- Even the most enthusiastic data analytics advocate would be hard pressed to argue that the healthcare industry is on the cutting edge of adopting and leveraging data science techniques to solve some of the most pressing problems of clinical care, business intelligence, and revenue cycle management. Chronically short of resources and burdened with a slew of competing initiatives that can throw analytics programs off track, healthcare lags far behind other industries, such as transportation, retail, and finance, in the way it uses data to generate insights, engage consumers, and drive improvements in quality care.
However, the cause is not a hopeless one. Many organizations are starting to make the commitment to clinical and financial analytics, and are turning towards those industries that have already traveled the rocky path towards a data-driven infrastructure to help them solve critical challenges. They are finding that other segments of the economy have plenty of lessons to teach hospitals and physician providers in the areas of big data management and predictive analytics.
Steven Escaravage, Principal in Booz Allen’s Strategic Innovation Group, believes this spirit of collaboration and shared learning is an important factor for the success of the healthcare market. Escaravage explained to HealthITAnalytics how the knotty problems of data science transcend industry boundaries.
To read the first part of this interview, please click here.
Why is it important to look to other industries for data analytics insights?
Every problem has its own nuances, and when we go into a situation, we’re always amazed at how experts in their field will understand the problem that they face with such a depth that it can be overwhelming at times. But when we start looking at the data and figuring out techniques to attack it, the problem starts to become demystified.
On one hand, yes, the data is very specific to the problem at hand and we do need a lot of experts to understand it, but all data is fundamentally the same at its core. Data sets are observations that occur over time. It’s data that describes the utilization of some type of resource. It’s some transaction that takes place between a citizen, a customer, a patient, and some system that’s trying to provide benefits to those people.
We’re looking for outliers, we’re looking for inliers, we’re looking for anomalies or irregularities in the data, we’re looking for patterns, and we keep coming back to the same set of methods or techniques. At the end of the day it’s really about how you group things with similar characteristics. You’re looking for those outliers, you’re looking for relationships and associations that you can use to predict future events and categorize observations in the past.
What are some of the main areas where you see a lot of commonalities between industries?
We’re all obsessed with time. We’re all obsessed with being more efficient and more effective. We have to be if we want to deliver better outcomes. We’ve got to find a way to be more efficient and reduce cost. So, what I’ve seen in in energy, transportation, and finance is a real effort focused on dealing with data that is captured over time. It produces a sort of imperfect understanding of events as they take place – time lags or time warps, as they call them. It’s not easy trying to figure out what’s going to happen in the future based on what the pattern has been in the past.
In healthcare, we need to predict how a patient is going to trend in terms of one of their key vital signs or statistics. In transportation, the question is whether or not there’ll be congestion on a highway or on a road, and in finance, it’s how a market is going to move in the future.
We have experts – PhD’s in computer science, mathematics, or statistics – and we find these same challenges materialize across different spaces. Every week, I will sit with my colleagues, who are really our finance analytics specialists, our transportation analytics specialists, and we’ll always be astonished that we’re all trying to solve fundamentally the same algorithmic challenges that have just materialized in completely different spaces.
Clients might be seeing their specific challenges for the first time. But as we break down the barriers between different domains and start reducing the problem to the data, and the methods and the techniques that they want to apply, in many cases we’ve solved that problem not once or twice but many, many times for many different clients. So we’re already armed with a space of a solid understanding of what we need to do and the opportunities and the pitfalls.
How can healthcare providers leverage the understanding and skills of professionals from different industries?
I think a real key is having those individuals that have a hybrid understanding. Let’s take wearable technologies in the health space, for example. We see a lot of literature being leveraged from signal processing and control systems that you would use in manufacturing, as well as algorithms and techniques that have been used to program autonomous robots to figure out things like determining exactly what the acceleration has been for an individual wearing a wearable device.
So, I think you’re going to continue to see, as the health space does become more and more data-driven, you’re going to continue to see analytic methods and techniques pulled from other disciplines to really supplement an expert medical professional’s information.