Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
End User Programming to Enable Closed-loop Medication
Management Using a Healthcare Robot
Chandan Datta, Hong Yul Yang, Priyesh Tiwari, I Han Kuo, Bruce A MacDonald
University of Auckland, New Zealand
Email: {chandan.datta,t.kuo,b.macdonald}@auckland.ac.nz
hongyul@cs.auckland.ac.nz, ptiw003@aucklanduni.ac.nz
Abstract
2005; Ryan, 1999; Higgins and Regan, 2004]. Many
medication safety related issues arise due to gaps in
the sharing of medication-use information amongst prescribers, pharmacists, individuals and caregivers. Interactive assistive technology can aid evidence-based safe
medication administration by providing decision support
to prescribers and by bringing doctors and pharmacists
to the same common understanding. It can also enable
care-givers and family members to monitor medication
regimes and actual medication usage by elders and thus
providing an opportunity to prevent potential medication errors. The medication management module in the
“Healthbots” project aims to reduce non-intentional
non-compliance among the older people by promoting patient-centered care, empowerment and autonomy
through the use of a social robot. Social robots offer a unique opportunity to add an interpersonal element to inform, empower and support older users. The
persuasive behavior of a social robot tends to promote
self management of health in older people [Looije et al.,
2010]. The main causes for older people not complying
with their medications [Lowe et al., 2000; Ryan, 1999;
Hughes, 2004] are:
Developments in healthcare delivery systems
are focusing on improved quality of care and patient satisfaction. Automated medication management on a mobile service robot is part of our
interdisciplinary Healthbots research project,
and has the potential to address both areas.
The robotic medication reminding system is the
result of an analysis of the workflow in medication management and a systematic application
of the workflow to an interactive mobile robot.
The robot is being evaluated to develop a better understanding of the technology required to
address the safety and adherence issues of medication management for seniors living independently in a retirement village. Medication data
and instructions presented during interaction
need to be programmed by end users according to the medication. Changes such as the
schedule, the names of drugs, and constraints
in taking the medication, are periodically updated by the caregiver after the medication is
prescribed by the doctor and dispatched by the
pharmacist. This paper presents an analysis of
the end user tools and the software architecture
essential to complete this medication workflow.
Initial prototypes of the software were used in
a pilot study at a retirement village. The pilot
system can support the quality of care provided
to the older people and prevent medication errors. The contribution the paper is a new analysis and novel design for enabling role based
customization for robotic medication reminding. These contributions will assist in the technical decision making of developers involved in
implementing service robots for similar tasks.
1
• Forgetfulness, the older person does not remember
to take their medication
• Change of medication schedule
• Confused by overwhelming number of instructions
• Overwhelming number of medications
• Taking medication at wrong times
• Confused by names of medication
Introduction and related work
Medication errors are one of the leading causes of morbidity and mortality amongst older people [Sokol et al.,
The organization of the paper is as follows. The background and definition of the problem are presented in
section 1, along with a brief literature review and existing solutions. Analysis of the context of the robot in the
aged care facility is presented through the description of
two field trials. The concept of closed loop medication
workflow was introduced in trial 2 and is described in
section 2.4. Our software architecture is presented in
Page 1 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
section 3.1 using UML diagrams shown in figure 6 and
figure 7. Finally, the resulting software prototypes are
presented in section 4. The software developed is shown
in figure 1 with the robot in the background.
Figure 1: Robogen running on Ipad
1.1
Related Work
Many solutions exist on the market today to tackle issues related to medication compliance [Corlett, 1996].
