International Journal of Medical Informatics 91 (2016) 44–59

Contents lists available at ScienceDirect

International Journal of Medical Informatics

journal homepage: www.ijmijournal.com

Review article

Smart homes and home health monitoring technologies for older

adults: A systematic review
Lili Liu a,∗ , Eleni Stroulia b , Ioanis Nikolaidis c , Antonio Miguel-Cruz a,d ,
Adriana Rios Rincon a,d
a
Department of Occupational Therapy, Faculty of Rehabilitation Medicine, University of Alberta, 2-64 Corbett Hall, Edmonton, T6G 2G4 AB, Canada
b
Department of Computing Science, Faculty of Science, University of Alberta, 307 Athabasca Hall, Edmonton, T6G 2E8 AB, Canada
c
Department of Computing Science, Faculty of Science, University of Alberta, 322 Athabasca Hall, Edmonton, T6G 2E8 AB, Canada
d
School of Medicine and Health Sciences, Universidad del Rosario, Calle 63D # 24-31, 7 de Agosto, Bogotá D.C, Colombia

a r t i c l e i n f o a b s t r a c t

Article history: Background: Around the world, populations are aging and there is a growing concern about ways that
Received 13 January 2015 older adults can maintain their health and well-being while living in their homes.
Received in revised form 12 April 2016 Objectives: The aim of this paper was to conduct a systematic literature review to determine: (1) the
Accepted 15 April 2016
levels of technology readiness among older adults and, (2) evidence for smart homes and home-based
health-monitoring technologies that support aging in place for older adults who have complex needs.
Keywords:
Results: We identified and analyzed 48 of 1863 relevant papers. Our analyses found that: (1) technology-
Frail elderly
readiness level for smart homes and home health monitoring technologies is low; (2) the highest level of
eHealth
Telehealth
evidence is 1b (i.e., one randomized controlled trial with a PEDro score ≥6); smart homes and home health
Gerontechnology monitoring technologies are used to monitor activities of daily living, cognitive decline and mental health,
Smart homes and heart conditions in older adults with complex needs; (3) there is no evidence that smart homes and
home health monitoring technologies help address disability prediction and health-related quality of life,
or fall prevention; and (4) there is conflicting evidence that smart homes and home health monitoring
technologies help address chronic obstructive pulmonary disease.
Conclusions: The level of technology readiness for smart homes and home health monitoring technologies
is still low. The highest level of evidence found was in a study that supported home health technologies
for use in monitoring activities of daily living, cognitive decline, mental health, and heart conditions in
older adults with complex needs.
© 2016 Elsevier Ireland Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2. Theoretical background: gerontechnology, smart homes and home-based consumer health technologies for older adults . . . . . . . . . . . . . . . . . . . . . . . . 45
3. Data and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.1. Data sources and search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.2. Studies selection process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.2.1. Inclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.2.2. Exclusion criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.2.3. Bias control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.3. Publications review and data abstraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

∗ Corresponding author.
E-mail addresses: [email protected] (L. Liu), [email protected] (E. Stroulia), [email protected] (I. Nikolaidis), [email protected],
[email protected] (A.M. Cruz), [email protected], [email protected] (A.R. Rincon).

http://dx.doi.org/10.1016/j.ijmedinf.2016.04.007
1386-5056/© 2016 Elsevier Ireland Ltd. All rights reserved.
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 45

3.3.1. Characteristics of the research conducted in smart homes and home health-monitoring technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.3.2. Features of smart homes and home health-monitoring technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1. Descriptive analysis of studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1.1. Characteristics of the research conducted in smart homes and home health-monitoring technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1.2. Features of smart homes and home health-monitoring technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.1. Gaps and implications for future research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.2. Study limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Author contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

1. Introduction 2. Theoretical background: gerontechnology, smart homes

and home-based consumer health technologies for older
In many countries of the world, populations are aging at an adults
increasing rate, as fertility rates decline and life expectancy rises
[1]. As a result, the population ratio of older adults relative to the In the United States, the prevalence of multiple chronic con-
total population is increasing. According to a report released by the ditions in older adults exceeds 60%. According to the American
United Nations, the proportion of persons aged 60 years and older, Geriatrics Society, complex needs are chronic conditions that fre-
compared to the total population, is expected to double between quently require services from different health care practitioners
2007 and 2050, and reach 2 billion by 2050 [2]. in multiple settings including frequent hospitalizations [5]. Older
Placement in a care facility, especially when it occurs against an adults with complex needs are limited in their ability to perform
individual’s wishes, has been associated with negative effects such basic daily activities due to physical, mental and psychosocial chal-
as depression, social isolation, and greater dependency in comple- lenges requiring complex continuing care [5]. Technology tools
tion of self-care tasks [3]. Therefore, older adults prefer to stay in provide the possibility for older adults with chronic conditions
their homes rather than enter a healthcare institution when they and complex needs to remain at home and maintain an accept-
need specialized care. For example, in a survey undertaken in the able quality of life [6,7]. Technological innovation for older adults
United States, 30% of those over 65 years stated they would “rather is occurring at an unprecedented rate [8]. “Gerontechnology” is a
die” than enter a nursing home [3]. Burden and costs of care not only term that combines gerontology and technology, coined to describe
rise in the healthcare system but also for informal caregivers (i.e., an interdisciplinary field of science for “designing technology and
family members, friends, neighbours). The significant costs associ- environment for independent living and social participation of
ated with the provision of care to an aging demographic, as well older persons in good health, comfort and safety” [9]. The term
as shortages in the health workforce, have spurred both industry “smart home” refers to a special kind of home or residence equipped
(e.g., Elite Care, Intel, Tunstall) and academia to undertake research with sensors and actuators, integrated into the infrastructure of
on the efficacy and feasibility of health monitoring and assistance the residence, intended to monitor the context of the inhabitant
provided in the home environment [4]. to improve his or her experience at home [10,6]. Smart homes
To the best of our knowledge, no literature review about smart can enable older adults to live independently at home longer and
homes and home health-monitoring technologies have character- reduce their reliance on informal or formal caregivers, or allow
ized the quality (as scientific evidence over health outcomes) of caregivers to better care for older adults. These technologies have
the studies carried out with these technologies or the level of tech- the potential to provide a cost-efficient approach to enhance one’s
nology readiness. The aim is to provide a comprehensive review quality of life and help older adults live safely in their homes [11]. A
that characterizes the state of the art of smart homes and home smart home can potentially provide a variety of services spanning
health monitoring technologies. The specific research questions of from simple task automation (e.g., room-temperature control), to
this review are: analysis or prediction of the location of a resident, to behaviour
or health status recognition of an occupant living at home, with
subsequent transmission of collected data for remote monitoring.
1. What is the clinical evidence of the outcomes in studies on
In health care, there are two key applications for smart homes:
smart homes and home health-monitoring technologies for
“(a) home automation: remote or automatic control of devices,
older adults with complex needs?
appliances, or systems at home to enhance an occupant’s quality
2. What is the level of technology readiness for smart homes and
of life, or to manage energy consumption; and (b) monitoring well-
home health-monitoring technologies intended for older adults
ness: monitoring an occupant’s health-status to maintain his or her
with complex needs?
well-being” [11,p. 4].
Technologies can be used to monitor wellness of older adults
This paper has six sections including the Introduction (Section with complex needs living at home. These are not necessarily
1). We present a general background about the concept of tech- embedded in the residence or building structure, but are designed
nologies for older adults in Section 2. Then we describe the data for home use by older adults with complex needs and their families,
sources and method used to locate and select the relevant literature and include wearable sensors to detect changes in vital signs [12].
in Section 3. Next, we present the results, where the research ques- Examples of home health-monitoring technologies are systems for
tions are answered in Section 4. In Section 5, we identify the main physiological monitoring (Phys), functional monitoring and emer-
research gaps and make suggestions for future research. Section 6 gency detection and response (Fx), safety monitoring and assistance
is our conclusion. (Saf), security monitoring and assistance (Sec), social interaction
46 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59

monitoring and assistance (Soc), and cognitive and sensory assis- third researcher for an assessment of the paper. We selected the
tance (Cog/Sen) (see Table A.1 in supplemental material for more third researcher from an interdisciplinary panel of experts, with
details) [10]. representation from the computing science and occupational ther-
There are three steps involved in monitoring the health and apy fields; they provided a final decision about whether or not to
well-being of older adults. First, one needs to monitor an occu- include a paper in the analysis (see Fig. 1 for more details). The
pant’s context using a set of appliances (devices, objects), sensors level of agreement between the raters was excellent, i.e., 93.45%
and actuators in an array of heterogeneous sensor “layer” plat- average agreement for abstracts (average kappa (␬) score of 0.807,
forms (also called the “acquisition layer”). The communication with p < 0.000), and 92.95% average agreement for full papers (overall
and among the devices, sensors, and actuator elements of a sensor average kappa (␬) score of 0.618, p < 0.000).
platform uses wired or wireless protocols (e.g., Wi-Fi, Bluetooth,
Zigbee, cable). Second, sensor data must be processed and ana-
lyzed to recognize the context of the inhabitant and environment 3.2.1. Inclusion criteria
(e.g., “service layer”). Third, applications available in the pervasive 1. Studies that included smart homes and home-health monitoring
space (“application layer”) activate and deactivate services and a technology that:
graphical environment which helps make smart spaces “visible” to a. Addressed technology use in home or supportive care envi-
users [13]. Finally, all data gathered from this pervasive environ- ronments for older adults with complex needs (i.e., private
ment can be stored either locally or remotely in a database. The residences, retirement villages, service-integrated housing
database can be processed and the derived knowledge can be used and independent living facilities were considered as potential
to alert caregivers by reporting an inhabitant’s situation or trig- home environments) regardless of whether the technology
gering alarms (data transmission) [14] or the data could be used was embedded in the building structure or was worn on the
to control the environment to improve an inhabitant’s experience person;
[4,15]. For example, devices and actuators automate actions such b. Addressed complex needs faced by older adults (physical,
as room temperature [16] or lighting control [13]. mental, or both); and
c. Included technology (ies) that has been implemented or
deployed at least in pilot form (with one older adult at a
3. Data and methods minimum) focused on supporting independence, or enabled
collection of data for health monitoring or communication of
3.1. Data sources and search strategy older adults with complex needs.
2. Studies that included participants who were older adults with
As smart home and home health-monitoring technologies complex needs:
research has been conducted in both health services and tech- a. Aged 60 years or older;
nology disciplines and scientific findings have been published b. Who required continuous care including permanent monitor-
in different literature repositories we decided to conduct our ing and who had chronic conditions generating limitations on
search in the following databases: Scopus, PubMed, Cinahl, IEEE their ability to perform basic activities of daily living (ADL1 ),
Explore, ISI Web of Sciences, and ACM Digital Library. We under- and instrumental activities of daily living (IADL2 ) at home due
took a search of literature published between 2010 to October to physical, mental or psychosocial impairments. If a given
2014 related to technologies that support older adults in their study addressed only one chronic condition such as COPD,
living environment. Papers were extracted from databases using diabetes or heart condition, we included the paper if it explic-
the following search terms alone or in four different combina- itly reported participants’ limitations in performing basic ADL,
tions using the logical operators of “AND” and “OR”: “smart home”, instrumental ADL, or both, at home due to physical, mental or
“health”, “gerontechnology”, “assisted living”, “tech*”, “monitor- psychosocial impairments.
ing”, “older adults”, “telesurveillance”, “telemonitoring”, “older 3. Studies published in English and available in full-text in peer-
adult”, “senior”, “elderly”. Table A.2 in the supplemental material reviewed journals or conference proceedings from electronic
presents terms used for different search queries. abstract systems.
4. Papers that used any type of study design or methodology, with
3.2. Studies selection process positive or negative results.

