The role of technical disruptions in the healthcare industries has been a litany of cautionary tales. Attempts by
Theranos to dramatically lower the cost of blood tests created news cycles of the lack of readiness of new
technology. The inaccuracies of wearable monitors generated many articles of debates and even ridicules of their
usefulness. But these discussions are incomplete without also understanding how new technologies can be used to
Healthcare is in essence a clinical exercise aided by statistics and technology. Using calibrated and certified
medical diagnostic equipment under the supervision of trained professionals; physicians call upon very accurate
snapshots of their patients to help confirm their diagnoses. But there is often little quantified information
available on the history of a patient, especially at the early stages of an illness and snapshots alone cannot
capture the symptoms without considerable uncertainty.
Of course medicine tries to improve on the odds. Patients are classified by sex, age, race, weight, self-reported
life styles and family medical histories to reduce the variance of the relevant population. Doctors order a battery
of tests to get multiple snapshots of the patient. But in the end, the most accurate snapshots are not enough to
overcome variations among the population and diagnoses become inconclusive until later stages.
Personal health monitors can add a dimension of precision and context. A hemoglobin A1C test provides the average
level of glucose in a patient over three months. However, a patient experiencing significantly high and low glucose
levels could present a seemingly normal average. A continuously worn health monitor captures how a health parameter
for a patient has changed, and the rate of that change: information that would be spotty and expensive to collect in
a clinical setting.
Medical practitioners often voice concerns over the accuracy of consumer wearable gadgets. The upgrade from novelty
gadgets to reliable health monitors requires progress in the following areas:
- Device accuracy at affordable cost. The faster product development and replacement cycles of mobile devices
allows them to leverage advances in sensing technology and to incorporate sophisticated data processing
algorithms. Many articles pushed in the Journal of American Medical Association (JAMA) have already noted that
mobile devices are matching the accuracy of professional instruments. We at NXP are also doing our part to make
instrument grade precision inexpensive and available to quantified health monitors.
- Better understanding and mitigation of installation/wearing style variations. We know health monitors that users
wear or install at home are subjected to different environments which can affect their results. But if we know
that a device is always used by the same user, it is possible to quantify and mitigate those environmental
“noises” and variables. Quantified health monitors will need to include intelligent algorithms
that can adapt to the baseline of a user and meaningfully report the changes and rate of changes from the
- Different health and reimbursement policies. Continuing the shift from fee-for-service to value-based care will
place incentives on health providers to find affordable means of managing and monitoring their populations. As
these reimbursement policies set accuracy requirements, they will also provide a better financial return for
personal monitors which takes on the responsibility of quantifying and mitigating their potential error sources.
- Better understand privacy. Diagnoses are better informed when the decision can consider information not just
from a single patient but current information about others patients in similar “communities.” A
holistic understanding of privacy must include not only security measures to protect personal data but also the
mechanism by which a pre-approved portion of the personal data can be shared in prescribed fashions. Such a
sharing infrastructure can only be created by collaborations from stakeholders among consumers, device makers
and healthcare providers.
- Expanding the horizons of medical decisions. Many physicians are alarmed that the fundamentals of medical
practice are changing too slowly. We have built up a knowledge base on making diagnoses based on snapshots
provided by medical tests. But we are ignoring the information carried by observing subtle changes in a patient
over long periods of time. We are ignoring consideration besides medical vital signs, such as changes in a
person’s activity levels. Instead of relying on an ever increasing battery of medical tests, medicine
needs to expand its knowledge base to include these other considerations.
Aligning all the necessary changes may amount to a paradigm shift from conventional “sickcare” to
precision (or personalized) healthcare and will require advances in sensing technologies, algorithms, machine
learning techniques, security and privacy protocols and medical practices. But it is a change for the better, and