Medical device based solutions can be placed in categories such as pill holders, alarm based pill holders, and
pill monitoring devices [Qudah et al., 2010]. Pill holders are passive solutions. They are designed as boxes
or containers to carry medications. Usually they are divided into different compartments. The medication is
loaded manually in the compartments. The compartments can be labeled for each dosage interval. Alarm
based aids [Systems, 2011] are active solutions. Some of
these devices consist of compartments to carry medication attached to a timer. When the medication is due,
an alarm is triggered. Pill monitoring devices such as
Med-eMonitor [Naditz, 2008] are home-based medication
devices which are sealed. They dispense medication and
some of them have voice and text alarms. They can be
connected to the internet and alert the patient/caregiver
via mobile phone if a dose is missed. The iMec dosing
medication monitor [Suzuki et al., 2010] uses sensors and
fuzzy inference methods for correct medication delivery.
The system is composed of an intelligent medicine case,
house-embedded sensors, and a database server. By estimating present living conditions from the sensory data,
the iMec can confirm the dosing [Suzuki and Nakauchi,
2009]. iMec can also recognize whether an elderly person has picked up correct medicine from its storage space
to confirm the quantity of dosing. While a lot of older
people still wouldn’t use PDAs and mobile phones, however mobile phone based solutions [Qudah et al., 2010;
Dean, 2005] are available to those who are willing to
enter their medication manually into the mobile phone
or PDA and set the dosage and reminders. Two com-
mercial solutions are pillPAL and OnTimeRx [Torrico et
al., 2005]. With these applications medication compliance data is not transmitted to health professionals to
monitor patient’s compliance. They focus on improving
adherence but do not attempt to measure quality use of
medication in relation to its desired or undesired impact
on disease and patient safety. While applications for
storing, dispensing and context aware reminding have
been developed, none of them brings collaboration between patient, pharmacist, caregiver and doctor.
The second kind of solutions are software based on personal health portals. Two major commercial products
are Microsoft Healthvault and Google Health [Siek et
al., 2009; Steinbrook, 2008](which has merged to HealthVault). Microsoft launched HealthVault [Microsoft,
2011] to provide users a centralized place to track health
information, prepare for health professional visits, manage health-related devices (e.g., participating blood pressure monitors), and utilize third party health services to
share information and receive personalized feedback. Its
suitable when people want to manage their own medication list over a web application. Medication list management includes viewing, deleting, and adding medications. The web application also prominently displays
sharing functionality.
Interactive robots have not been studied to this detail
in the literature yet. Experiments presented in preliminary works [D’Este et al., 2008; Takacs and Hanak, 2008;
Takahashi et al., 2006; Takahashi et al., 2002] have not
mentioned the clinical context which needs to be incorporated to make the deployment of the robot successful
in the real world.
Table 1 summarizes some of the existing robots or software agents which have interactive robot applications for
assistive purposes.
2
Evolution of system requirements
The development of the mobile robot medication reminding requirements is documented in this section, as it
evolved from analyses and practical studies.
2.1
Field Trial 1 design and findings
Research findings [Tiwari et al., 2010] by our health
informatics team at the School of Population Health,
University of Auckland revealed the requirements of an
autonomous robot that could facilitate medication data
sharing for elderly people in aged care facilities. Following that research, a pilot field trial was conducted [Tiwari
et al., 2011b] at the aged care facility. The aim was to
assess the usability, feasibility and appropriateness of a
medication management support system on the healthcare robot system. The robot could provide the patient
with reminders to take their dispensed medications and
engage in dialog to collect information about any issues
Page 2 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
Robot Name
Research
Institution
CMU
Robot Type
Primary Application
Interactive Mobile
HANC
University of
Rochester
UCSF Medical Center
Tevital
Intelligent computer assistant
Dispense pills to medical
stations
Voice activated robot
Monitoring and recognizing an elder’s activities [Dishman, 2004]
Dialogue systems for health communication [Allen
et al., 2006]
Medication scheduling [Yeh et al., 2007]
ConnectR
IRobot
Interactive Mobile
Pearl
Chester
Waldo
Reminders for medication compliance [Warner,
1998]
Visiting robot, operated by family and
friends [Daniel et al., 2009]
Table 1: Interactive robots or agents which are used for assistive purposes
Figure 3: Layered software components for the medication workflow
Figure 2: Robot medication management field trial 1
workflow
that may be inhibiting adherence to the medication plan.