The database search and the initial removal of duplicates 3.2.2. Exclusion criteria
were performed by experienced researchers. Two independent 1. Studies published in books, book chapters, PhD or Masters’ the-
researchers evaluated the titles and abstracts of the remaining ses.
articles and compared them to the inclusion and exclusion cri- 2. Papers that were lecture notes in conferences, theoretical
teria. We paired each researcher with a research assistant. Both papers, narrative reviews, meta-analysis, and other types of
pairs of researcher and research assistant met to reconcile dif- literature review.
ferences through discussion. If there was any disagreement on 3. Research conducted in hospitals, nursing homes or rehabilita-
a paper’s abstract suitability, the abstract was included. Next, tion facilities. These facilities were excluded because assistance
the two researchers and two assistants reviewed the full texts (i.e., physical and psychological) was provided to residents in
of the selected papers. Each independently assessed a quarter of these settings [17].
the papers to determine suitability for inclusion in data analysis. 4. Studies that were not health-related, focusing on home-based
The same four independent raters completed data abstraction of technology for other purposes such as energy efficiency or
the final selected papers, and annotated the operationalization of
variables in a codebook. Finally, as a test of agreement between
researcher and research assistant pairs during the full text articles
1
review, and data abstraction, two new researchers independently ADL (activities of daily living) are self-care activities such as bathing, dressing,
eating, and personal grooming.
reviewed 20% of randomly selected articles and 20% of all vari- 2
Activities that are not necessary for fundamental functioning, but allow an indi-
able operationalization written in the codebook. In case of any vidual live independently in a community. For example: taking medications, use of
disagreement regarding the suitability of a paper, we consulted a telephone or other form of communication or, using technology.
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 47

Fig. 1. Process flow of our literature review.

home security (e.g., sensors or cameras solely used to monitor 11. Studies that reported mixed interventions (e.g., not control-
either energy consumption or to detect intruders). case studies) including remote-monitoring technologies with
5. Studies on assistive devices such as canes, walkers, wheelchairs other interventions such as home care education. Studies that
and hearing aids. incorporated interventions in addition to technologies under
6. Studies on other types of technologies not related to smart study would not allow us to know whether a change in the
home or home monitoring, e.g., Nintendo Wii. outcome was attributed to the technologies or to other inter-
7. Conference proceedings. ventions administered at the same time.
8. Abstracts or papers that were not available. 12. Papers that did not provide enough information for catego-
9. Studies published before 2010. rizing the paper (e.g., description of participants, technology
10. Papers on telemonitoring, telemedicine or telehealth programs readiness).
that involved self-monitoring using lower complexity tech-
nologies such as interactive voice response protocols through 3.2.3. Bias control
landline phones or cell phones; blood pressure measurement By including a variety of databases, we were able to guarantee a
using sphygmomanometers because we were interested in more thorough search, to achieve greater levels of sensitivity, and
high tech applications that involved electronic data transmis- to reduce source publication bias [18]. The inclusion of papers with
sion. positive and negative results eliminated the possibility of the publi-
48 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59

cation bias described by Scherer et al. [19]. The inclusion of papers studies, clinical-oriented studies, usability, or a combination of
registered in electronic abstract systems was as the first ‘quality them), study design (i.e., qualitative or quantitative research
filter’, ensuring some degree of scientific levels of conceptual and method), main outcomes of the study, and setting where the tech-
methodological rigor [20]. Studies published before 2010 were not nology was tested (i.e., house, apartment etc.).
included because, as suggested by [17], the rapid development of
smart homes and home-based consumer health technologies uti- 3.3.1.2. Condition/behaviors addressed or therapeutic application.
lized before this time would likely be obsolete. The exclusion of This parameter refers to the type of need that was addressed by
non-English papers from our sample represented only 0.48% of the the proposed technology (e.g., falls prevention or detection, demen-
total (see results section for more details). The inclusion of two tia support, chronic disease management, assisted living activities
pairs of raters during the selection process for relevant articles (i.e., either for daily living (ADL) or instrumental activities daily living
abstract and full paper reading), and a third reviewer in case of dis- (IADL)).
agreement, substantially reduced rater-bias that may have arisen
from the subjective nature of applying the inclusion and exclu-
3.3.2. Features of smart homes and home health-monitoring
sion criteria. Although inclusion criteria (1c) and (2a) may appear
technologies
restrictive, we chose them because we were interested in knowing
3.3.2.1. Level of technology readiness. This indicator assesses the
the actual levels of implementation, use, acceptance and effective-
maturity of evolving technologies during their development and
ness of smart home and home health-monitoring technologies for
early operations. We used the readiness scale of United States
older adults with complex needs.
Department of Energy [23], in which 9 levels are used as follows:
TRL1 (basic principles of technology are observed and reported);
3.3. Publications review and data abstraction
TRL2 (technology concept and/or application is formulated); TRL3
(analytical and experimental critical functions and/or characteris-
Each selected paper was carefully reviewed for the follow-
tic proof of concept); TRL4 (component and/or system validation is
ing attributes: characteristics of the research conducted in smart
conducted at laboratory environment); TRL5 (technology is devel-
homes and home health-monitoring technologies (i.e., therapeu-
oped and tested both in laboratory and validated in relevant
tic application and clinical implications (outcomes), sample size of
environments); TRL6 (technology demonstration, in which engi-
studies, client age, and level of clinical evidence of outcomes), and
neering and pilot-scale, similar prototypical system validation in
features of smart homes and home health-monitoring technologies
relevant environment is conducted); TRL7 (system commissioning
(i.e., level of technology readiness).
I, in which full scale similar prototypical system is demonstrated
in relevant environments); TRL8 (system commissioning II, actual
3.3.1. Characteristics of the research conducted in smart homes
system is completed and qualified through test and demonstra-
and home health-monitoring technologies
tion in relevant environments); and TRL9 (system operation, actual
3.3.1.1. Type of study and level of clinical evidence and outcomes in
system operate over the full range of expected mission conditions).
the studies regarding smart homes and home health-monitoring tech-
nologies. The quality of a study was measured by the Physiotherapy
Evidence Database (PEDro) scale.3 This scale allows assessment of 3.3.2.2. Application of technology. This parameter refers to the
the quality of clinical trials in terms of a study’s structure including classification provided by Ref. [10] for smart homes and home
randomization, blinding, attrition, design and statistics [21]. The health-monitoring technologies. According to this classification,
PEDro scale has 11 criteria. Criteria 2–11 are rated “0” for “no” and smart homes and home health-monitoring technologies were
“1” for “yes”. Criterion 1 that relates to the generalizability of the classified as either systems for physiological monitoring (Phys),
trial was not used to calculate the PEDro score, thus, the maxi- functional monitoring/emergency detection and response (Fx),
mum score that a trial can achieve was 10. Teasell and colleagues safety monitoring and assistance (Saf), security monitoring and
(2013) assessed the methodological quality of Randomized Con- assistance (Sec), social interaction monitoring and assistance (Soc),
trolled Trials (RCTs) based on their PEDro score. Scores of 9-10 were or cognitive and sensory assistance (Cog/Sen) (see Table A.1 in Sup-
considered as “excellent” quality; 6-8 indicated “good” quality; 4-5 plemental material).
were of “fair” quality; and below 4 were “poor” quality. After assess-
ing the clinical trials using the PEDro scale, we used an adaptation of 3.3.2.3. Technology type used in the acquisition layer. This param-
the modified Sackett criteria proposed by Teasell et al. [22] in order eter refers to the name and technology used for sensing and
to summarize the findings. Using these criteria, raters assigned a capturing data (e.g., accelerometer, gyroscope, switches, magne-
level of evidence for a given therapeutic intervention based on a 7 tometer, pressure sensors, GPS, cameras).
level scale. Table 1 shows the seven levels of evidence. The high-
est levels of evidence are achieved by RCTs and cross-over designs 3.3.2.4. Type of method used for communicating the data. This
(1a, 1b) and, the lower levels are provided by quasi-experimental parameter refers to the protocol and system technology for com-
and uncontrolled designs (Levels 2–4) or, conflicting evidence [22]. municating the captured data (e.g., Bluetooth, RFID, UWB radio,
As the field of smart homes and home health-monitoring technolo- Wi-Fi, Zigbee, ultrasound, infrared).
gies was broad in relation to the condition or behaviours addressed
through these technologies, we assessed the levels of evidence for
3.3.2.5. Technology type used in the application layer to analyze and
each condition or behaviour addressed in the papers analyzed in
process the data. This parameter refers to the algorithms, computer
this review and associated the paper with the highest level of evi-
programs, computing methods used for processing, transferring
dence achieved per condition or behaviour.
and storing data.
We collected data on sample size, the length of the experiment (in
years), type of topic or problem undertaken (i.e., technology-oriented
3.3.2.6. Approach or policy taken towards ethics and privacy of the
individual. This parameter refers to whether any ethics and privacy
3
The PEDro scale is a valid measure of the methodological quality of clinical
issues were addressed or taken into account to guarantee a privacy-
trials. Available at: http://www.pedro.org.au/wp-content/uploads/PEDro scale.pdf. sensitive storage and communication of individual’s data.
Accessed on July 11, 2014.
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 49

Table 1
Levels of evidence [22[22, p. 9].

