1. INTRODUCTION
Hospitals must maintain records of all the patients that use the
hospital’s services. This includes all the patients in the outpatient,
inpatient and emergency care. One of Taiwan’s largest hospitals,
the Taipei Veterans General Hospital (VGH), recorded over 2.7
million visits by patients in 2012. In order to be able to quickly
retrieve and/or update the necessary records, the VGH maintains a
vast database maintaining all the electronic health records (EHRs)
of each of these patients. Doctors can access their patients’ EHRs
at designated work computers such as at nurse stations, clinic
examination rooms or in their offices. However, only attending
physicians are given assigned offices, which means resident
doctors are sharing the work stations most of the time.
These EHRs largely confine the doctors, and even the nurses, to
desktop computers. Some doctors find that they must return to
workstations a few times during their morning rounds only to
search for available ones. This creates an overhead of workload
for the doctors.
This gave attention to the personal digital assistant (PDA) [5, 7].
Many hospitals have begun to adopt mobile EHRs accessible via
iPads or similar tablets, and have studied doctors’ reactions and
preferences when using the iPad applications [5, 6]. However,
these are studies with data that were self-reported by doctors.
Unlike these previous studies, we collected live, motion data on
top of surveys and interviews from each doctor during their use of
the iPads. We deployed Dr. Pad, an EHR application on the iPad,
to resident doctors at the Taipei VGH, and looked at 179
physicians’ mobility patterns derived from this data to see how the
doctors’ actual activities match with their claims and our
observations. Our system gives us concrete data on the doctors’
usage patterns such as time of day, and frequency of use.
Additionally, we trained an unpruned J48 decision tree classifier
to tell us each doctor’s type of mobile activity while using Dr.
Pad, like walking, standing or sitting at a desk, up to 96.6%
accuracy. This previously unavailable collection of data allow us
to illustrate the mobility of doctors using a mobile EHR, and see
how mobile they have become. Furthermore, this data lets us
observe use trends that supported their reports of how they used
the mobile EHR.
In the remainder of this paper, we will discuss previous studies on
similar applications of the iPad at other health service
establishments, and some of the shortcomings in these studies. We
will look at what their results tell us about doctor preferences, and
the pros and cons of implementing an iPad EHR. Then we will
describe the Dr. Pad application and its deployment at the Taipei
Veterans General Hospital. This is followed by our study on the
doctors’ behaviors before and after the introduction of Dr. Pad
through surveys and shadowing sessions, which will establish our
motivation to capture motion data. Then we present the design and
implementation of the motion tracking and logging code, as well
as how our classifier is trained to predict the activities. We will
show; with live, quantitative data, the resident doctors’ mobility
patterns and how it complements the qualitative reports. Finally,
we will conclude with the implications derived from these new
results.
2. RELATED WORKS
A lot of a doctor’s work is time-critical; therefore, iPads and
similar tablets are attractive for their claims of increased mobility
and the possibility of saving time before doctors get to their
Permission to make digital or hard copies of all or part of this work for personal
or classroom use is granted without fee provided that copies are not made or
distributed for profit or commercial advantage and that copies bear this notice
and the full citation on the first page. Copyrights for components of this work
owned by others than the author(s) must be honored. Abstracting with credit is
permitted. To copy otherwise, or republish, to post on servers or to redistribute
to lists, requires prior specific permission and/or a fee. Request permissions