The software components to complete the workflow are
shown in figure 2.
In the observational study, a small sample of residents
were involved to provide feedback. The following results
were concluded from a post interaction questionnaire and
structured interview:
• The current state of the art for talking pill boxes
or automated dispensers is not sufficient to enable engaging interaction, provide detailed instructions/education, invoke affect and social support,
build trust by addressing safety, troubleshoot errors,
and report and call for human help in real time,
which many older people often need
• By virtue of its mobility, anthropomorphic presence,
wireless connectivity and an empowering mixedinitiative user interface, a robotic platform offers potential opportunities that were not previously possible using simple standalone reminder devices and
pill management systems
• The use of a touch screen interface supported a well
structured dialogue sequence that enabled timely
and personalized medication reminding
2.2
Motivation for a portal to enable End
User Programming
RoboGen [University of Auckland, 2011] is a web-based
application available 24 hours, 7 days a week through
any browser. RoboGen enables End User Programming (EUP). EUP trades the expressiveness of programming languages [Lieberman et al., 2006] with usability and human factors so that non-programmers can
use the system. While the system ideally should execute strictly as per physician prescription, often the
pharmacists changes brand names or packaging, and patients personalize medications and self prescribe nutritional supplements. Therefore adhering only to the prescription could be confusing to patients. Some degree
of EUP is required, which can be done by patient or
caregiver.
The motivation was to separate the clinical workflow
from the robot’s software architecture, so that the users
(as shown in table 2) had to access and modify only the
part of the system they were concerned with. This was
based on the requirements from the research summarized
in section 2.3 and described in the paper [Tiwari et al.,
2010]. The tiers of interaction between various software
components are shown in figure 3. This form of layered
design enables users, engineers and developers of the system to be isolated from each other’s domain and use the
system without knowing all the system components.
Page 3 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
2.3
Field Trial 2
The aim was to enhance the functionality from trial 1
and enable EUP of the data and dialog as per the user
role definition and workflow defined in section 2.4.
User role definition Research based on grounded
theory was conducted at the aged care facility by our
Health Informatics team [Tiwari et al., 2010]. It revealed the principal actors in the medication management workflow, their roles and concerns. The research
found that the doctor and pharmacist are more interested in preventing errors and helping their patients take
the medicines according to clinical guidelines for best
outcomes, while caregivers are focussed on solving dayto-day problems as they help these same people with
their medicines. The patients and their family members
are not concerned about the processes but interested in
resolving symptoms, remaining safe and healthy, and being treated well. The challenge for us was to incorporate
the medical implications into our research and design the
customization requirements accordingly to the provided
guidelines. Table 2 shows the different users’ roles for
completing the medication management workflow and
the features required in the application for each group
of users.
RoboGen enables EUP of the data and dialog as follows
1. main medication instructions by the doctor
2. medication names by the doctor and verification by
the caregiver
3. medication timings and schedule by doctor and
caregiver
4. side effects monitoring questions by the doctor
5. patient educational information by the clinicians
6. doctor appointments and refill reminders by caregiver
2.4
Closed Loop Medication Workflow
The
concept
of
closed-loop
medication
(CLMW) [Cousins, 1998; Williams, 2004] is associated with medication delivery mechanisms in hospital
information systems (HIS), where 56 percent of potential medication errors occur during the ordering
process, compared with 10 percent during dispensing
and 34 percent during administration [Brown, 2006,
p.330]. The situation supports a role, such as a nurse,
for maintaining medication safety [Pillow and Smith,
2007]. Closed-loop medication systems are composed of
the pharmacy information system, the Computerized
provider order entry, bar code and the point of care
dispensing device such as an automated drug dispensing
cabinet.