Levels of evidence Number of RCT PeDRO Scale Score Additional characteristics

1a More than one 6 or higher Includes within subjects comparison with randomized conditions and
1b One 6 or higher crossover designs.

2 One 6 Non-RCTS and Cohort studies (using at least 2 similar groups with one
exposed to a particular condition).

3 NA NA Case-Control: a retrospective study comparing conditions, including historical

controls.

4 NA NA Case Series: retrospective chart review. Pre-Post: a prospective trial with a

baseline measure, intervention, and a post-test using a single group of
subjects. Post-Study: a prospective post-test with two or more groups,
intervention. Post-test using a single group of subjects.

5 NA NA Observational, studies using cross-sectional analysis to interpret relations;

clinical consensus, expert opinion without explicit critical appraisal, or based
on physiology, biomechanics or “first principles”; or Case Report, pre-post or
case studies (n = 1).

Conflicting NA NA • Absence of evidence: agreement by a group of experts on the appropriate

treatment course.
• Disagreement between the findings of at least two RCTs.
• Where there are more than four RCTs and the results of only one was
conflicting, the conclusion was based on the results of the majority of the
studies, unless the study with conflicting results was of higher quality.

4. Results

A total of 3378 papers were identified during the initial search

phase. After removing duplicates the number of papers decreased
to 1863. After applying the inclusion and exclusion criteria, only 48
of studies were included in the final analysis and data-abstraction
phase (2.58%, 48/1863). We found 40 of the included papers in
Pubmed (83.33%), 6 papers in CINAHL (12.5%) and, 2 papers ISI
Web of Science (4.17%) (see Table A.3 in Supplemental material).
Nine (only 0.48%) studies were excluded for not being in English
(i.e., German, Italian, Portuguese, Spanish and Serbian). Therefore,
the so-called Tower-of Babel bias described by Dickersin et al. [24]
for excluding non-English language papers was minimal. Interest-
ingly, half of the studies, i.e., 50.51% (941/1863) were excluded
because the papers did not meet inclusion criteria (1c) or (2a) or
both, i.e., the technology(ies) under study were not implemented
or deployed at least in a pilot (TRL 5), participants did not have
complex needs and were not aged 60 years or higher. Notably, 3%
of the studies (57/1863) were removed due to exclusion criteria 12,
i.e., after reading the full paper, there was not enough information
for categorizing the paper (e.g., description of participants, tech-
Fig. 2. Paper selection process of our literature review.
nology readiness). Other relevant reasons for exclusion of papers
from the final data-abstraction phase were that abstracts or papers
were not available (5.1%, 95/1863), and papers were conference plemental material for more details). Studies were characterized by
lecture notes, theoretical papers, narrative reviews, meta-analysis, low impact factor (i.e., SNIP mean 1.32 SD 1.86; 95% CI [0.73, 1.91]).
or other types of literature review (5.9%, 111/1863). Fig. 2 shows a However, the papers were published in journals located mostly
flow chart of the search and studies selection process.4 in Q1 journal quartile according to SJR journal classification [25].
Of the studies included, 96% were journal papers and 4% were Regarding the location of the home studies, we found that stud-
conference papers (see Fig. A.1 in Supplemental material for more ies were undertaken in 16 different countries. Almost a half of the
details). Interestingly, in a linear regression carried out between the studies were conducted in United States of America (48% US), with
variables year of article publication versus number of papers pub- 11% at multicenter or collaborations (see Fig. A.3 in Supplemental
lished and analyzed, we found a positive correlation between these material for more information).
two variables (correlation coefficient b = +0.48, and proportion of
the total variability r2 = 0.64, p < 0.000). This means that although
4.1. Descriptive analysis of studies
the total number of papers analyzed was small compared to the
total sample after duplicates analysis (1863), there was a clear
4.1.1. Characteristics of the research conducted in smart homes
upward trend of number of published papers (see Fig. A.2 in Sup-
and home health-monitoring technologies
4.1.1.1. Smart homes and home health-monitoring technologies − age
of participants, the size and the length of studies (number of par-
4
The% may not total 100% because several papers were excluded due to more ticipants/patients). The reviewed papers described studies with a
than one inclusion or exclusion criteria. mean age of participants of 75.40 years old (SD 7.17). The age ranges
50 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59

were from 60 to 96 years, accounting for 11,282 participants in There was conflicting evidence regarding home health-
total, with a high dispersion in the number of participants, i.e., mean monitoring technologies for patients with chronic obstructive
of n = 240.06 and SD 952.19. 68.75% of reviewed papers can be con- pulmonary disease (COPD). On one hand, there was level 1b evi-
sidered as small trials with a total number of participants <50 (i.e., dence from a RCT of a good quality that indicated older adults with
mean n = 16.78 SD 12.7), whereas the remaining trials can be con- COPD experience a lower rate of exacerbations and hospitalizations
sidered as medium-large (i.e., >50), with mean in the number of when wearing a Bluetooth wristband that telemonitored their vital
participants of n = 766.35 SD 1668.22 (see Table A.4 in Supplemen- signs. On the other hand, there was level 1b evidence from a RCT
tal material for more detail). The lengths of the studies were diverse, of good quality to indicate that a home-health physiological mon-
with a mean of 0.82 years and a SD 0.92 years itoring system did not have any impact on the number of hospital
admissions or hospital length of stay in older adults with COPD.
4.1.1.2. Smart homes and home health-monitoring technologies—type Regarding cognitive decline and mental health, there was level
of topic or problem undertaken. Almost 40% of the reviewed papers 1b evidence from a RCT of a good quality that older adults with
undertook the problem of technology acceptance, and 33.33% of chronic illness and comorbid depression exhibited reduced symp-
studies reported investigation in terms of patients’ clinical out- toms and post-discharge emergency-department visits due to the
comes. However, only 2.08% of studies had incorporated any sort of use of home health-monitoring technologies or devices for physi-
economic analysis (see Table A.5 in Supplemental for more detail). ological and functional monitoring.
Of all included papers, 60.41% used quantitative experimental In the category of disease, disability prediction, health-related
design (randomized controlled trials, cross-sectional, descriptive, quality of life, there was level 1b evidence from a RCT of a good
etc.), whereas, 35.42% of papers used qualitative experimental quality that use of technology to measure and track biometric data
design (e.g., grounded theory, phenomenology, qualitative case by itself did not improve or reduce the decline in frailty status in
study). In 4.17% of cases, studies used mixed methods. Only 18.75% older adults.
of all papers reported outcomes of randomized controlled trials, the Regarding fall prevention, there was level 2 evidence from a
gold standard for medical and health research. The other two quan- nonequivalent pretest-posttest control group design to indicate
titative experimental designs with relatively high specific weight that the provision of a fall detector for older adults did not make a
are technology feasibility studies (14.58%), and cross-sectional difference in reduction of fear of falling and in number of falls. How-
studies and one group pre-test post-test with a 6.25% of papers (see ever, older adults who used the device experienced improvements
Fig. A.4 in Supplemental material for more details). There were four in safety, independence and confidence.
medical conditions or behaviours that smart homes/home health- Finally, regarding monitoring of heart conditions, there was
monitoring technologies addressed, in which either no randomized level 1b evidence from a RCT of a home health-monitoring system
controlled trial (RCT) had been conducted or RCT studies were still (wired blood-pressure monitor, wireless weight scale, a function
very few, i.e., diabetes type II, falls prevention, reduction of use of programmed to ask participants about their health status, activities,
healthcare services, and cognitive decline and mental health (see and medication adherence, and that showed educational videos)
Fig. A.5 in Supplemental material for more details). that there was no statistically significant difference in rates of hos-
pitalization, emergency department visits or death in older adults
with heart failure who were monitored with a telehealth system
4.1.1.3. Smart homes and home health-monitoring compared with those who had case management. Under the same
technologies—what are the outcomes?. Of the clinically-oriented medical-condition category, there was level 1b evidence from a RCT
and economic-assessment studies that reported any outcome of good quality that older adults with chronic kidney disease and
(37.5%, n = 18), overall, 66.66% (12/18) of studies reported that hypertension using a wireless monitoring device improved sharing
the smart homes and home health-monitoring technologies of data with the clinic and showed a trend toward improvements
showed advantages in terms of clinical outcomes compared to no in blood pressure control.
intervention or other types of interventions (i.e., positive versus We also did a correlation analysis between the level of evi-
negative clinical outcomes was statistically significant, ␹2 = 5.56, dence and the technology readiness level, and found a positive,
df = 1, p < 0.05, see Table 2a). This provides reasonable evidence statistically significant relationship between these two variables
for the feasibility of smart homes and home health-monitoring (Spearman rho correlation coefficient rxy = +0.533, p < 0.000).
technologies (see Section 4.1.1.4 for more details). Tables 2a and 2b
show the number of papers classifying the positive and negative 4.1.1.5. Smart homes and home health-monitoring
outcomes per medical condition and disability addressed (n = 18), technologies—usability and technology acceptance studies. Of the
also, details about the total number of participants and study 48 studies included in this literature review, 66.67% of them (32
design type are provided. papers) aimed to study the usability and technology acceptance of
smart homes and home health-monitoring technologies. Of these
4.1.1.4. Smart homes and home health-monitoring 32 papers, 21 (65.63%, 21/32) examined older adults’ acceptance
technologies—where is the evidence?. Research question 1: what of monitoring technology (see Table A.6 in Supplemental material
is the clinical evidence of the outcomes in the studies regard- for more detail). In general, after analyzing the outcomes of this
ing smart homes and home health-monitoring technologies for kind of studies one can say that acceptability and usability of these
older adults with complex needs? Of the clinically-oriented and technologies were high among older adults. For example, in one
economic-assessment studies that reported any outcome (37.5%, study aimed to make a comparative evaluation of the interfaces of
n = 18), the evidence depends on the areas of medical conditions an enSAVE Ò-Prototype (wearable system), during and after the
and disabilities addressed by smart homes and home health- development, three-quarters of the older adults said they would
monitoring technologies. Here, we present the highest level of use the service [45] (n = 22).
evidence by condition (see Table 3). Another study that aimed to characterize older adult partic-
Regarding monitoring of ADL, there was level 1b evidence from ipants’ perceived usefulness (n = 7) of in-home sensor data and
one RCT of excellent quality that demonstrated older adults main- to develop a novel visual display for sensor data from Ambient
tained physical and cognitive status, and function in their ADL and Assisted Living and Smart Homes showed that the participants’
mobility when a smart home system for functional monitoring was perceived usefulness was high [46]. As a result, the acceptability
installed in a home. of this technology was also high because it allowed older adults
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 51

Table 2a
Positive and negative outcomes per Medical condition & disability addressed clinical-oriented and economic assessment studies (n = 18).

