Rare Birds in Analytics

Posted on October 7, 2015 Published by Leave your thoughts

Companies across all industries are trying to figure out how to use their data to measure and predict outcomes, make smarter decisions and increase customer satisfaction. Often they have so much data that it’s hard to know where to start. It’s always interesting to learn from those who have taken the plunge what has and hasn’t worked.

In a webcast entitled “Health Analytics: Find Data Beyond the Hospital Doors”, Bob Gladden, vice president of the Center for Analytics at CareSource, described a couple initiatives he’s leading. He also gave some helpful tips to listeners.

CareSource is an HMO that specializes in the Medicaid sector of the market in Ohio, Kentucky and Indiana. Gladden explained how he’s taking claims data and crossing it with demographic data to reduce the probability of hospital re-admissions.

By mining and analyzing data, his team identified a vicious cycle for members who suffer from asthma. They go to the hospital, get better, go home, get sick again, and eventually end up back in the hospital. What they also discovered is that these members often live in housing conditions that aggravate asthma.

So now CareSource is leveraging the analysis to create solutions that will break a cycle that doesn’t improve members’ health and increases healthcare costs. They’re looking at crossing their claims data with data from housing experts on the existence of mold, insects or other environmental issues that could potentially make members sick. What’s even more innovative, they’re investigating whether it’s more cost-effective to remediate the cause of the problem than to pay for repeat hospital stays.

CareSource is also using data analytics to solve health problems in prison populations. When people are arrested, the police typically take away their medications. Then when they’re released from jail, eventually they end up in the emergency room because they’ve become either physically or mentally sick. If the company could partner with county agencies and use their data – and there are legal and privacy issues here – they may be able to stop this cycle as well.

Gladden advised listeners not to ignore the low hanging fruit. Just grab exploratory or informational data from anywhere and in any format, put it in a sandbox and work with it to see what insights emerge. Then perhaps pull in some operational data (data that comes from the organization’s operations), run it through an algorithm and do some predictive modeling.

But the real challenge lies in taking the analysis and executing on it. For that, you need someone who understands data analysis, can put it in the context of the business and drive innovative solutions. You need a good communicator who is aware of industry trends and speaks the language of statisticians and line managers. Now that’s a rare bird.

Tags: , , ,

Categorized in:

This post was written by Sententia Partners

Leave a Reply

Your email address will not be published. Required fields are marked *