We adapt this concept and apply it to an automated
robotic reminder system in the aged care facility setting
to enable regular monitoring of medication intake as well
as to enable improved medication safety. The prescribers
need to know compliance as he or she needs to balance
dosage in light of adequacy of symptom control and adverse drug events. Too little will not be effective and too
much will cause adverse events. Often the prescriber adjusts dosage by guessing or an using an estimate based
on subjective reporting that is often inaccurate.
CLMW for pilot study 2 is shown in figure 4 and
explained as follows. The medication information is
initially generated by the prescribing physician using
his/her clinical judgement and expertise. The prescription is usually a set of coded information (e.g. Tab Peroxetine 40 mg TID PO, mitte 90, repeats 2 ) that has
to be decoded by the pharmacist to choose drugs and
the quantities that need to be dispensed. Often the
pharmacist detects potential problems and seeks clarifications from the physician, they also change brand
names depending upon availability and funding eligibility and sometimes actual medication or the dose or its
formulation is changed in consultation with the physician. The patients then need to convert these multiple medications and instructions to their daily routine and try their best to comply with the instructions
given. This issue is minimized by customizing the medication instructions and medication details using RoboGen. Patients have their own way of assigning meaning
to the medication instructions and calendaring the regimen. The cognitive challenges of old age make it difficult to manage this complex task, often compounded
by apathy and shifting clinical status that demands frequent change in medications [Morrow and Leirer, 1999;
Park, 2000]. The problems faced by patients in day
to day adhering with the regimen, adequacy of symptom control or appearance of side effects are rarely conveyed to physicians with the level of granularity that
can influence clinical decision making [Waitzkin, 1984;
Ong et al., 1995]. To tackle these issues RoboGen enables tailoring of the medication schedule and timings.
As the robot does the medication reminders, it logs the
medication activity and then uploads the data to RoboGen, so that doctors, family and caregivers can track the
proper medication intake. This feedback in form of analytics is visualized in a dashboard. It is very helpful to
the doctors to quickly review the patients following the
medication regimen. This is discussed in more detail in
section 4.
3
“RoboGen” design
Since users in the healthcare sector have limited knowledge of advanced devices and the ability to use state-ofthe-art technologies, we tried to simplify the workflow
Page 4 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
Physician
Generate Prescription
Define
Monitoring
Parameters
Evaluate Compliance
Care-giver
Customize user preferences
Enter reporting parameters
Older Person
Personalize name
Personalize dosage
timings
Respond to robot
Pharmacist
Review and edit prescription
Enter medication education information
Table 2: User role definition for robotic medication management
Figure 4: Closed Loop Medication Workflow (CLMW)
by limiting the functionality of the EUP tools to the
specific items required by each role as described in section 2.3. Different roles have access to their data and
can collaborate in innovative ways. For doctor, pharmacist and care-giver roles, who are not always with the
robot, there are benefits in having customization over a
web application:
• Cross-platform support Most web-based applications are far more compatible across platforms
than desktop installed software. Typically, the minimum requirement would be a web browser which
are usually installed in most systems and even
portable devices. The web browsers are available
for a multitude of operating systems and so whether
the users use Windows, Linux, Mac OS or iOS, they
can still run the web application
• Manageability Web-based systems need only be
installed on the server placing minimal requirements
on the end user workstation. This makes maintaining and updating the system much simpler as usually it can all be done on the server. Client updates
can be deployed via the web server with relative ease
• Easy deployability Due to the manageability and
cross-platform support, deploying web applications
to the end user is far easier. They are also ideal
where bandwidth is limited and the system and data
are remote to the user
• Security of older people’s data Typically,
Page 5 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
robot.
