Keywords
Remote monitoring technology - assistive living - COVID-19 - pandemics - user requirements
- safety - public health informatics - health informatics
1 Introduction
Globally, the COrona VIrus Disease 2019 (COVID-19) has infected 23 million people
and the number of infections continues to grow [[1]–[4]]. This new, emergent infectious disease has led to significant healthcare and societal
costs as well as human suffering. From a healthcare perspective, emergency departments
and hospital units have been overwhelmed by the rapid influx of critically ill patients
with moderate to severe disease [[1]–[4]]. Yet, a significant proportion of the population infected by COVID-19 remain asymptomatic
or experience only a mild symptoms of the disease necessitating the need to self-isolate
or quarantine to prevent spread to other individuals in our communities [[5]]. Researchers are attempting to predict the development of COVID-19 symptoms to
distinguish between those that develop no, mild, moderate or severe symptoms of disease
[[5]–[7]]. Such knowledge is necessary for effective health resource planning, allocation
and utilization (i.e., deployment and reallocation of human and health technology
resources). The approach also allows for tailoring of technology interventions to
individual needs; for example, research has illuminated a relationship between COVID-19
severity and sex, race, age, the presence of underlying health conditions (i.e., diabetes,
renal disease and chronic pulmonary disease) and socio-economic status [[7]–[9]]. To date, there remains a continuum of technology interventions that could be deployed
for use in those infected with COVID-19 from sensors that allow for remote monitoring
of vital signs and oxygen values in a persons’ home through to the pulmonary ventilation
of a patient in an intensive care unit (ICU) in a hospital. There is also a need,
from a societal perspective, to isolate and quarantine those people who may be infected
with COVID-19 while they await the outcome of test results [[10]]. Lastly, there is a need to maintain population health after an individual has
tested positive for the disease, while at the same time ensuring their safety in the
home (i.e., should they develop severe symptoms requiring medical intervention and/or
immediate hospitalization) [[10], [11]]. In such cases there is a need to quarantine to contain the disease by asking individuals
to self-isolate in their homes while at the same time being able to engage with caregivers
such as family and friends as well as health professionals such as public health officers,
telehealth nurses and physicians using virtual care approaches [[11]–[13]]. In this paper we aim to describe a methodology for modelling a health technology
system to support self-isolation and remote monitoring of individuals at risk for
developing COVID-19, or moderate to severe COVID-19 disease. This research aims to
fill a gap in the literature on health technology systems requirements modelling [[14]] for remote monitoring technologies (RMTs) applied to individuals exposed to and/or
infected with COVID-19 in the published peer reviewed literature. We begin this paper
by discussing the Background and Gaps in the Research Literature.
2 Background and Gaps in the Research Literature
2 Background and Gaps in the Research Literature
2.1 Public Health Informatics and COVID-19
Public health informatics is an important area of research in the sub-field of health
or biomedical informatics. From time to time, a new infectious disease or a disease
that is currently being controlled emerges and affects significant parts of the human
population; for example, SARS-CoV-2 (severe acute respiratory syndrome coronavirus
2), MERS-CoV (Middle East respiratory syndrome-related coronavirus), or Ebola [[15]]. Mitigation strategies (i.e., self-isolation and quarantine of infected individuals)
employed during infectious disease outbreaks have emerged as an important public health
strategy that involved collaboration with public health informatics professionals.
Here, public health researchers focus on reducing disease spread while public health
informatics researchers use technology to develop new ways of isolating and quarantining
individuals who have been exposed to or have contracted the disease while at the same
time enabling contact with family and health professionals. Such research has been
critical to controlling the spread of COVID-19 while at the same time enhancing the
safety of the healthy population in a given country jurisdiction (as is the case in
our current global COVID-19 pandemic) [[5], [10]]. One of the most important mitigation strategies in a public health practitioner’s
toolbox is the act of isolating or quarantining individuals who may be carriers of
an infectious disease [[10], [11]]. This public health strategy protected populations from possible exposure to the
disease and aims to prevent, decrease the number, and size of outbreaks [[10], [11]]. With the development of new health technologies used in hospitals and in the community
by the individuals, public health informatics has emerged as a technology-oriented
discipline aimed at preventing the spread of disease using health technology systems
[[10],[15],[16]].