Medical condition & disability No. of papers (%) Total no. of participants Study type Study
addressed

Negative or mixed outcomes Positive outcomes

Monitoring of Activities of Daily 0 (0.0%) 2 (11.1%) 118 Case study, RCT [26,27]
Living
Chronic obstructive pulmonary 1 (5.6%) 1 (5.6%) 99 RCT [28, 29]
disease
Cognitive decline and mental 1 (5.6%) 2 (11.1%) 173 RCT, case study, one [30 – 32]
health group pretest-posttest
Disease/disability 2 (11.1%) 3 (16.7%) 9344 RCT, observational, [33, 34, 35 – 37]
prediction/Health-related nonequivalent
quality of life pretest-posttest
control group
Fall Prevention 1 (5.6%) 2 (11.1%) 144 nonequivalent [38–40]
pretest-posttest
control group,
descriptive
Monitoring Heart conditions 1 (5.6%) 2 (11.1%) 749 RCT [41 – 43]
Total 6 (33.33%) 12 (66.66%) 10671

Notes: Monitoring Heart conditions: heart failure, control of hypertension, monitoring heart conditions.
Disability prediction and health-related quality of life: social and spatial barriers/isolation reduction, self-perception of health wellbeing, monitoring overall health status,
sleep problems, low physical activity, changes in activity patterns which can indicate emerging health problems, rate of deterioration into frailty state and death (weight
loss, weakness, exhaustion, low activity, slow gait speed) cognitive decline and mental health: depressive symptoms, chronic conditions, dementia, wandering detection,
slight cognitive decline). Paper in bold indicates negative outcomes or none significant differences were found.
RCT: randomized controlled trial.

Table 2b
Medical conditions and disabilities addressed and main outcomes.

Medical condition & disability Study Main outcome(s)

addressed

Monitoring of Activities of [26] Cost savings were achieved in all cases, and the benefits to older people and their caregivers were
Daily Living also considerable.
[27] The treatment group maintained physical and cognitive status, whereas the control group
declined significantly in both.

Chronic obstructive pulmonary [28] The patients experienced a lower rate of exacerbations and Chronic obstructive pulmonary disease
disease related hospitalizations compared to patients followed up using the standard model of care
[29] No statistically significant differences in the number of Chronic obstructive pulmonary disease
related hospital admissions, length of stay between treatment and control group

Cognitive decline and mental [30] (1) Lower emergency department usage rates for older adults, (2) improved the clinical and
health healthcare use outcomes for depressed older adults receiving homecare.
[31] (1) The system detected that patient was very active, (2) reduced the care visits, it save $2500
(pounds); and (3) reduced the felling of intrusion of caregivers
[32] No change was reported in the elders’ quality of life and daily activity abilities.

Disease/disability [33] Home telemonitoring in older adults with multiple co morbidities does not significantly improve
prediction/Health-related self-perception of mental well-being.
quality of life
[34] The use of technology to measure and track biometric data by itself does not improve or lessen the
decline in frailty status
[35] The telemonitoring system in rural agencies reduced the number of nursing visits and the overall
cost of care
[36] Automatic monitoring of movement allowed early detection of dementia
[37] Reduction in hospitalization rates among the telehealth group compared to the non-telehealth
group

Fall Prevention [38] Dynamic assessment of the variability of everyday movements, when combined with other known
risk factors for falls, can significantly improve the accuracy of fall prediction
[39] The balance testing apparatus detected balance issues and demonstrated to be an effective
method for monitoring balance and fall risk remotely
[40] (1) Both intervention and control groups experienced a reduction in fear of falling and also in fall
reduction, and (2) Improvements in safety, independence and confidence.

Monitoring Heart conditions [41] No significant difference in rates of hospitalization, emergency department visits, death between
intervention(telehealth case management) and control group (case management)
[42] Telemonitoring facilitated better ambulatory management of heart failure patients, including
fewer emergency department visits
[43] The Blood Pressure statistically significantly decrease in both groups but participants using this
device transmitted more than 30 BP readings/m

and family caregivers, to manage one’s health and activity patterns, adults, 58% of participants perceived that they had improved inde-
and identify changes in health status (sleep patterns). In another pendence with use of technology. Moreover, 61% considered that
study that examined the usability of a fall detector (n = 47) for older it had improved their safety, 72% of them felt more confident with
52 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59

Table 3
Level of Evidence Scale per Medical condition & disability addressed clinical-oriented and economic assessment studies (n = 18).

Medical condition & disability No. of papers Technology Study type PeDro Scale Level of Study
addressed readiness level Evidence Scale

Monitoring of Activity Daily 2 9 RCT 10 1b [27]

Living
9 Case study NA 5 [26]

Chronic obstructive pulmonary 2 8 RCT 7 1b [44]

disease
RCT 6 1b [29]

Cognitive decline and mental 3 9 NA NA 5 [31]

health
9 RCT 8 1b [30]
9 One group pretest-Posttest NA 4 [32]

Disease/disability 5 9 RCT 4 2 [33]

prediction/Health-related
quality of life
7 RCT 6 1b [34]
7 Non-equivalent pretest-posttest control group NA 2 [35]
6 Observational NA 5 [36]
9 Retrospective NA 3 [37]

Fall Prevention 3 8 Predictive correlational NA 4 [38]

5 Descriptive NA 5 [39]
9 Non-equivalent pretest-posttest control group NA 2 [40]

Monitoring Heart conditions 3 7 RCT 4 2 [42]

7 RCT 6 1b [41]
9 RCT 6 1b [43]

Notes: RCT: randomized controlled trial.

Level of evidence scale cell highlighted in bold means the highest level of evidence in this category.

the use of the technology, and 90% of them were pleased that they tance was quantitative, and they were conducted with low sample
had a fall detector [40]. One study assessed the perceptions and sizes. Further, none of them was based on theories that helped to
expectations of seniors concerning technology installed and oper- explain intention to use and the usage behaviors of participants.
ated in their homes (n = 15). Participants stated that they accepted
the smart home because they had the perception that these types 4.1.1.6. Smart homes and home health-monitoring technologies: tar-
of technologies would: (1) benefit them in an emergency, (2) pro- geted users and environments. In 75% of the papers, the smart homes
vide assistance with hearing and visual impairment, (3) prevent and and home health-monitoring technologies were oriented toward
detect falls, (4) monitor physiological parameters (e.g., blood pres- satisfying individuals’ needs, in other words, to support physi-
sure, glucose levels), (5) provide safety control, (6) ensure property cal or mental health, and to maintain independence and quality
security (i.e., intruder alarm), (7) announce upcoming appoint- of life in their own living environment. In the remaining 25%
ments or events, and (8) provide timely and accurate information of the papers, the technologies were oriented toward support-
on adverse drug events and contraindications [47]. ing individuals and either formal, informal or family caregivers.
Participants in usability studies highlighted barriers that hin- With respect to targeting users, in 87.5% of the papers, the smart
dered the use of smart homes and home health-monitoring homes and home health-monitoring technologies were installed or
technologies. For example: (1) participants stated that they were tested in either home, private dwellings or independent retirement
reluctant to accept the smart homes and home health-monitoring facilities, whereas only 6.25% of the smart homes and home health-
technologies if these technologies did not allow them to remain monitoring technologies were installed or tested in assisted living
in their own homes and to age in place; (2) participants would facilities (see Tables A.7 and A.8 in Supplemental material for more
only accept technologies if they improved their quality of life (i.e., details).
high perception of usefulness) [47,48]; (3) in some studies partici- We found a statistically significant and moderate positive corre-
pants felt anxious either due to loss of privacy (i.e., possible privacy lation between targeted environments and targeted users variables
violation resulting from the use of camera) [49] or due to the risk (Spearman rho correlation coefficient rxy = +0.301, p < 0.043). This
that their information would fall to people or organizations with reinforced the older adults’ primary criteria about the use of smart
no authorization to use it [50]; (4) some technical problems per- homes and home health-monitoring technologies, i.e., participants
sist, such as cohabitation or multiple individuals residing in a single in the analyzed papers stated that they accepted smart homes and
dwelling can result in sensors failing to correctly detect daily activ- home health-monitoring technologies only if these allowed to them
ities [12], buttons on the touch screen too small, incorrect feedback to remain in their own homes and to age in place.
messages [51], level of false alerts can cause distress to users [40],
the low-power wireless connection of low cost devices makes a 4.1.1.7. Smart homes and home health-monitoring
system unreliable during long-term home usage [52]; (5) technolo- technologies—medical condition/disability addressed. Table 4
gies under test were in their initial development stages and did not and Fig. 3 show the type of medical condition and disability
monitor or take into account all the aspects participants felt as very addressed by the proposed technology. From this table, one
important for them [53]; (6) users’ personal characteristics, includ- can see that Monitoring of Activities of Daily Living and, Dis-
ing age, gender, cognitive abilities and personality traits, influenced ease/disability prediction/Health-related quality of life categories
users’ success with technologies under study [54]; and, (7) par- are associated with the highest number of papers, i.e., 27.08%
ticipants considered some technologies as obtrusive systems [45]. and 22.92% respectively. Fall prevention (16.67%), and Cognitive
Notably, none of these studies on usability and technology accep- decline and mental health (12.50%) are the two others important
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 53

Table 4
Condition/behaviors addressed by technologies (n = 48).

