|Artificial Intelligence: Healthcare Can Benefit From Smart Use of Data|
|By: Alan Cudney, RN-BC, CPHQ, PMP, FACHE|
The term Artificial Intelligence (AI) is routinely used in
the media. Through fictional movies like The
Terminator series and I, Robot,
we have become de-sensitized to the idea, and many view it as something that is
still “way out there,” leaving consideration to aficionados of fiction. However,
with advances in information technology and, more specifically, analytics, healthcare
delivery will see big changes over the coming decades. Here are some possible
uses for AI and examples of how it could work.
Intelligent clinical decision support at the point of care
An endocrinologist evaluates options for Mary, a patient
with Type 2 diabetes, to help better manage her health. Using data sources
aggregated and normalized across episodes and sites of care, the physician can instantly
see her complete medical history, including a related hospitalization in a
different city, her prescription refill history, and her completion of diabetes
management education. As he begins to place an order for a new type of insulin,
the clinical system responds that, due to Mary's clinical history and genetic
map, a different drug combination would be more effective.
Predictive models that continuously re-evaluate health status and best ways
to help people make healthy choices
Ron hates to fill out surveys and has missed two of his
last five doctor appointments. He ignores voice mail but responds to text
messages and automated medication reminders from his cell phone. AI is used to
continuously update a predictive model, which projects the most effective way to
communicate with Ron and encourage him to make healthy choices, based on his
previous activities. The model includes data feeds from social media,
healthcare patient portals, credit card companies and pharmacy benefit
In a like manner, an ACO or medical home will stratify
groups of patients according to clinical measures, health status and lifestyle
characteristics, automatically assigning patients to specific categories of
care management intervention. AI revises the predictive models as new data are available.
These models then re-run the cohort selection, risk stratification, and
Medical supply companies will predict future supply and demand needs for
durable medical equipment at the unit, facility or service area levels
Imagine AI software is able to predict that, based on public
health and hospital data, demand for home oxygen will increase in a six-county
area of the state. The software suggests re-direction of specific oxygen
canisters volumes and numbers of delivery staff to that geographic area. The AI-driven
models anticipate patient needs and respond much more quickly than traditional
reporting and surveillance processes.
companies will customize prescription drugs, based upon understanding of
individual DNA and likelihood of intended response
Marla is taking a new
medication to control her hypertension. The molecular structure of the
medication, as well as the dosage, have been slightly modified to account for
her genetic predispositions and previous response to other anti-hypertensive
medications. The predictive models continuously optimize themselves as new
clinical data comes in from hospitals and doctor offices.
AI with deep learning
of radiological images and nanotech sensors will take clinical decision support
to a new level
Pathologic lesions can be
missed on diagnostic images. Detecting these is often dependent upon the
training and expertise of the individual clinician. AI with image detection
software will be able to aggregate experience across thousands of patients and
suggest additional diagnostic options to the physician. Combining insights from
these types of analyses with more traditional analytics should help providers
to standardize diagnoses and assist radiologists, increase efficiency and
Such uses can go beyond
imaging to include nanotechnology sensors embedded a pill or ingested fluids.
These sensors can contribute to the data used to create recommended diagnoses
and treatment plans.
How will AI affect the
AI will enable the delivery
personalized, proactive healthcare that is more efficient, more effective and
less expensive. In order for it to work properly, AI is best enabled in a
unified analytic structure that has already learned to govern and manage data
within the organization. This structure accompanied by the right blend of
software, clinicians and data scientists can make AI a reality for improving
healthcare. Just as the EMR required major focus and development of
infrastructure, AI will take some work to realize its potential.
How can we begin to
prepare for AI?
· Take time to become more “data savvy” by learning about ways
others are using analytics to improve care.
· Support efforts of your organization to better manage data and
turn it in to useful information.
· Offer refresher training in basic statistics and data management,
so no one is left behind.
· Make sure you are using software that not only handles day-to-day
reporting needs but is able to scale for the future.
· Consider adding to your team a data scientist, who can develop
data models and provide guidance in practical use of advanced techniques.
Interested in other perspectives? Consider
predictions on how artificial intelligence will change clinical care – for the
better or worse?” by Laura Dyrda of Becker’s Hospital Review. The article
includes insights from me and several other industry experts.
This should get you started
and better prepared to take advantage of AI and other new technologies that are
on the horizon. Healthcare can surely benefit from smarter use of data!