2.2 Public Health Informatics and Health Technology
In recent years, we have seen a number of health technologies used and repurposed
for use to address outbreaks of infectious disease increase in their use by public
health informatics professionals, public health officers and government policy makers
[[15], [16]]. For example, jurisdictional and national public health information systems during
the SARS-CoV-2 outbreaks had been used to detect and track those affected by new and
emergent infectious diseases [[16]]. Other technologies may have been re-purposed or used in novel ways to focus on
addressing outbreak management. Here, re-purposing or novel use of a technology includes:
using existing search engines such as Google¯ to track searches about specific symptoms, modifying mobile self-assessment and symptom
monitoring, customizing electronic health records for tracking and visualization of
patient characteristics, symptoms and treatment decisions, and updating decision support
systems to reflect real-time research about the disease and its management [[15], [16]]. This includes using existing consumer health technologies in novel ways to support
activities such as social distancing and quarantine to protect those people who are
at risk of contracting COVID-19 [[15], [17]]. Lastly, during a pandemic, new technologies may be deployed to assess their ability
to predict population and patient risk, such as machine learning algorithms [[18], [19]], or to support an individual’s decision making such as through a mobile app decision
support tool [[20]], or a web-based Chabot [[21]]. Under pandemic conditions, existing technologies may be quickly modified and used
in novel ways even while new technologies are being developed and implemented to prevent
and manage specific aspects of an infectious disease [[15],[19],[21]]. Therefore, there is a need to understand how existing technologies can be re-purposed
and selectively used to develop new technologies that may improve pandemic responses
in a strategic manner. Such research is essential to create efficient health technology
systems that can be rapidly deployed while at the same time allowing for more fulsome
allocation of funding to focus on human resource needs during a pandemic.
2.3 Pandemics, Health Technology Systems and Unified Modelling Language (UML)
Health technologies have an important role during pandemics. They allow for dissemination
of information to the public about how best to protect oneself. They also allow for
widespread dissemination about the presentation, control, mitigation, treatment and
management of the new or emergent disease to public health professionals and health
professionals [[15], [16]]. Technology has been used to effectively disseminate information during past pandemics
(e.g., SARS-CoV-2, MERS-CoV, Ebola) [[15],[16]]. The COVID-19 pandemic has shown that health technologies can be deployed as public
health interventions, helping individuals to self-isolate after exposure to disease
and health professionals to remotely monitor those who are positive for the disease
in their homes. Unified Modeling Language (UML) is used to develop graphical models
of complex, software systems [[13], [22]]. UML is widely used in many phases of software development to draw diagrams, analyze
model consistency, create design patterns, generate programming code, and produce
reports and documentation. These functions are performed using varying UML tools and
to define the ways in which objects interact with each other in varying phases. The
majority of this work is focused on software systems rather than technology systems
for supporting patients in home settings. To date there have been some published articles
on the related topics of use of UML for emergency plan development [[23]], creation of business processes as part of emergency planning [[24]], and service design for resilience during public health emergencies [[25]]. There are also a few published works on use and modeling of multi-agent systems
in medicine generally [[26]] as well as on use of UML modeling in of emerging infectious diseases [[27]]. However, there has been a lack of articles that have been published with an aim
to develop a health technology virtual care system using UML to monitor individuals
for and with COVID-19 symptoms [[22]]. There is a need to define RMT requirements for individuals who have been exposed
to and/or infected with COVID-19 in the peer reviewed literature. Furthermore, the
application of UML modelling for use in education and the design and testing of use
cases for potential home care applications could have potential benefits for creating
sharable and reusable models.
3 Method
In our current research the authors developed a method for gathering requirements.
We used reviews of the literature and published evidence-based, guidelines to design
a health technology system using UML for those individuals, who were exposed to or
infected with COVID-19. The health technology system allows for RMT to be used by
family and health professionals to monitor affected individuals [[10]].