Medical condition & disability addressed No. of papers (%) Total number of Total number of Study
participants participants
Mean (SD)

Monitoring of Activities of Daily Living 13 (27.08%) 277 17.46 (29.6) [26,27,49,51,55–63]

Disease/disability prediction/Health-related quality of life 11 (22.92%) 9527 866 (1893.16) [29,33–37,52,53,64,65,66]
Fall prevention 8 (16.67%) 269 33.63 (26.72) [38,39,40,46,67,68,69,70]
Cognitive decline and mental health 6 (12.50%) 312 52 (50.77) [30–32,50,71,72]
Monitoring Heart conditions 4 (8.33%) 771 192.75 (187.68) [41–43,45]
Diabetes Type II 2 (4.17%) 38 19 (12.72) [54,73]
Do not report 2 (4.17%) 38 19 (5.65) [47,48]
Chronic obstructive pulmonary disease 2 (4.17%) 143 99 (38.89) [28,44]

Notes: monitoring heart conditions (i.e. heart failure, control of hypertension, monitoring heart conditions).
Disability prediction and health-related quality of life: social and spatial barriers/isolation reduction, self-perception of health well-being, monitoring overall health status,
sleep problems, low physical activity, changes in activity patterns which can indicate emerging health problems, rate of deterioration into frailty state and death (weight
loss, weakness, exhaustion, low activity, slow gait speed)).
Cognitive decline and mental health: depressive symptoms, chronic conditions, dementia, wandering detection, slight cognitive decline).

Fig. 3. Number of studies and the total number of participants per condition/behaviors addressed by in the areas of smart homes and home health monitoring technologies
(n = 48).

in laboratory and validated in relevant environments (TRL5);

or the technologies were in demonstration or pilot phase, and
prototypical system validation in relevant environments (TRL6).
The remainder 44% of smart homes and home health-monitoring
technologies in the analyzed studies were mature technologies in
which the actual system operates over the full range of expected
mission conditions (TRL7, 8, and 9). Fig. 5 shows the technology
readiness level versus number of participants per condition or
behavior addressed. From this figure, one can reach the conclusion
that the types of conditions or behaviors addressed by smart homes
and home health-monitoring technologies that have greater levels
Fig. 4. Technology readiness level (n = 48) [23]. of technology readiness were disease/disability prediction/health-
related quality of life, chronic obstructive pulmonary disease, and
medical condition and disabilities to which smart homes and home monitoring of heart conditions.
health-monitoring technologies have been applied. In general, we found a statistically significant positive cor-
relation between level of technology readiness and number of
4.1.2. Features of smart homes and home health-monitoring participants in the experiments conducted (Spearman rho cor-
technologies relation coefficient rxy = +0.439, p < 0.002). In addition, we found
4.1.2.1. Smart homes and home health-monitoring a statistically significant positive association between the level
technologies—technology readiness level (TRL). Fig. 4 shows the of technology readiness (8 and 9) and the study type, i.e., ran-
level of technology readiness of the papers included in our analysis domized controlled trials (Pearson ␹2 = 5.998, p < 0.014, df = 1, Phi
according to the scale of United States Department of Energy [23]. association coefficient = +0.451). In other words, randomized con-
For research question 2, we asked, “What is the level of technology trolled trials were conducted to assess technologies in the highest
readiness for smart homes and home health-monitoring technolo- technology-readiness levels (8 and 9). We also find a significant
gies for older adults with complex needs?” One can say that 56% positive association between the level of technology readiness (6)
of the studies were still either in development and testing phases and the study type, i.e., technology feasibility (Pearson ␹2 = 11.357,
54 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59

Fig. 5. Technology readiness versus number of participants per condition/behaviors addressed by in the areas of smart homes and home health monitoring technologies
(n = 48).

p < 0.001, df = 1, Phi association coefficient = +0.621). In other words, blood pressure cuff, stethoscope, pneumotachograph, electrocar-
the technology-feasibility studies were conducted when technol- diogram, thermometer, scales) were technologies most frequently
ogy was implemented in pilot studies at small-scale trials. used in the acquisition layer in smart homes and home health-
monitoring technologies. These two modalities of technology type
4.1.2.2. Smart homes and home health-monitoring used in the acquisition layer in smart homes and home health-
technologies—application. The most and the least used appli- monitoring technologies accounted for 66.66% of studies analyzed.
cations of smart homes and home health-monitoring technologies A total of 7 papers (14.58%) reported the use of cameras and video
were physiological monitoring (Phys) and cognitive and sensory in the acquisition layer for monitoring the special needs in older
assistance (Sen), presented in 41.7% and 2% of the studies analyzed adults. A distinctive aspect of this literature review is the use of
respectively. In addition, 61% of studies analyzed had only one type mobile technologies in combination with additional devices (i.e.,
of application (i.e., functional monitoring/emergency detection and pulse-oxymeter, blood pressure cuff, stethoscope, pneumotacho-
response component (Fx, 34%), physiological monitoring (Phys, graph, electrocardiogram, thermometer, scales), as technology type
23%), safety monitoring and assistance (Saf, 2%), social interaction used in the acquisition layer. However, mobile technologies in com-
monitoring and assistance (Soc, 2%)); and 39% of studies analyzed bination with these additional devices were used only in 8.3% (4
examined more than one type of application. Only 8.0% of the papers) of the studies analyzed. This seems to be a contradiction
studies had 3 or more applications: physiological monitoring, func- in an era when mobile technologies are ubiquitous. Regarding the
tional monitoring/emergency detection and response component, type of software used in the application layer to analyze and process
and safety monitoring and assistance (Phys, Fx, Saf, 4%); physio- the data gathered, it was found that 43.75% of papers did not report
logical monitoring, functional monitoring or emergency detection the specific type of software used in the application layer, 14.6% of
and response component, and safety monitoring and assistance papers used web applications through broadband internet connec-
(Phys, Fx, Saf, Sec, 2%); and physiological monitoring, functional tion for administration and data access, and the remaining 31.75%
monitoring/emergency detection and response component, safety used custom-made software for administration and data access. Of
monitoring and assistance, social interaction monitoring and the papers under study, 41.6% reported that smart homes and home
assistance, and cognitive and sensory assistance (Phys, Fx, Saf, health-monitoring technologies used internet for external trans-
Soc, Cog/Sen, 2%) (see Fig. A.6 in Supplemental material for more mission of data (27.1% used solely internet, 13/48) and, internet in
details). combination with either Bluetooth or Zigbee wireless platform for
data transmission between the physical and the application layer
4.1.2.3. Smart homes and home health-monitoring (i.e., 14.5%, 7/48) (see Table A.9 in supplemental material for more
technologies—technology type used. In examining the technologies detail).
used in smart homes and home health-monitoring technologies, it
is worth examining the technologies used in the acquisition layer, 4.1.2.4. Approach/policy taken towards ethics and privacy of the indi-
in the application layer (to analyze and process the data gathered), vidual. Table 5 shows the approach or policy taken to guarantee
and the types of methods used to communicate the data. privacy of an individual either in communication or storage of infor-
Stand-alone devices and computers either connected to a set of mation processes. Almost a quarter of the papers reported any kind
sensors or combined with additional devices (i.e., pulse-oxymeter, of approach or policy taken to guarantee privacy of the individuals
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 55

Table 5
Approach/policy taken towards ethics and privacy of the individual that use smart homes and home health monitoring technologies.

Ethics and privacy No. of papers (%) Technology Study

Readiness (mode)

Not reported 35 (72.92%) 6 [26–29,31–33,35–42,45–49,51–54,56,57,63,64,66–68,69,70,72,73]

Storage of data 5 (10.42%) 6 [34,43,44,58,59]
Communication + Storage 4 (8.33%) 6 [50,55,62,65]
Communication, transmission of data 4 (8.33%) 6 [30,60,61,71]

that used smart homes and home health-monitoring technologies. most common technology readiness level was 6, which means that
The most common approaches were: (1) removal of user iden- technologies studied were implemented at a level that was located
tity and data encryption using unsupervised learning (Continuous in the bottom or in the lowest level of the last third (the last third
Varied Order Multi Threshold Method (COM)) [55,65,58]; (2) in goes from level 5–9) of the technology readiness level scale [23].
video-based fall detection, sensor algorithms were implemented The low number of studies included (48) in comparison with the
to identify persons’ patterns in the image that extract a silhou- initial potential sample of studies after duplicate removal (1863),
ette of the person instead of using the real image [59]; and (3) revealed that although a large number of papers mentioned either
web-site interfaces provided a secure interface to family members in their titles or abstract the words “older adults”, many of them
and individuals [43,62] (i.e., secure HTTPS protocol compliant login included younger participants in the respective studies. In fact, out
structure, according to Health Insurance Portability and Account- of the 356 full papers that were analyzed after the abstract filter,
ability Act of 1996—HIPAA). 69 were excluded because the participants were adults younger
than 60 years of age. As a result, we want to issue a wake-up
call on two aspects. First, when researchers analyze the scientific
5. Discussion evidence of the outcomes in the application of smart homes and
home health-monitoring technologies in older adults, they verify
The aim of this systematic review was to examine evidence whether scholars agree on what age criteria defines “older adults”.
for home health technologies that support aging in place, and the Second, despite the relevance of this topic, there actually exists lit-
level of technology readiness. To this end, we included 48 studies tle research on the measurement of the impact of technologies on
(out of 1863) from 2010 to October 2014. Specifically, we wanted older adults.
to identify and characterize the best available evidence about In our literature review, we excluded 3.1% of the studies because,
the implementation of smart homes and home health-monitoring after reading the papers, we did not have enough information (e.g.,
technologies and to respond to two research questions: (1) what is description of participants, technology readiness) to categorize the
the clinical evidence of the outcomes in the studies regarding smart papers. Despite efforts in the standardization of good practices in
homes and home health-monitoring technologies for older adults reporting scientific papers in medical informatics such as [74,75],
with complex needs?; and (2) what is the level of technology readi- lack of quality in reports persists. This implies that the editors of
ness for smart homes and home health-monitoring technologies for journals indexed and conference proceedings reviewers must con-
older adults with complex needs? tinue to improve the peer-review processes.
To answer the first research question, we found that smart We found in our literature review a considerable number of
homes and home health-monitoring technologies have been used studies (66.67%) dedicated to understanding how users accept
to address several medical conditions and disabilities. The highest home health technologies. In general, after analyzing the outcomes
level of evidence was 1b in regard to smart homes and home health- of these studies, one can say that the acceptability and usability
monitoring technologies used to monitor ADL, cognitive decline of these technologies was high among older adults. Older adults
and mental health, and heart conditions in older adults with com- highlighted that the main deciding factors, on whether to use or
plex needs. The evidence supports that home health-monitoring accept technologies, were that these technologies must allow them
technologies for cognitive decline and mental health reduce symp- to remain in their own homes and to age in place, must improve
toms of depression and visits to the emergency department in older their quality of life, and there must be a high perception of use-
adults with chronic illness. There is evidence to support that tech- fulness. Privacy is a major concern that hinders the adoption and
nologies for monitoring heart conditions improve patients’ sharing the use of home health technologies, e.g., possible privacy violation
data with clinicians and their blood pressure control. There is resulting from the use of cameras. Major privacy concerns reported
no evidence that smart homes or home health-monitoring tech- in these studies can be divided into two categories: (a) privacy of
nologies help to address the conditions of disease or disability the occupant, and (b) security and privacy of the collected data. The
prediction, health-related quality of life or fall prevention. Finally, privacy of the occupant should be considered both in details of col-
the evidence that smart homes or home health-monitoring tech- lected data and type of devices and sensors. Traditionally, there are
nologies help to address the conditions of COPD is conflicting. controversial privacy concerns when cameras are used to collect
To answer the second research question, we can say that data. Therefore, many home health technologies should consider
level of technology readiness for smart homes and home health- this issue in their design of data-capture methods and avoid using
monitoring technologies is still low. We can assert this based on two cameras or index frames for privacy protection and ethical rea-
facts. First, after removing the duplicated papers we had an initial sons (e.g., capturing silhouettes instead of an image that allows the
potential sample of 1863 studies of which we had to exclude 591 identification of the client).
papers (one third or 31.72%) because the studied technology(ies) None of the studies about usability of home health technolo-
were not in at least pilot phase. This means that one third of the stud- gies included in this review used theories that helped to explain
ies that investigated smart homes and home health-monitoring the intention to use and the usage behaviors in participants (e.g.,
technologies were conducted in artificial environments such as Technology Acceptance Model (TAM-TAM2) [76], and the Unified
laboratories or academic institutions. In other words, these stud- Theory of Acceptance and Use of Technology (UTAUT) [77]). This is
ies were in the stage of proof-of-concept or even cases of design consistent with what was found in [78] where most of the papers
projects that propose but not implement a system even in labora- in their literature review about acceptance of technology for aging
tory settings. Second, of the 48 papers in our literature review, the
56 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59

