The Journal of mHealth Vol 3 Issue 1 (Feb/Mar 2016) | Page 25

The Path to Digital Health Nirvana is Paved with Data
So how can we change this? I firmly
believe that answer is in the intelligent
use of patient data. When we talk about
“patient data” we really mean two things;
subjective and objective data. Subjective
data is data that is reported by the patient.
How they are feeling, what their pain rating is or how many times did they take
their medication last week? This data is
extremely important for understanding
what an individual patient needs from
their intervention. Just as a physician
would ask the patient questions when
they are in the office, digital health technology can use patient reported information to add utility to the intervention.
However, because of it’s inherent biases,
there are downsides to relying solely on
subjective data. Numerous studies have
shown that subjective assessments often
do not line up with their objective counterparts. A good example is sleep. When
patients are asked to rate their sleep quality, it often has little or no correlation
with how well their brainwaves say they
actually slept3. In this way, objective data
can be very powerful, providing an unbiased assessment of an important health
parameter. The other advantage of using
this type of data is that it can often be
collected passively, eliminating the need
for a patient to input the information.
The advent of multiple wearable technologies has really expanded the types
of objective data that can be collected.
Blood glucose, activity, and sleep patterns are just some examples of available
data that could be used by digital health
interventions to increase relevance (and
utility) for patients.
At MEMOTEXT, we believe in the
power of data and we use it to build
digital health interventions that improve
patient outcomes by targeting medication adherence. To test how “actioning”
patient data could improve patient outcomes, in late 2013 we partnered with
Green Shield Canada (GSC) to develop
a free hypertension and high cholesterol support program for eligible plan
members. The goal of the program
(Stick2It) was to offer supportive messaging and personalized reminders to
improve adherence to medication and
thereby achieve better long term health
outcomes.
Plan members who signed up for the
support program completed an intake
survey and were asked how well they
were sticking to their medication, what

kinds of supportive educational messaging they would like to receive, and what
factors might prevent them from adhering to their treatment. In addition to collecting this subjective data, pharmacy
claims information was collected (in real
time) for each member. This allowed us
create automated refill reminders while
also giving us an objective source of data
to measure the outcome.
Based on their responses, members
received messages (phone calls, texts,
or emails) about their illness, diet and
exercise advice and tips for maintaining a healthy lifestyle. The amount of
messages members received from each
category was based of their own subjective (reported) and objective (claims)
data, creating a personalized program for
each individual. Six months later, members completed another survey asking
them again about their preferences for
messaging and how well they were sticking to their medication regimen. These
responses were used to further finetune the proportion of messages they
received from each category.
This data-driven personalization strategy had a significant impact on patient
medication adherence. The proportion
of patients who stopped taking their
medication a gFW"