3.1 Setting
This work took place in a partnership with the University of Victoria (Canada) Smart
Home Laboratory in the School of Health Information Science (https://www.uvic.ca/hsd/hinf/index.php),
Canada (AK, EB and RK), the Department of Computer and Information Science (IDA),
School of Engineering and Technology at Linköping University, Sweden (VM), the Department
of Biomedical Informatics, University of Pittsburgh (YS), U.S.A. and Harvard School
of Medicine, Harvard University, U.S.A. (YQ). The Smart Home Laboratory at the University
of Victoria was designed to simulate a home environment and allows for studies of
the use and usability of a range of health technologies. As part of this research,
we also worked with a technology supplier to identify technologies for testing candidate
devices and processes that can be used to promote population and individual safety
using RMT during a pandemic. Our work was intended to be vendor agnostic so we partnered
with a health technology device vendor to identify candidate devices and technologies
integrating our knowledge of health technologies, medical science and health systems.
The research initially began with the testing of varying devices in a laboratory at
the University of Victoria in the School of Health Information Science in Canada.
The lab is currently functional, and we have developed teaching and educational materials
for remote use of the technologies as part of our teamwork as well as integrated the
methodology into our health informatics curriculum for rapid provincial and country
level deployment through trained health informatics professionals.
The lab was accessible for pilot work during the pandemic by the research investigators
both in–person and virtually with the addition of virtual collaborative tools such
as Zoom¯ (AK, EB, RK). Public health protocols were followed. The team is currently working
remotely as a group using virtual collaborative technologies both within the University
of Victoria and with Linköping University, University of Pittsburgh and Harvard University.
In addition to this, the researchers have password protected, VPN access to secure
drives, and intranet services at the University of Victoria to collaborate on documents
(much as we would before social distancing requirements were introduced by the province).
3.2 Procedure
We developed our methodology for gathering requirements, designing and testing a system
of technologies to be enacted in a series of stages so our approach could be easily
re-shared and used by other public health informatics groups. We present our methodology
in a series of stages.
3.3 Stage 1
We began our work by searching PubMed.gov and the grey literature for articles, and
government websites with guidelines describing the signs and symptoms of COVID-19,
the progression of the disease and self-isolation as well as quarantine. We specifically
focused on sources where there is a description of the following four aspects of COVID-19:
-
Signs and symptoms of COVID-19;
-
The progression of the disease;
-
When there was a need to self-isolate;
-
Who is at risk for developing moderate to severe disease.
This included identification of references [[1]] through [[12]] which dealt with the above aspects. In addition to this, we took note of how COVID-19
patients can quickly overwhelm an emergency department during an outbreak and how
some patients will be hospitalized so they can receive life-saving treatment. The
articles were reviewed by two researchers to identify gaps and unfulfilled patient
and health system requirements where RMT could potentially be applied.
3.4 Stage 2
In stage 2, the researchers derived use cases, sequence diagrams and design patterns
for selecting candidate technologies and identifying objectives for supporting patient
self-monitoring, data collection and virtual care [[13]]. The modelling was based on requirements gathered and extracted from articles and
guidelines in Stage 1 (described above). Here, we employed modelling approaches from
the health systems analysis and design literature and used object oriented modeling
approaches (i.e., with each health technology acting as an object in this new health
system). Our reasoning was this: (1) we needed to understand what technologies could
be used (including what functions they performed), and (2) how these technologies
could be integrated into an existing healthcare system. Published evidence about the
disease and our knowledge of public health information systems, telehealth systems
and consumer technologies (including differing types of devices that perform the same
functions (and can be used interchangeably in a RMT system) was essential to designing
the use cases and sequence diagrams. Technologies could be added to the health system,
if the patient’s symptoms required more intensive monitoring. Technologies could also
be removed from the system (i.e., when there was a resolution of some symptoms and
a progression towards wellness). Telehealth professionals in conjunction with health
informaticists who specialize in public health informatics would assess and determine
the system of technologies that would be used and their usage over time as part of
the remote monitoring.