4
Figure 5: RoboGen Core and interaction with other components
in larger more desktop-oriented systems data are
stored and moved around separate systems and data
sources. In web-based systems, these systems and
processes can often be consolidated, reducing the
need to move data around. Web-based applications
also provide an added layer of security by removing
the need for the user to have access to the data and
back end servers
3.1
RoboGen components
The software design for the CLMW in section 2.4, is
shown in figure 5. The design is further explained by
the UML sequence diagram in figure 6 and UML collaboration diagram in figure 7. Unlike the sequence
diagram, the collaboration diagram shows relationships
among component roles and does not express time as a
separate dimension. Therefore, the messages in the collaboration diagram are numbered to indicate their sequence. The proposed web-based tool’s architecture is
sketched in figure 7. The user’s browser on a PC or mobile device connects to the web server using an HTTP
request. The web server hosts the web application on
the IIS (Internet Information Server) web server, which
receives and sends HTTP requests. The web application
is a Microsoft ASP.NET application that handles these
requests. The application data are stored in one or more
databases. The application can access the database using SQL commands or a web service. The ASP.NET
application dynamically creates HTML pages, which IIS
posts to the users’ web browser. The browser then displays the page for customization. A request for medication dialog and associated medication data from the
Healthbot framework returns the customized information for the particular older person interacting with the
Results and RoboGen Prototype used
in Field Trial 2
The implementation of the requirements discussed in section 2.3 was evaluated in a user study conducted at Selwyn Village, a residential aged care facility housing over
300 older people in March 2011 [Tiwari et al., 2011a]. We
conducted a formative study to access patient acceptance
and feasibility of our design. Following an ethics approval from the institutional committee, six participants
were recruited by a random selection and expression of
interest in participation. Their personal health records
were obtained from their General Physicians and designated pharmacies. The trial was conducted for a two
week period. During this period the researchers took
the robot to the participants’ apartments. The medication information was entered into RoboGen prior to
conducting the interaction with each participant. At the
end of each interaction, a semi-structured interview was
conducted. Clinical data was collected from questionnaires, video recordings of interactions, researcher notes
and activity logs in RoboGen.
The age of participants ranged from 83 to 92 years
with a mean of 86.6, out of which 50% were male and
50% female. Half of the participants used pill boxes to
organize their medication, while others used loose pills.
The participants did not receive personal assistance from
the care-giver for supervising the robot interaction or
medication administration. Table 3 [Tiwari et al., 2011a]
shows the results of interactions with the robot, extracted from the recorded videos. The participant could
interact with the robot everyday. The total number of
interactions logged was 45, out of which 42 were successfully completed. In the other 3 instances, the participants had already taken their medications without
waiting for the robot to arrive.
The closed-loop medication system along-with the
robot was intended to tackle the following distinct causes
of medication error :
• Forgetfulness and missed doses
• Confusion about timings and quantity of medications
• Complex medication regimes demanding different
medications on different days
• Changed prescriptions on hospitalization and acute
episodes of illness
These issues were addressed by incorporating the following in our system design :
• Timely reminders of medication as per patient preference
Page 6 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
Figure 6: UML Sequence Diagram for CLMW
: RobogenDatabase
4:result set
3:* sql command/web service
7:dialog request
: ASP.NET MVC Application
: Healthbot Framework
8:dialog response
2: Data customization request
5:autogenerated DHTML
Vision
Engine
: IIS
Voice
Engine
Robot software engines
Camera
Motors
Robot Hardware engines
1: Data customization request
6:autogenerated DHTML
: Browser
Figure 7: UML Collaboration Diagram for CLMW
• Precise dialog and instruction detailing names,
doses and method to take each medication
• Simplification of the medication regimen where the
cognitive burden of handling complexity has been
handled using the robot and RoboGen
• Changes in prescription translates into an accurate
reminder dialog helping the patient to cope with the
change
An interaction screen showing the participant receiving an accurate instruction from the robot is depicted
in figure 8. The results which enable the other stakeholders in the workflow to minimize the error and delivery higher quality of care is shown in figure 9.
A summary of the formative study is as follows :
• The intended goal was achieved in 42 out of a total
Figure 8: An interaction screen showing the participant
an accurate instruction
Page 7 of 10
of 45 interactions: successful intake of medication
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
Participant
Number
Total
No.