in place by older adults lacked a theoretical approach. A theoretical Most of the papers were journal articles. Articles published in
framework in usability studies would allow researchers to achieve conference proceedings typically reported partial results, there-
a better understanding about the reasons older adults accept or fore, it was difficult to find in conference proceedings sufficiently
reject home health technologies. detailed studies of smart homes and home health-monitoring tech-
We found that, in 75% of the papers, the smart homes and home nologies with technology readiness level higher than 5. Thus,
health-monitoring technologies were oriented toward individuals, papers of, or below, technology readiness level 5 were excluded
i.e., older adults. This means that designers of smart homes and during the selection process of this review. It seemed that results
home health-monitoring technologies were taking into account the on proof-of-concept research (≤level 5) were published mainly as
main claim of older adults: they stated they were willing to accept conference papers, while studies reporting clinical outcomes and
the smart homes and home health-monitoring technologies if these usability results were published mainly as journal papers.
allowed them to remain in their own homes and to age in place. As The origin of papers, according to the location of the authors, was
well, participants accepted technology if it was perceived to be use- 100% from developed countries. This is not surprising given the high
ful and improved their quality of life [47,48]. We found that in 25% cost of investigations on smart homes and home health-monitoring
of papers, smart homes and home health-monitoring technologies technologies, which prevents researchers from conducting such
were oriented toward supporting individuals and formal or infor- investigations in developing countries. Another possible explana-
mal caregivers, in contrast to other literature reviews that found tion for the absence of authors from developing countries may be
few papers included caregivers as targeted users of smart homes that the health systems of these countries have other priorities:
and home health-monitoring technologies (e.g., Ref. [12]). Design- (1) inequity in access, (2) the cost coverage, and the (3) quality
ers and developers of smart homes and home health-monitoring of the health care systems. Thus, the resources of the systems in
technologies appeared to understand that involving family mem- those contexts may not pay for these types of technologies or ser-
bers of older adults and other stakeholders is important for the vices that, in turn, affects the progress of research in this field.
success of such technologies, because their involvement allows Other possible explanations may be sound intact family structures
older adults with cognitive and functional limitations to live in the that reduce need for professional or automatized services, or life
community [12]. expectancies lower than 60 years of age in some developing coun-
In our literature review, we found that only 2.08% of the included tries which make smart homes and health-monitoring technologies
papers incorporated any sort of economic analysis. As evidence not a research priority.
of cost-effectiveness is the most persuasive argument to moti-
vate healthcare decision makers to implement at larger scale 5.1. Gaps and implications for future research
the home health technologies, more studies on cost-effectiveness
analysis are needed. Other literature reviews assert that one • There is no evidence that smart homes or home health-
of the factors leading to limited adoption of smart homes and monitoring technologies help to address the conditions of
home health-monitoring technologies are the lack of information disability prediction and health-related quality of life or fall
related to technology costs and sustainable reimbursement mod- prevention, and there is conflicting evidence that home health
els [12]. This can explain why we found a low level of technology technologies help to address the conditions of COPD. More RCTs
readiness for smart home and home health-monitoring technolo- on these topics would provide evidence to determine whether
gies. home health technologies are clinically effective to address these
We found a statistically significant positive association between medical conditions. The conflicting evidence in the benefits of
the level of technology readiness (levels 8 and 9) and RCT study home health technologies for patients with COPD may be due to
type, and a positive association between the level of technology the field of telemonitoring interventions for patients with COPD
readiness (level 6) and the “technology feasibility” study type. This is relatively new and presents some usability problems as Cruz
is expected because technology-feasibility studies are conducted and colleagues [79] pointed out.
when a given technology is implemented in pilot studies on a small • None of the studies about usability of home health technologies
scale (i.e., technology readiness level 6), whereas RCTs are typically included in this review used theories for explaining the intention
conducted to respond to more complex questions such as whether to use and the usage behaviors in participants. Usability studies
or not a given technology is effective or sustainable. This kind of should use a theoretical framework to explain the main determi-
question can be answered only when technologies achieve mature nants of smart homes and home health-monitoring technologies
stages of implementation (technology readiness level 8 and 9). In adoption. In addition, as most usability studies of home health
addition, we found a statistically significant positive correlation technologies used qualitative approaches, an increase in quan-
between level of technology readiness and number of participants titative research approaches in this topic will help to produce
in the studies (Spearman rho correlation coefficient rxy = +0.439, evidence on the adoption of home health technologies by users.
p < 0.0.002). It is expected that researchers want to demonstrate • Typically, the length of most studies was about one year (mode
clinical evidence of smart home and home health-monitoring tech- 1 year, and mean 0.82 years, SD 0.92 years). More longitudinal
nologies with higher levels of implementation, therefore, they want studies are required to gain an understanding of the effectiveness
to have larger sample sizes with randomized controlled trials. and sustainability of smart homes and home health-monitoring
Surprisingly, we found that in an era of ubiquitous mobile tech- technologies on the functional status of older adults over several
nologies, only four papers showed the use of smart mobile phones years.
in combination with smart homes and home health-monitoring • Economic assessment studies regarding the cost-effectiveness of
technologies for older adults. Computers such as desktop or lap- smart homes and home health-monitoring technologies are still
top, and video communication, in combination with sensors or relatively rare (2.08%). More studies of this type are essential
other devices were the most common devices used to implement to provide high quality evidence of cost-effectiveness of smart
the smart homes and home health-monitoring technologies solu- homes and home health-monitoring technologies to guarantee a
tions. Possibly, smart mobile technologies are more difficult to use widespread adoption.
and less accepted by older adults. Although there were few studies • The origin of the included papers was 100% from developed coun-
that reported the costs of implementation and development, low- tries, therefore, it is necessary to promote the development and
cost, readily available technologies including a combination of the research of smart homes and home health-monitoring technolo-
Internet, Zigbee, and Bluetooth were increasingly used. gies in developing countries because the aging of the population
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 57

in developing countries [80] is becoming a demographic concern

as it is in developed countries. Summary points
• There was a small number of studies that implemented low-cost What is already known on the topic?
technologies such as a combination of Internet plus Zigbee wire-
less platform and Bluetooth. If research and development of smart
homes and home health-monitoring technologies continue using • In many countries of the world populations are aging at an
these low-cost technologies in clinical studies in real settings, increasing rate.
• The proportion of persons aged 60 will reach 2 billion by
they can achieve higher levels of technology readiness, i.e., TRL
2050.
5–6 or higher. • “Gerontechnology” is a term that combines gerontology and
technology.
5.2. Study limitations
What this study adds to our knowledge?
Our study suffers from some limitations. First, we acknowledge
that we were unable to perform a meta-analysis because of the het- • Technology readiness for smart home and health monitoring
erogeneity of the technologies used, the medical conditions and technologies is still low.
disability addressed with those technologies, and the outcomes • There is a high level of evidence for monitoring function,
reported in the studies included and reviewed. Second, there were cognitive and mental health.
a considerable number of parameters that could not be examined • There is minimal evidence technology predicts disability, pre-
because the papers did not report them. Third, we were only able vents falls, helps quality of life.
• There is conflicting evidence home health technologies help
to assess quantitatively the strength of studies that used RCT due
conditions of chronic obstructive pulmonary disease (COPD).
to the absence of standardized scales that determine the quality of
either qualitative or quantitative non-RCT studies. Although there
are guidelines, tools and checklists available for performing a criti-
cal appraisal of research literature, the result was a proxy measure
of quality; without a scale, comparison of the relative quality of mented at pilot levels (TRL 5–6). (Pearson ␹2 = 11.357, p < 0.001
papers was not possible. Fourth, despite our use of a thorough df = 1, Phi association coefficient = +0.621).
• Although the total number of papers analyzed was small com-
search strategy, some empirical studies on smart homes and home
health-monitoring technologies may not have been identified for pared to the total sample after duplicates analysis (i.e., 1863),
this systematic review (e.g., gray literature such as unpublished there is a clear upward trend in the number of published papers
documents and reports), because we only included papers that over the years (␤=+0.48, R2 = 0.64, p < 0.000).
• Economic-assessment studies regarding the cost-effectiveness of
were published through the peer-review process.
smart home and home-based technologies are still rare.