3.5 Stage 3
In some cases, individuals testing positive for COVID-19 are also self-isolating,
when they were asymptomatic or when experiencing symptoms of the disease [[10]]. Of note, those infected with COVID-19 may develop no, mild, moderate or severe
symptoms, and therefore need RMT. RMT is necessary so that those people who begin
to experience moderate to severe symptoms of the disease seek appropriate medical
attention in a timely manner [[5]]. During our current COVID-19 pandemic, it has become essential for individuals
to self-monitor their symptoms and to reduce contact with others unless there is a
need for health professional intervention (i.e., health service use is necessary only
in specific circumstances). Such patient RMT is necessary and appropriate use of health
care services by those with COVID-19 and those individuals requiring other urgent
treatment for cardiac conditions, cancers …etc. [[5]]. In our lab, we began to identify and test RMT that could fulfill the above outlined
requirements with a view towards how these technologies would become part of a health
technology system for remote monitoring. This involved identifying devices currently
available for purchase by consumers, reviewing the functions they could perform and
conducting a walkthrough of the device in the context of the Home of the Future. The
walkthrough included determining if the device could be used without significant effort
by a person who was wearing or using the device to collect health data.
In Table 1, we identify the types of RMT that we identified could be used to monitor
a COVID-19 patient based on physiologic needs.
Table 1
Remote Monitoring Technologies
|
Device
|
Function
|
|
Thermometer
Blood pressure device
Smart Watch
|
Monitor for fever and temperature.
Monitor blood pressure
▪ Heart rate, Respiration, Pulse Oximeter
▪ Alerts (e.g., Text, Calls and Emails to connect with Nurse or physician)
▪ GPS (e.g., to track compliance with quarantine)
▪ Sleep monitor
|
3.6 Stage 4
After reviewing several candidate devices that could be used to conduct RMT of COVID-19
patients, we met and collectively designed a health technology system using existing
consumer technologies and modeled it using UML. Use case and sequence diagrams were
developed and created (see [[22]] for detailed UML modelling steps). [Figure 1] shows requirements for a health technology system in terms of actors and activities
using UML. As indicated in the diagram, users (actors) of a system include patients,
caregivers (i.e., family, friends and formal caregivers), public health officers,
telehealth nurses and researchers. The diagram illustrates a first cut at defining
the individual use cases such as summarizing patient data, various forms of measurement
and recording of symptoms. Tracking for symptoms would be undertaken through use of
vital sign measurement devices (i.e., watch, blood pressure cuff and thermometer),
and devices could also include alerts designed for the smart watch for social distancing,
information about the number of days into the quarantine period, and educational information
focusing on the need to maintain quarantine. Lastly, the device acts as a communication
tool for connecting to other devices to collect and send data, voice communication
and contact with family, friends, a telehealth nurse and the emergency services (see
[Figure 1]). All of these functions would be performed with access to cellular networks (i.e.,
the person would not need wireless internet access) as is sometimes the case for older
adults and those living in poor socioeconomic circumstances [[8]].
Fig. 1 UML Use Case Diagram of a Health Technology System for Remote Monitoring of COVID-19
Patients to Ensure Patient Safety.
[Figure 2] shows a UML sequence diagram to detail the internal processing involved in one of
the activities indicated in the use case diagram - i.e. RMT alerting. We are using
these UML sequence diagrams for reasoning about and documenting user requirements,
system design processes and potential issues around security and privacy [[28], [29]]. This is leading to the development of a library of reusable and shareable design
patterns for application in this domain (as UML is an international modeling standard).
The researchers identified health technology configurations and methods of implementing
the technology in homes as an “in-the-box” health care technology solution in the
context of a healthcare system. Knowledge about signs and symptoms of COVID-19 including
incubation period and minimum distance needed to stay apart to ensure social distancing
was also considered in the design of the technology system. In this case the patient
uses a Smartwatch that has an app that allows for RMT. Similar sequence diagrams were
created for each of the individual twenty-two use cases (i.e., for each bubble shown
in the use case diagram in [Figure 1]).