Interactions
Success
intake
1
2
3
4
5
6
9
5
8
8
7
8
9/9
5/5
8/8
8/8
6e /7
6f /8
of
Correct
Prompting/
appropriate
dialog
Instances
needing
researcher
help
Yes to side
effects
Non-tech.
errors
(if any)
8a
4c
8
8
7
8
day
day
day
0
0
day
no
Yes
no
no
no
no
1b
0
0
1d
0
1
1
1,2,3
1
1
Table 3: Result of the interactions with the robot from the recorded videos
a
Taken medication without robot
b
Loose medication was arranged mistakenly in the pillbox
c
Unclear when to swallow pill
d
Missed supplements
e
Taken medication without robot
f
Taken medication without robot
Figure 9: Medical Compliance Summary on RoboGen accessed by a physician
while being prompted by the automated dialogue
system, in turn driven by the Robogen database.
• Moreover 43 of 45 prompts were correctly delivered
while being sequentially relevant and contextually
appropriate.
• RoboGen correctly collected and displayed the session outcome logs as intended at each instance in
real time. No errors were recorded during the trial.
• Medication side effect monitoring dialogues were delivered 11 times, and one was reported positive.
• Overall the users rated their experience positive.
After the initial days the ratings ranged from good
to excellent. At the end of the study the users gave
similar ratings.
• The intended goal of increasing participants medication knowledge, by providing detailed medication
information as an option, was not successful in the
current design. Despite being made available as an
option and being demonstrated in the beginning the
users neither accessed it nor remembered it.
These results show that the new role based analysis of
end user tool requirements, the software architecture,
and initial prototype can support the medication management workflow required for quality of life for older
people, and prevent medication errors.
Page 8 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
5
Conclusion
This paper presents our mobile healthcare robot medication management system based on an analysis of the
medication management process in an aged care facility, along with the customization tools and the software
architecture essential to complete a medication management workflow. The main objective of improving medication compliance with an interactive robot and the software tools is to reduce mortality and morbidity rates
and reduce the cost and burden on the health system,
due to preventable hospitalizations. The integration of
the medication management module into the robot suite
of software services allows for collecting valuable medication compliance data and side effects responses. As
a result we hope to see health improvement and better
patient awareness with respect to their medication and
also the other healthcare professionals involved in the
workflow to be better informed about their patients.
Acknowledgment
This work was jointly supported by the R&D program of
the Korea Ministry of Knowledge and Economy (MKE)
and Korea Evaluation Institute of Industrial Technology
(KEIT). [KI001836], Development of Mediated Interface
Technology for HRI] and the New Zealand Foundation
for Research, Science and Technology IIOF (13635). We
thank Electronics and Telecommunications Research Institute (ETRI) for their valuable contributions and help
with the research. We would also like to thank Yujin Robot for their technical support and our colleagues
of University of Auckland HealthBots research team for
their ongoing support.
References
[Allen et al., 2006] J. Allen, G. Ferguson, N. Blaylock,
D. Byron, N. Chambers, M. Dzikovska, L. Galescu,
and M. Swift. Chester: towards a personal medication advisor. Journal of Biomedical Informatics,
39(5):500–513, 2006.
[Brown, 2006] T.R. Brown. Handbook of institutional
pharmacy practice. American Society of Health System, 2006.
[Corlett, 1996] AJ Corlett.
Aids to compliance
with medication. BMJ: British Medical Journal,
313(7062):926, 1996.
[Cousins, 1998] D.D.M. Cousins. Medication use: A systems approach to reducing errors. Joint Commission
Resources, 1998.
[Daniel et al., 2009] Kathryn Daniel, Carolyn L. Cason,
and Sherry Ferrell. Assistive technologies for use in the
home to prolong independence. In Proceedings of the
2nd International Conference on Pervasive Technologies Related to Assistive Environments, PETRA ’09,
pages 26:1–26:4, New York, NY, USA, 2009. ACM.
