5/28/2013
Data provides us with the information to answer empirical
questions. Every fifth grader in America is familiar with the Scientific Method
and its ability to prove or disapprove a hypothesis. But what about Big Data?
How do we modify our approach to find tangible value for our consumers?
Ask the Right
Questions
Big Data is messy and not precise. Two plus two equals three
point nine, and that’s okay. When we have twenty data points in a system,
accuracy is essential; but with twenty million sensors, accuracy is not so
important.
We know that predictions based on correlations lie at the
heart of Big Data, but to unlock all of its potential we must adjust our
expectations.
The entire conversation of Big Data in the healthcare
industry must change. Instead of asking “Why” we must first ask “What”.
The “Big Data”
Mindset
You can lead a horse to water, but you can’t make him drink.
Once we begin to ask the right questions, we must open our minds up to “see”
the solutions. This shift towards a “Big Data” mindset will enable us to see
things differently and allow us to act on opportunities before others do.
Creating Solutions
from Correlations
To extract value from data we must first be able to see the
possible. This unique sight is what empowers innovators to create disruptive
solutions.
The Medical Star Washington Hospital Center in Washington, D.C.
and Microsoft Research recently teamed up in an effort to reduce hospital re-admissions.
They wanted to identify “What” correlations existed between patients that were
re-admitted to the hospital within 30 days.
With the help of Microsoft’s Amalga software they uncovered all
the usual “predictors” but also found something else. One of the top predictors
of re-admission within 30 days was highly correlated with mental status upon
admission. If a patient’s H&P included a negative mental status with words
such as “depression” the patient was more likely to return to the hospital.
With this knowledge the hospital can now create new initiatives
to prevent re-admissions; thus adding value to both the consumer and the
hospital.
In Denmark, a study
combined 3 million cellphone records with the National Cancer Registry to see
if they could find a link between cell phones and cancer. Their study didn’t
show any correlation, but the idea of combining multiple data sets to predict
points of intervention allow us to view the world around us in a much different
way.
The solution to our healthcare crisis is far from simple,
but I believe the answer lies somewhere in the masses of data we have captured.
What is your data telling you? Are you acting on the
opportunity? If not, why?
No comments:
Post a Comment