Fig. 2 UML Sequence Diagram for Home Alert Monitoring.
4 Application of the Approach
4 Application of the Approach
We have applied the approach to exploring how a range of devices (e.g., thermometers,
pulse oximetry devices, blood pressure cuffs …etc.) can be integrated into a home
alert monitoring system. The use of this in the Smart Home Laboratory has been multi-fold.
Firstly, we have used the modelling approach as a basis for teaching upper year undergraduate
students and graduate students in health informatics about the potential use and application
of home alerting monitoring devices. This work has been the basis for development
of prototypes that can then be tested in the lab. In addition, we have extended the
use of the UML modelling and subsequent prototype development (based on the modelling)
for use in collaborative projects with regional health authorities (developing home
monitoring solutions) and a range of vendors of healthcare IT products. The application
of UML for this has allowed for improved communication, sharing of design ideas and
patterns, and the ultimate advancement of new approaches to remote monitoring. This
has had particular importance and impact as the move towards virtual care and home
health monitoring has been accelerated by the COVID-19 pandemic.
5 Conclusions
In this paper we address a significant gap in the literature, we have created a methodology
for designing a health technology system for pandemic response that is based on published
articles and guidelines. Much of the response to the COVID-19 has been reactionary.
RMT and other technologies used in virtual care have been implemented without much
consideration to developing effective and viable health technology systems using UML
modelling. In this research we outlined a methodology that can be used to gather requirements
and derive use cases, sequence diagrams and design patterns that can be reused across
health systems to effectively deploy technologies for RMT of COVID-19. This modeling
is now being used to drive the design and implementation of new forms of RMT for testing
in the Smart Home Laboratory at the University of Victoria. The approach begins with
an evidence-based knowledge of the disease and its progression. This is followed by
a focus on pandemic management using health technologies that are currently being
used and can be re-purposed for use during a pandemic.
Future research will focus on testing the approach with potential and then actual
patients to validate the design pattern for RMT (as a limitation of our current work
is the need for validation of the models in real-world contexts). Another limitation
of the work in the context of COVID-19 would be that such approaches to RMT would
only be used by those who believe there to be a pandemic (i.e., that there is a contagious
disease that requires monitoring) [[30]]. Our future work will include interviewing patients and health professionals about
their implementation experience and self-monitoring approaches. Data from the interviews
will be used to refine the selection of technologies and the implementation approach
from a user perspective, vendor perspective and health professional view (i.e. telehealth
nurses and physicians). Data collected from participants will include: demographics,
eHealth literacy levels, ability to set up the RMT effectively, ability to discuss
health issues with a nurse who is reviewing the collected physiologic data online,
user experience, and workflow data. As well, health informatics professionals used
to support this health technology system will be interviewed about the implementation
approach and process of providing care and information technology services respectively.
This research will also be extended to the remote monitoring of vulnerable patients
(e.g. cancer, respiratory and cardiovascular disease patients) to prevent deterioration
of illness, hospitalization and death during the pandemic. The research could be extended
to chronic disease patients during and after the pandemic to improve self-monitoring
and self-management of diseases. In addition, we will be exploring applying the approach
to include consideration of use of the technologies for people who are entirely asymptomatic,
but who wish to have monitoring and alerting available at home to warn at an early
stage of physiological indicators (e.g., using pulse oximeter) that are consistent
with potential COVID-19 infection (this alerting would lead them to get tested to
confirm presence or absence of COVID-19). Our future work will involve creating UML
diagrams for RMT for COVID-19 patients with chronic, long term symptoms [[31]]. We are also currently integrating this approach into health informatics workforce
training and building a health informatics workforce to initiate the deployment of
the technologies. There is a further need to conduct research on the development of
health technology systems to provide support to individuals self-isolating or quarantining
in the home. To date, this paper represents a novel application of UML modelling to
the development of health technology systems and the development of RMT approaches
for COVID-19 modelling.