[Dean, 2005] Laura Dean. Handhelds for Patients, pages
59–63. John Wiley Sons, Ltd, 2005.
[D’Este et al., 2008] C. D’Este, D. Reid, and B.H. Kang.
A Robotic Interface to a Medication Review Expert
System. In 2008 International Symposium on Ubiquitous Multimedia Computing, pages 145–150. IEEE,
2008.
[Dishman, 2004] E. Dishman. Inventing wellness systems for aging in place. Computer, pages 34–41, 2004.
[Higgins and Regan, 2004] N. Higgins and C. Regan. A
systematic review of the effectiveness of interventions
to help older people adhere to medication regimes. Age
and ageing, 33(3):224, 2004.
[Hughes, 2004] C.M. Hughes. Medication non-adherence
in the elderly: how big is the problem? Drugs &# 38;
Aging, 21(12):793–811, 2004.
[Lieberman et al., 2006] H. Lieberman, F. Paternò,
M. Klann, and V. Wulf. End-user development: An
emerging paradigm. End User Development, pages 1–
8, 2006.
[Looije et al., 2010] R. Looije, J. Arendsen, J. Saldien,
B. Vanderborght, J. Broekens, and M. Neerincx.
Robots that care. In Proceedings of the 28th Annual
European Conference on Cognitive Ergonomics, pages
301–302. ACM, 2010.
[Lowe et al., 2000] C.J. Lowe, D.K. Raynor, J. Purvis,
A. Farrin, and J. Hudson. Effects of a medicine review and education programme for older people in
general practice. British journal of clinical pharmacology, 50(2):172–175, 2000.
[Microsoft, 2011] Microsoft.
Microsoft HealthVault.
www.microsoft.com/health/en-us/products/
Pages/healthvault.aspx, 2011. [Online; accessed
March-2011].
[Morrow and Leirer, 1999] D. Morrow and V.O. Leirer.
Designing medication instructions for older adults.
Processing of medical information in aging patients:
Cognitive and human factors perspectives, pages 249–
265, 1999.
[Naditz, 2008] A. Naditz. Medication ComplianceHelping Patients Through Technology: Modern Smart Pillboxes Keep Memory-Short Patients on Their Medical
Regimen. Telemedicine and e-Health, 14(9):875–880,
2008.
[Ong et al., 1995] L.M.L. Ong, J.C.J.M. De Haes, A.M.
Hoos, and F.B. Lammes. Doctor-patient communication: a review of the literature. Social Science &
Medicine, 40(7):903–918, 1995.
Page 9 of 10
Proceedings of Australasian Conference on Robotics and Automation, 7-9 Dec 2011, Monash University, Melbourne Australia.
[Park, 2000] D.C. Park. The basic mechanisms accounting for age-related decline in cognitive function. Cognitive aging: A primer, 11:3–19, 2000.
Robotics and Vision, 2002. ICARCV 2002. 7th International Conference on, volume 2, pages 1025–1030.
IEEE, 2002.
[Pillow and Smith, 2007] M. Pillow and V. Smith. The
Nurse’s Role in Medication Safety. Joint Commission
on, 2007.
[Takahashi et al., 2006] Y. Takahashi, H. Goto, and
H. Shigemoto. Facial expressions of simplified robot
face for health care system. Transactions of the Institute of Measurement and Control, 28(2):163, 2006.
[Qudah et al., 2010] I. Qudah, P. Leijdekkers, and
V. Gay. Using mobile phones to improve medication compliance and awareness for cardiac patients.
In Proceedings of the 3rd International Conference on
PErvasive Technologies Related to Assistive Environments, pages 1–7. ACM, 2010.
[Ryan, 1999] A.A. Ryan. Medication compliance and
older people: a review of the literature. International
journal of nursing studies, 36(2):153–162, 1999.
