

Also, we can expect those with higher digital literacy to take a more active part in the telemonitoring process (e.g., manually entering some of the monitoring data, in addition to those that can be automatically collected via blue-tooth technology) as they may not view a task like data entry as difficult, but for those who are less familiar with digital devices, an automated data recording scheme may become a necessity for a successful telemonitoring program. In general, the TAM perspective, with its focus on user perception and acceptance of technological design’s functional characteristics, tends to guide us towards a more client-centered or user-sensitive design, which is critical for a successful telemonitoring program.

3 Understanding Sociological Determinants of Digital Divide

In contrast to the TAM model that focuses on the inner psychological processes of technology adoption, another emerging approach is primarily derived from the digital divide phenomena and related literature that points towards the importance of the broad contextual perspective.

The digital divide literature emerged from the observation of a broad variation or gap among large groups, populations, or geographical regions. The gap reveals the inequality of substantial social consequences in terms of access to, knowledge about, and use of the information and communication technology [24, 25]. If HIT is believed to be a promising solution to modern health care delivery to general population, the digital divide phenomenon suggests a profound challenge to the above belief, that is, the population-based heterogeneity as revealed in digital divide indicates the possibility that a large group of digitally disadvantaged people may not be able to access, use, and benefit from the HIT-based heath care, while the digitally advantaged group can. If the digital divide is a proxy reflection of the social-economic status, education level, age, and/or geographic region such as rurality, as many digital divide researchers are concerned, then advance of HIT may expand the persisting health disparities in our society rather than reducing them.

The notion of digital divide has evolved and broadened, mainly in the past decade, to become a multi-layered concept that includes at least three waves of conceptualization. Prior to the term of digital divide was ever coined, there was discussion in 1980 s about differences between those who had computers and those who did not [26]. In the late 1990 s, the initial concept of “digital divide” first merged and was primarily centered on the inequality in access to digital hardware of equipment and infrastructure, such as computer, internet, and mobile technology, that are necessary for enabling information and communication technology. Despite the increased availability of ICT products, which seems to reduce the general technological divide across age, gender, and income levels, [25] inequality persists particularly in terms of access to latest or most advanced technological products.

The second wave of the digital divide concept had more to do with human resource development than technological access. What Hargittai [27] called “Second-Level Digital Divide” focused on skills or what was later referred to as digital literacy (i.e., knowledge and familiarity with ICT), and further e-health literacy (e.g., able to use HIT to address health needs), as barriers to the use of digital technologies. The digital divide discourse at this second level turned the access concern about “Haves and Have-Nots” to a deeper inquiry of those “don’t-want-tos” (i.e., when the computing device is a convenient social setting, at no cost, some people would still choose not to use) [28]. Furthermore, the digital divide at the literacy level also entails how leaders or “early adopters” play a role in facilitating training and other community activities that improve related literacy and skills of community members to use and engage with HIT. For the older population, such enabling efforts may also need to take into account the disability conditions or physical limitations such as visual deficits, arthritis and joint pain that often reduce one’s confidence in learning new ways of managing one’s health, thus limiting the adoption of technology for healthcare purposes [29, 30]. In addition to physical impairments like decreased dexterity and vision loss, older adults’ restrictions also include cognitive decline such as decreased concentration, which, even as a result of normal aging rather than disease, may contribute to perceived difficulty in use of technology [31].

The third wave of the digital divide conceptualization further extended the notion of the human resource development to its context, with a focus particularly on building an organizational or community culture that is committed to promoting the pro-technology, pro-health, and pro-responsibility values, and to integrating these values with institutional and policy support for use of digital technology. Those who live in a community without such a culture may witness and experience technology phobia or distrust, concerns about cost or low return of investment of money and effort, and over-dependence on external authority such as doctors to make decision on care management, thus prohibiting behavioral digital engagement, including initial use, subsequent use, and/or sustained use of health information technology [7].

Different from TAM, which takes a micro-analytic approach to the problem of variation in adopting or using digital technologies and tends to psychologize all relevant determinants of technology acceptance, the trajectory of the conceptual evolution of the digital divide is heading towards an increasingly macro-contextual approach, representing a rather intuitive realism and providing a broad framework to allow actual technological entities such as equipment and software as well as social processes including training on digital literacy and community activities to promote pro-health and pro-technology values and culture. Such a framework makes it easier to link an intervention design to its implementation setting. Based on this broadened digital divide perspective, efforts aiming to improve a digitally disadvantaged group’s access to and engagement with telemonitoring programs should include not only educational interventions to improve digital and ehealth literacy, but also social and community interventions for promoting values of self care and active participation in technology-based care improvement schemes.

