Big data builds individualized healthcare strategies


This week, I spoke at The Center for Excellence track as part of the 2015 Workers’ Compensation and Education Conference in Orlando, FL. The responses from attendees reconfirmed how much people are seeking to understand how big data and predictive analytics can play a role in their organizations.

One overwhelming theme emerged from those in my session. Big data must be used to build and sustain personalized healthcare and productivity models.

As you might imagine, there was a lot of information covered and for those unable to attend I want to share some of my thoughts on this topic. So here are some key things you need to know.

Big data, actionable data, predictive data…the industry is abuzz with discussions about the importance of data.  However, data, even organized data, provides no value in business until it is paired with purposeful action. Industry systems and practices have evolved to give business leaders unprecedented access to information. One of our biggest opportunities to revolutionize our injury management solutions is to use big data to develop and sustain models that personalize the injured worker’s healthcare experience and provide guidance to all stakeholders on how to use this information to affect change. The overall concept is to turn big data into results!

I want to share some important areas in which data can be used to build models that will individualize healthcare management strategies.

You have identified the factors that lead to large loss development.  Now what?

Match treatment intervention solutions and injury management team skill sets to injuries and circumstances that will benefit from those solutions and provide training and guidance to this team on how best to improve the outcome for each injured worker. A multi-disciplinary team consisting of claims adjusters, nurse case managers, behavioral health specialists, vocational rehabilitation specialists, pharmacists and physicians should be interconnected through systems and practices.

Triggers derived through predictive modeling must get specific injury situations into the hands of the team members whose skills will provide the best and speediest remedy. Seamless transitions across the injury management team and exchanges of information must be automated. For example, psychosocial and co-morbidity information systematically captured and stored is included in a trigger that can engage a behavioral health specialist or provide the examiner or nurse with the information needed to tap into the most effective strategies to improve outcomes. Strategic prescription drug safety solutions are deployed by systematically connecting prescriber, dosage and drug duration information across dispensers, networks and claims administrators. A vocational rehabilitation specialist may be deployed for a claim without continuing medical issues in which return to work has not yet been achieved.

You have scored providers through analysis of medical cost or wholesale claims outcomes. What do you do with the information?

Measure the injured employee population receiving treatment with the providers whose scores reflect their affiliation with the best recovery and return to work outcomes. Look at the outcomes at a state level, customer and office level so that claims adjusters, case managers and other stakeholders in care direction can be notified of improvement needs.

Be sure that provider performance measures include important claims management components such as lower rates of litigation, faster return to work and lower expense area claims cost. Our experience shows that factors such as faster return to work, less opioid prescriptions and fewer litigated claims are some of the best value areas affiliated with providers with proven good outcomes.

Make sure the provider information is valid. Be sure system flags or symbols identify providers with recently validated information. Sending an injured worker to the incorrect address of a good provider is a terrible outcome wrought from good intentions. Make sure that the claims adjusters and nurse case managers using your provider search tools are accessing valid and verified provider information.

Please watch for future posts as we share more information on big data and predictive analytics. If these points raise more questions than answers, please leave a comment here or visit to learn more about what Sedgwick is doing in this space.

George Furlong, SVP Managed Care Program Outcomes Analysis
Managed Care Management | Sedgwick

Back to Blog
Back to top