Saeed, what does Decision Point do?
We are a healthcare engagement analytics company…essentially we help clients that are “at risk” organizations to improve performance, including STAR ratings. We do this by providing data driven insights to more effectively engage members and providers.
What type of data do you use to make these recommendations?
Well, taking better care of members is about emotionally involving them in their care. Information to help do this resides in data that plans already have available, i.e. utilization patterns, distance to doctors, if they are compliant with evidence based guidelines, do they call into the call center. We also seek to include information about their behavior as a consumer. such as their lifestyles, their access to technology, and so forth.
Claims data makes sense, everyone has that but the other data you mentioned, that can be harder to capture. Why does non-claims oriented data matter?
We develop predictive models that are unique for each client – specifically based on the demographics and variables of their population. Variables like exercise and technology access matter because — for example, exercise habits influence mood and access to technology demonstrates a way to contact them or invite them to participate in online communities with other members like themselves.
The predictive models then determine which members are at most risk?
Yes, yes they do but they can also determine a member’s barriers to desired behavior, and their likelihood of responding to and acting on health plan communications. For example, if we identified a diabetic member as high risk of non-compliance, found their primary barrier to compliance as health literacy, and determined that the member will likely respond positively to a combination of health coaching and mobile health initiatives, we would recommend outreach that directly addresses these findings..
Noreen, when you were working on the payer side of the house, how were you going about determining which members were in your at risk population?
We had teams of people doing mining of claims data and we were asking members to complete surveys. This made for more data but the sheer volume of data made it complex to accurately review and assess which members were at highest risk. It was very challenging to take into consideration all of the variables that impact each member. Taking data from so many disparate sources and bringing it together is a big challenge.
What made it (and continues to make it) it so challenging, specifically to STARS?
So much of the data is collected as surveys or in other non-standard formats. Members inherently are unique which creates a lot of variability and it is often difficult to interpret the relationships that exist between members and primary care physicians, specialists, facilities and the rest of their care team. Relationships are important because they can provide insights into utilization patterns, potential overlaps or gaps in care and how we can most effectively engage those members in their care.
What are Informatica and Decision Point doing together?
To optimize the predictive models, as Saeed described, it’s imperative to feed them as much data and as accurate of data as possible. Without data, insights will be missed… and insights are the path to discovery and to improving CMS STARS ratings. Informatica is the data integration company — we ensure that data is reliable), connected (from any source to any target) and safe (avoiding data breaches or HIPAA violations). Informatica is delivering data to Decision Point efficiently and effectively so that clients have access to the best data possible to derive insights and improve outcomes. Our technology also provided the Star team with a member profile which brings together that disparate data and organizes it into the 360 degree view of that member. In addition to fueling Decision Point’s powerful algorithms, this is a tool that can be used for ongoing insights into the members.
Excellent, how can readers learn more?
They can meet with us next week – email Noreen or Saeed to arrange time.
The post Optimizing Outcomes for Medicare Advantage appeared first on The Informatica Blog - Perspectives for the Data Ready Enterprise.