[Siek et al., 2009] K. Siek, D. Khan, and S. Ross. A usability inspection of medication management in three
personal health applications. Human Centered Design,
pages 129–138, 2009.
[Sokol et al., 2005] M.C. Sokol, K.A. McGuigan, R.R.
Verbrugge, and R.S. Epstein. Impact of medication
adherence on hospitalization risk and healthcare cost.
Medical Care, 43(6):521, 2005.
[Steinbrook, 2008] R. Steinbrook. Personally Controlled
Online Health DataThe Next Big Thing in Medical Care?
New England Journal of Medicine,
358(16):1653–1656, 2008.
[Suzuki and Nakauchi, 2009] T.
Suzuki
and
Y. Nakauchi. Dosing monitoring system for caretakers
using intelligent environment. In System Integration,
2009. SII 2009. IEEE/SICE International Symposium
on, pages 107–112. IEEE, 2009.
[Suzuki et al., 2010] T. Suzuki, Y. Oyama, and
Y. Nakauchi. Intelligent medicine case system with
distributed RFID readers. In Engineering in Medicine
and Biology Society (EMBC), 2010 Annual International Conference of the IEEE, pages 344–347. IEEE,
2010.
[Systems, 2011] Medcenter
Systems.
Medcenter Monthly Medication System.
www.
medcentersystems.com/zOrganizerStand.html,
2011. [Online; accessed 19-March-2011].
[Takacs and Hanak, 2008] B. Takacs and D. Hanak. A
prototype home robot with an ambient facial interface
to improve drug compliance. Journal of Telemedicine
and Telecare, 14(7):393, 2008.
[Tiwari et al., 2010] P. Tiwari, J. Warren, and K. Day.
Can a robot facilitate medication information sharing in an aged care facility? In HIC 2010 Conference: Informing the Business of Healthcare, 24-26 August 2010, Melbourne Convention and Exhibition Centre, page 156. Health Informatics Society of Australia,
2010.
[Tiwari et al., 2011a] Priyesh Tiwari, Jim Warren,
Karen Day, and Chandan Datta. Comprehensive support for self management of medications by a networked robot for the elderly. In Health Care and Informatics Review Online, www.hinz.org, HINZ ’11, 2011.
[Tiwari et al., 2011b] Priyesh Tiwari, Jim Warren,
Karen Day, Bruce MacDonald, Chandimal Jayawardena, I.H. Kuo, Aleksandar Igic, and Chandan Datta.
Feasibility study of a robotic medication assistant for
the elderly. In Proceedings of the Australasian Conference on User Interface, AUIC ’11, Darlinghurst, Australia, Australia, 2011. Australian Computer Society,
Inc.
[Torrico et al., 2005] S. Torrico,
F.D. McGuire,
C. Roberts, D.J. Verhaagh, G.C. Douglas, SE McGee,
and G. Hibbins. More serious letters about obesity.
Medscape General Medicine, 7(1), 2005.
[University of Auckland, 2011] Department of ECE
University of Auckland, Robotics Research Group.
RoboGen Portal. https://robogen.ece.auckland.
ac.nz/, 2011. [Online; accessed August-2011].
[Waitzkin, 1984] H. Waitzkin. Doctor-patient communication. Journal of the American Medical Association,
252(17):2441, 1984.
[Warner, 1998] I. Warner. Telemedicine in home health
care: The current status of practice. Home Health
Care Management & Practice, 10(2):62, 1998.
[Williams, 2004] C.T. Williams. Inside a closed-loop
medication strategy. Nursing management, 35:8, 2004.
[Yeh et al., 2007] Han-Chun Yeh, Chi-Sheng Shih, and
Jane W. S. Liu. Workflow-based integration framework for medication-use process. In Proceedings of
the 2007 HCMDSS-MDPNP, pages 161–163, 2007.
[Takahashi et al., 2002] Y. Takahashi, H. Goto, and
T. Saito. Health care system using face robot and
tele-operation via Internet. In Control, Automation,
Page 10 of 10