6. Conclusions Author contributions

• The level of technology readiness for smart homes and home LL, ES, and IN contributed to the conception of the study and to
health-monitoring technologies is still low. the text of the paper. AMRR and AMC contributed to the conception
• The highest level of evidence was 1b when considering whether and text of the paper and performed the literature search.
home health technologies can help to address monitoring of ADL,
cognitive decline and metal health, and heart conditions in older
adults with complex needs. Conflict of interest
• There is no evidence that home health technologies help address
the conditions of disease or disability prediction and health- The authors have no conflicts of interest to declare.
related quality of life or fall prevention.
• The evidence that home health technologies help to address the Acknowledgments
conditions of COPD is conflicting.
• The higher the level of evidence of experiments conducted in This work was supported, in part, through a grant by the Alberta
home health technologies, the higher the technology readiness Addiction and Mental Health Research Partnership Program, Col-
level of these smart homes and home health-monitoring tech- laborative Research Grant Initiative: Mental Wellness in Seniors
nologies (i.e., Spearman rho correlation coefficient rxy = +0.533, and Persons with Disability in 2010–2011. Corinne Schalm, Beth
p < 0.000). Wilkey, Suzanne Maisey, Carmen Grabusic and Sharla King pro-
• The smart homes and home health-monitoring technologies vided early insights in formulating the scope of the project. Angela
were oriented toward being installed or tested in either home, Sekulic, Katie Woo, Ran Ran Zhang and Koosha Golmohammadi
private dwellings or independent retirement facilities to be were research assistants who conducted an initial literature review
used directly for older users rather than for family caregivers on this subject. The authors thank Lindsay Hoehne, Jacquelyn Lar-
(rxy = +0.301, p < 0.043). den, for their assistance with the literature search and papers’
• The higher the number of participants in the studies conducted analyses. Finally, we are grateful to the reviewers for their con-
in smart homes and home health-monitoring technologies, the structive feedback.
higher the level of technology readiness (rxy = +0.439, p < 0.0.002).
• RCTs were conducted using smart homes and health-monitoring
technologies of the highest technologies readiness levels (8 and Appendix A. Supplementary data
9). (Pearson ␹2 = 5.998, p < 0.014 df = 1, Phi association coeffi-
cient = +0.451). Supplementary data associated with this article can be found,
• Technology feasibility studies were conducted when smart in the online version, at http://dx.doi.org/10.1016/j.ijmedinf.2016.
homes and home health-monitoring technologies are imple- 04.007.
58 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59

References education and activation of mood (I-TEAM) study, J. Am. Geriatr. Soc. 62 (5)
(2014) 889–895.
[1] UNDP, Human Development Report 2013. The Rise of the South: Human [31] C. Price, Monitoring people with dementia—controlling or liberating? Qual.
Ageing Older Adults 8 (3) (2007) 41–44.
Progress in a Diverse World, Published for the United Nations Development [32] C. Vincent, D. Reinharz, I. Deaudelin, M. Garceau, L. Talbot, Public
Programme (UNDP), UN Plaza, New York, NY, 2013. telesurveillance service for frail elderly living at home, outcomes and cost
[2] UN, “UN Global Issue: Ageing,2̈0 May 2014 [Online]. Available: http://www. evolution: a quasi experimental design with two follow-ups, Health Qual. Life
un.org/en/globalissues/ageing/. Outcomes (2006) 4–41.
[3] T. Mattimore, N. Wenger, N. Desbiens, J.M. Teno, M. Hamel, H. Liu, R. Califf, [33] J.L. Pecina, G.J. Hanson, H.V. Houten, P.Y. Takahashi, Impact of telemonitoring
A.F. Connors Jr., L. Lynn, R. Oye, Surrogate and physician understanding of on older adults health-related quality of life: the Tele-ERA study, Qual. Life
patients’ preferences for living permanently in a nursing home, J. Am. Geriatr. Res. 22 (9) (2013) 2315–2321.
Soc. 45 (1997) 818–824. [34] B. Uptasing, G.J. Hanson, Y.L. Kim, S.S. Cha, Y. Yih, P.Y. Takahashi, Effects of
[4] D. Cook, Health monitoring and assistance to support aging in place, J. home telemonitoring on transitions between frailty states and death for older
Ubiquitous Comput. Syst. 12 (2006) 15–29. adults: a randomized control trial, Int. J. Gen. Med. 6 (2013) 145–151.
[5] D. Hickam, J. Weiss, J. Guise, D. Buckley, M. Motu’apuaka, E. Graham, N. [35] D. Hansen, A. Golbeck, K. Lee, V. Noblitt, J. Pinsonneault, J. Christner, Cost
Wasson, S. Saha, Outpatient Case Management for Adults with Medical Illness factors in implementing telemonitoring programs in rural home health
and Complex Care Needs, Agency for Healthcare Research and Quality (US), agencies, Home Healthc. Nurse 29 (6) (2011) 375–382.
Rockville, 2013. [36] W. Kearns, V. Nams, J. Fozard, Tortuosity in movement paths is related to
[6] V. Frisardi, B. Imbimbo, Gerontechnology for demented patients: smart cognitive impairment, Methods Inf. Med. 49 (6) (2012) 592–598.
homes for smart aging, J. Alzheimer’s Dis. 23 (2011) 143–146. [37] H. Chen, C. Kalish, J. Pagan, Telehealth and hospitalizations for medicare home
[7] N. Archer, K. Keshavjee, C. Demers, R. Lee, Online self-management healthcare patients, Am. J. Manag. Care 17 (6) (2011) 224–230.
interventions for chronically ill patients: cognitive impairment and [38] W.D. Kearns, J.L. Fozard, M. Becker, J.M. Jasiewicz, Path tortuosity in everday
technology issues, Int. J. Med. Inf. 83 (2014) 264–272. movements of elderly persons increases fall prediction beyond knowledge of
[8] K. Chen, A. Chan, Predictors of gerontechnology acceptance by older Hong fall history, medication use, and standardized gait and balance assessments, J.
Kong Chinese, Technovation 34 (2014) 126–135. Am. Med. Dir. Assoc. 13 (7) (2012) 665.e7–665.e13.
[9] ISG, International Society for Gerontechnology, 2014 [Online]. Available: [39] Z. Matjacic, K. Bohinc, I. Cikajlo, Development of an objective balance
http://www.gerontechnology.info/index.php/journal/pages/view/isghome assessment method for purposes of telemonitoring and telerehabilitation in
(accessed 22.05.14). elderly population, Disabil. Rehabil. 32 (3) (2010) 259–266.
[10] G. Demiris, B. Hensel, Technologies for an aging society: a systematic review [40] S. Brownsell, H. Hawley, Fall detectors: do they work or reduce the fear of
of smart home applications, Yearb. Med. Inform. 2008 (2008) 33–40. falling? Hous. Care Support 7 (1) (2004) 18–24.
[11] L. Liu, Collaborative Research Grant Initiative: Mental Wellness in Seniors and [41] M. Wade, A. Desai, C. Spettell, A. Snyder, V. McGowan-Stackwicz, P. Kummer,
Persons with Disabilities Seed/Bridge Fund Final Report, Alberta Addiction M. Maccoy, R. Krakauer, Telemonitoring with case management for seniors
and Mental Health Research Partnership Program, Edmonton, 2012. with heart failure, Am. J. Manag. Care 17 (3) (2011) 71–79.
[12] B. Reeder, E. Meyer, A. Lazar, S. Chaudhuri, H. Thompson, G. Demiris, Framing [42] C. Tompkins, J. Orwat, A randomized trial of telemonitoring heart failure
the evidence for health smart homes and home-based consumer health patients, J. Healthc. Manag. 55 (5) (2010) 312–322.
technologies as a public health intervention for independent aging: a [43] D. Rifkin, J. Abdelmalek, C. Miracle, C. Low, R. Barsotti, P. Rios, C. Stepnowsky,
systematic review, Int. J. Med. Inf. 82 (2013) 565–579. Z. Agha, Linking clinic and home: a randomized, controlled clinical
[13] S. Helal, W. Mann, H. El-Zabadani, J. King, Y. Kaddoura, E. Jansen, The gator effectiveness trial of real-time, wireless blood pressure monitoring for older
tech smart house: a programmable pervasive space, Computer 38 (2005) patients with kidney disease and hypertension, Blood Press. Monit. 18 (1)
50–60. (2013) 8–15.
[14] G. Abowd, A. Bobick, I. Essa, E. Mynat, W. Rogers, The Aware Home: A Living [44] C. Sicotte, G. Paré, S. Morin, J. Potvin, M. Moreault, Effects of home
Laboratory for Technologies for Successful Aging, Automation as Caregiver, telemonitoring to support improved care for chronic obstructive pulmonary
American Association for Ambient Intelligence (AAAI), 2002. diseases, Telemed. J. E Health 17 (2) (2011) 95–103.
[15] E. Stroulia, D. Chodos, N. Boers, H. Jianzhao, P. Gburzynski, I. Nikolaidis, [45] Lorenz, R. Oppermann, Mobile health monitoring for the elderly: designing
Software engineering for health education and care delivery systems: the for diversity, Pervasive Mob. Comput. 5 (2009) 478–495.
smart condo project, in SEHC ‘09: Proceedings of the 2009 ICSE Workshop on [46] B. Reeder, J. Chung, T. Le, H. Thompson, G. Demiris, Assessing older adults’
Software Engineering in Health Care (2009). perceptions of sensor data and designing visual displays for ambient
[16] S. Das, D. Cook, Designing Smart Environments: A Paradigm Based on environments, Methods Inf. Med. 53 (3) (2014) 152–159.
Learning and Prediction. Pattern Recognition and Machine Intelligence, [47] A. Mihailidis, A. Cockburn, C. Longley, J. Boger, The acceptability of home
Springer, Berlin/Heidelberg, 2005, pp.80–90. monitoring technology among community-dwelling older adults and baby
[17] M. Morris, B. Adair, E. Ozanne, W. Kurowski, K. Miller, A. Pearce, N. boomers, Assist. Technol. 20 (2008) 1–12.
Santamaria, M. Long, C. Ventura, C. Said, Smart technologies to enhance social [48] K. Wild, L. Boise, J. Lundell, A. Foucek, Unobtrusive in-home monitoring of
connectedness in older people who live at home’, Aust. J. Ageing (2014) 1–11. cognitive and physical health: reactions and perceptions of older adults, J.
[18] K. Dickersin, R. Scherer, C. Lefebvre, Identifying relevant studies for Appl. Gerontol. 27 (2) (2008).
systematic reviews, BMJ 309 (6964) (1994) 1286–1291. [49] G. Demiris, M. Rantz, M. Aud, D. Mark, H. Tyrer, A. Hussam, Older adults’
[19] R. Scherer, K. Dickersin, P. Langenberg, Full publication of results initially attitudes towards and perceptions of ‘smart home’ technologies: a pilot study,
presented in abstracts. A meta-analysis, JAMA 272 (2) (1994) 158–162. Med. Inf. Internet Med. 29 (2) (2004) 87–94.
[20] R. Light, D. Pillemer, Summing Up: The Science of Reviewing Research, [50] L. Boise, K. Wild, N. Mattek, M. Ruhl, H. Dodge, J. Kaye, Willingness of older
Harvard University Press, Cambridge, 1984. adults to share data and privacy concerns after exposure to unobtrusive home
[21] L. Portney, M. Watkins, Foundations of Clinical Research Applications to monitoring, Gerontechnology 11 (3) (2013) 428–435.
Practice, third ed., Pearson Prentice Hall, Toronto, 2009, p. 891. [51] J. Vermeulen, J. Neyens, M. Spreeuwenberg, E. Rossum, W. Sipers, H. Habets,
[22] R. Teasell, S. Marshall, N. Cullen, M. Bayley, L. Rees, M. Weiser, P. Welch-West, D. Hewson, L. Witte, User-centered development and testing of a monitoring
C. Ferri, J.-A. Aubut, Evidence-based review of moderate to severe acquired system that provides feedback regarding physical functioning to elderly
brain injury. Executive summary, December 2013. [Online]. Available: http:// people, Patient Preference Adherence 7 (2013) 843–854.
www.abiebr.com (accessed 29.07.14). [52] J. Merilahti, J. Pärkkä, K. Antila, P. Paavilainen, E. Mattil, E.-J. Malm, A.
[23] DOE, Department of Energy Technology Readiness, Assessment Guide, Office Saarinen, I. Korhonen, Compliance and technical feasibility of long-term
of Management. Department of Energy, U.S., Washginton, DC, 2009. health monitoring with wearable and ambient technologies, J. Telemed.
[24] K. Dickersin, Y. Min, C. Meinert, Factors influencing publication of research Telecare 15 (2008) 302–309.
results, JAMA (1992) 374–378. [53] J. Pecina, K. Vickers, D. Finnie, J. Hathaway, G. Hanson, P. Takahashi,
[25] SCImago, SCImago journal & country rank: un nuevo portal, dos nuevos Telemonitoring increases patient awareness of health and prompts
rankings, El profesional de la Información 16 (6) (2007) 645–646. health-related action: initial evaluation of the Tele-ERA study, Telemed. E
[26] L. Magnusson, E. Hanson, Supporting frail older people and their family carers Health 17 (6) (2011) 461–466.
at home using information and communication technology: cost analysis, J. [54] O. Blanson Henkemans, W. Rogers, A. Fisk, M. Neerincx, J. Lindenberg, C. Van
Adv. Nurs. 51 (6) (2005) 645–657. der, Usability of an adaptive computer assistant that improves self-care and
[27] R. Tomita, W. Mann, K. Stanton, A. Tomita, V. Sundar, Use of currently health literacy of older adults, Methods Inf. Med. 47 (1) (2008) 82–88.
available smart home technology by frail elders process and outcomes, Top. [55] P. Rashidi, J. Cook, COM: a method for mining and monitoring human activity
Geriatr. Rehabil. 23 (1) (2007) 24–34. patterns in home-based health monitoring systems, ACM Trans. Intell. Syst.
[28] C. Pedone, D. Chiurco, S. Scarlata, R.A. Incalzi, Efficacy of multiparametric Technol. 4 (4) (2013) 1–20.
telemonitoring on respiratory outcomes in elderly people with COPD: a [56] F. Palumbo, J. Ullberg, A. Stimec, F. Furfari, L. Karlsson, S. Coradeschi, Sensor
randomized controlled trial, BMC Health Serv. Res. 13 (1) (2013) 82–90. network infrastructure for a home care monitoring system, Sensors 14 (3)
[29] N. Antoniades, P. Rochford, J. Pretto, R. Pierce, J. Gogler, J. Steinkrug, K. Sharpe, (2014) 3833–3860.
C. Mcdonald, Pilot study of remote telemonitoring in COPD, Telemed. J. [57] A. Khan, Y. Lee, S. Lee, T. Kim, Accelerometer’s position independent physical
e-Health 18 (8) (2012) 634–640. activity recognition system for long-term activity monitoring in the elderly,
[30] Z. Gellis, B.L. Kenaley, T.T. Have, Integrated telehealth care for chronic illness Med. Biol. Eng. Comput. 48 (12) (2010) 1271–1279.
and depression in geriatric home care patients: the integrated telehealth
 L. Liu et al. / International Journal of Medical Informatics 91 (2016) 44–59 59