4 Finding Technology Accelerators for Older Adults

Consistent with, but independent of the above digital divide literature and TAM discussions, researchers from the Center for Technology and Aging proposed a model for Accelerating Diffusion of Proven Technologies (ADOPT) [32]. Their ADOPT model suggests that in addition to technology developers, external collaborators (organizations and individuals who work directly with older adults) could play an important role in facilitating technology diffusion. These collaborators include but are not limited to aging service organizations, formal/informal caregivers, family members, medical providers, and health plans. The presence of collaborators is especially important for older adults, as this population often has a lower level of familiarity/awareness of technology, cognitive or physical limitations that make it difficult to use technology, and limited resource to access technology. Without collaborators, technology access and user education can be difficult, if not impossible. Besides collaborators, Wang et al. also pointed out that the wider context in which older adults live could also impact their intention and ability to use technology. Two key context factors that they indentified are policy (including reimbursement, privacy considerations, and interoperability) and resources/access relating to technology. For example, reimbursement and policies that either reimburse technologies directly or incentivize health outcomes could likely increase diffusion of beneficial technologies.

According to the ADOPT model, in order to successfully promote a telemonitoring program among digitally disadvantaged people, it would be critical to form partnerships with “collaborators” including not only caregivers and family members but also those associated with a service organization such as aging service agency, medical center, and health insurance company, instead of promoting it alone by e-health developers or health care innovators. These collaborators can help accelerate the diffusion of technologies through adjusting incentivizing mechanisms and service arrangements to encourage and enable at-risk older adults to enroll in, engage with, and continually use and benefit from the telemonitoring care services as well as incentivizing their providers for participation in the service program. The notion of using a group of members of a community to help another group of people to accelerate diffusion or adoption of technology can be traced back to original writing of Everett Rogers [33] who explained the importance of innovation champions and opinion leaders in the technology diffusion process.

5 HCI Design Recommendations

These three conceptual frameworks lead to different emphases and directions—i.e., centering on technological design, human resource development, and service system collaboration, respectively—in finding possible solutions to the problem of designing telemonitoring care services for technologically, socially, and clinically disadvantaged groups.

As summarized in Table 1, from the TAM perspective, the key strategy to improve patient engagement and adherence is through careful product design that aims to enhance usability (e.g., “perceived usefulness”) and lower usage barriers (i.e., “perceived ease of use”), so that those who already have access to the monitoring device will feel motivated to use the telemonitoring services. From the broadened digital divide perspective, effective designs of telemonitoring intervention for digitally disadvantaged users, who typically are low-income, low-literacy, home-bound older adults, should focus on supporting, strengthening, and enabling digitally disadvantaged individuals and communities at various levels. From the technology accelerator perspective, the technology-based intervention design efforts should focus on integration of telemonitoring programs with the social, clinical, and financial environments.

Table 1. HCI design recommendations for effective delivery of telemmonitoring care to digitally disadvantaged users: based on three conceptual perspectives
Full size table

6 Limitations and Under-Addressed Issues

These theory-based recommendations are consistent with the emerging literature on factors contributing to differential technology and information use in general and adoption of telemonitoring technology among older adults in particular [8, 34–40]. Nevertheless, several issues, which may be regarded as beyond the scope of the aforementioned conceptual frameworks, remain under-addressed by the research community interested in delivering telemonitoring care to digitally disadvantaged older adults. One, most of the measures used for analyzing the digital divide assume a quantitative difference (e.g., ehealth literacy) between digitally high-access and low-access groups, while less attention is paid to possible qualitative differences between these two groups. For instance, if values, priority of goals, meanings of usefulness, exposure to health care systems and HIT services, types of digital literacy, and disability conditions of the target users are all different from general population, it would argue for the need of more qualitative inquiry, such as ethnographic research to deepen our understanding of the profile of various digitally disadvantaged groups, before quantitative research is planned. Two, most of the discussion focuses on initial behavioral acceptance, use, engagement, or adoption of telemonitoring technology, with much less attention paid to retention, or temporal variation in user intention and usage behavior. While recognizing that the barriers to sustained usage or adherence are of critical clinical significance, we know very little about how to model or analyze the temporal variation of these barriers and their complex inter-relations, while taking into account normal age-related physical and cognitive changes in late life, as well as disease related decline, fatigue of engagement over time, and psychological adaptation to and habituation of routine activities. Three, although for many older, frail adults, the decision of using or continuing to use a telemonitoring care service is made by, or in collaboration with, informal and formal caregivers, very limited research effort and intervention designs are centered on caregivers to elders in need of telemonitoring services.

The future success of telemonitoring care will depend on whether we can effectively spread the technology-based service across the digital divide that has prevented a large population from access to, adoption of, and benefit from telemonitoring care programs. Much work, including perhaps both qualitative policy research and quantitative intervention research, is still ahead of us to develop better and more aging friendly devices and programs, to strengthen and build the user community, and to integrate the telemonitoring program within existing clinical, financial, and social contexts.