[58] N. Noury, M. Berenguer, H. Teyssier, M. Bouzid, M. Giordiani, Building an [70] L. Lines, K. Hone, Multiple voices, multiple choices: older adults’ evaluation of
index of activity of inhabitants from their activity on the residential electrical speech output to support independent living, Gerontechnology 5 (2) (2006)
power line, IEEE Trans. Inf. Technol. Biomed. 15 (5) (2011) 758–766. 78–91.
[59] G. Demiris, B. Hensel, M. Skubic, M. Rantz, ‘Senior residents’ perceived need of [71] S. Junnila, H. Kailanto, J. Merilahti, H.-M. Vainio, A. Vehkaoja, M. Zakrzewski,
and preferences for smart home’ sensor technologies, Int. J. Technol. Assess. H. J, Wireless, multipurpose in-home health monitoring platform: two case
Health 24 (1) (2008) 120–124. trials, IEEE Trans. Inf. Technol. Biomed. 14 (2) (2010) 447–555.
[60] G. Demiris, P. Oliver, D. Dickey, M. Skubic, M. Rantz, Findings from a [72] J. Van Hoof, H. de Kort, Unattended autonomous surveillance in
participatory evaluation of a smart home application for older adults, community-dwelling older adults: a field study, Gerontechnology 7 (2)
Technol. Health Care 16 (2) (2007) 111–118. (2008) 121.
[61] G. Demiris, D. Oliver, J. Giger, M. Skubic, M. Rantz, Older adults’ privacy [73] A.M. Lai, D. Kaufman, J. Starren, S. Shea, Evaluation of a remote training
considerations for vision based recognition methods of eldercare approach for teaching seniors to use a telehealth system, Int. J. Med. Inf. 78
applications, Technol. Health Care 17 (1) (2009) 41–48. (2009) 732–744.
[62] D. Feeney, D. Mahoney, E. Mahoney, E. Liss, AT EASE: automated technology [74] N. Keizer, J. Talmon, E. Ammenwerth, J. Brend, M. Rigby, P. Nykänen,
for elder assessment, safety, and environmental monitoring, Gerontechnology Systematic prioritization of the STARE-HI reporting items, an application to
8 (1) (2009) 11–25. short conference papers on health informatics evaluation, Methods Inf. 51 (2)
[63] E. Hanson, A. Clarke, The role of telematics in assisting family carers and frail (2012) 104–111.
older people at home, Health Soc. Care Community 8 (2) (2000) 129–137. [75] J. Talmon, E. Ammenwerth, J. Brender, N. de Keizer, P. Nykänen, M. Rigby,
[64] V. Winstead, W.A. Anderson, E.A. Yost, A.R. Cotton, A. Warr, R.W. Berkowsky, STARE-HI—statement on reporting of evaluation studies in health informatics,
You can teach an old dog new tricks: a qualitative analysis of how residents of Int. J. Med. Inf. 78 (2009) 1–9.
senior living communities may use the web to overcome spatial social [76] F.D. Davis, Perceived usefulness, perceived ease of use, and user acceptance of
barriers, J. Appl. Gerontol. 32 (5) (2012) 540–560. information technology, MIS Q. 13 (3) (1989) 319–339.
[65] M. Rantz, S. Scott, S. Miller, M. Skubic, L. Phillips, G. Alexander, R. Koopman, K. [77] V. Venkatesh, M. Morris, G.B. Davis, F.D. Davis, User acceptance of information
Musterman, B. J, Evaluation of health alerts from an early illness warning technology: toward a unified view, MIS Q. 27 (3) (2003) 425–478.
system in independent living, Comput. Inf. Nurs. 31 (6) (2013) 274–280. [78] S. Peek, E. Wouters, J. van Hoof, K. Luijkx, H. Boeije, H. Vrijhoef, Review factors
[66] A. Essen, The two facets of electronic care surveillance: an exploration of the influencing acceptance of technology for aging in place: a systematic review,
views of older people who live with monitoring devices, Soc. Sci. Med. 67 (1) Int. J. Med. Inf. 83 (2014) 235–248.
(2008) 128–136. [79] J. Cruz, D. Brooks, A. Marques, Home telemonitoring in COPD: a systematic
[67] S. Sadasivam, T. Luger, H. Coley, B. Taylor, T. Padir, C. Ritchie, H. T, review of methodologies and patients’ adherence, Int. J. Med. Inf. 83 (2014)
Robot-assisted home hazard assessment for fall prevention: a feasibility 249–263.
study, J. Telemed. Telecare 20 (1) (2014) 3–10. [80] OISS, IMSERSO, Situación, necesidades y demandas de las personas mayores
[68] S. Redmond, M. Scalzi, M. Narayana, S. Lord, S. Cerutti, L. N, Automatic en Bolivia, Colombia, Costa Rica, Ecuador y México, 2008. [Online]. Available:
segmentation of triaxial accelerometry signals for falls risk estimation, http://www.oiss.org/IMG/pdf/Situacion necesidades y demandas de las
Annual International Conference of the IEEE Engineering in Medicine and personas mayores en Bolivia Colombia Costa Rica Ecuador y Mexico
Biology (2010) 2234–2237. ampliado marzo 2008 .pdf.
[69] L. Bourke, S. Prescher, F. Koehler, V. Cionca, C. Tavares, S. Gomis, V. Garcia, J.
Nelson, Embedded fall and activity monitoring for a wearable ambient
assisted living solution for older adults, 34th Annual International Conference
of the IEEE EMBS (2012).