Experts agree that proper population health management strategies are the key drivers to better member outcomes. Done right, they also improve a health plan's confidence in its IT investments.
EBA spoke with Teri Mullaney, VP of healthcare strategy, and Amy Salls, director of healthcare analytics, at DST Health Solutions, a provider of health technology services, about the role of analytics in PHM.
How is data incorporated into population health management?
Salls: The challenge is to integrate a wide variety of transactional information to support population health monitoring. You can aggregate this information and identify, for example, within your population and ask are they getting all the services that would be recommended for them? That could be things like preventive care - immunizations, cancer screenings, all of those preventive things that we want our individuals to adopt.
At a more granular level, we can identify patients that have certain disease states and determine if they're getting all of the services that would be recommended for them. We can also use it to help us describe an at-risk population. Risk in this context could be risk of future costs or risk of certain high-utilization events like in-patient admissions.
Once we have the ability to identify that population, we use the predictive modeling tool from Johns Hopkins University, called ACGs [Adjusted Clinical Groups]. That allows us to identify that at-risk population in order to offer them specific programs to improve their health.
Those interventions might include:
* Identifying patients that don't seem to have a strong relationship with a primary care provider and helping refer them into a medical home.
* Identifying patients that are really best treated by specialists, help them find referrals to appropriate, cost-effective specialists.
* Providing them with some online research or handouts related to their condition to make sure they understand their primary condition.
* Helping them clarify their benefits and how they can best utilize their benefits related to their condition.
* Helping them identify their personal health goals, whether that's managing chronic pain, managing their obesity, improving their mental health, cessation of smoking.
* Differentiating their copays in order to incent the appropriate behavior around medication use, healthy habits and preventive care.
* Differentiating their copays if they select a provider in the appropriate network.
So there are a number of types of interventions that can be performed at a population level that fall short of that one-on-one intensive case management.
Mullaney: Amy brings up a very good point and I think it's very pertinent. The industry's changed. Before, interventions were pretty singularly focused and a lot of that revolved around someone sitting in a call center and doing outreach to the member to talk to them about their disease state.
I'm not saying that there's not a role for those classic disease management programs, but [now] you see things like benefit design based on their ability to meet certain goals. All of those types of flexible benefit designs have given the industry a greater ability to actually facilitate change.
The one-angled disease management programs of the past, they were partially successful with those folks who were already motivated to change, but I think that it wasn't enough support for the rest of the population.
Is it a challenge to get this data?
Salls: It can be. [For example, ] if medication use is critical to managing a population eliminating copays improves your data collection. If your copay is $10 and members choose to go to Walmart and pay $4 for medication you lose the visibility and ability to capture the data of that transaction. It's no longer on benefit. So sometimes to get data it's worth paying the higher copay for that drug to make it go onto your benefit system.
There's some information that we would prefer to get that provides a level of granularity beyond what you can get from a transaction. And that would include a third-party based health risk assessment. What's the member's own perception of their health and are there things they would like to change about their health?
That gets toward more of the health coaching. Just because someone is obese doesn't mean they're ready to change. So those health risk assessments can help with more of the behavioral part of health risk.
What are your future projections for the field of PHM?
Salls: The idea of touching every member in some way is where we are heading, but then really stratifying down to those that need the greatest level of support and intervention and informing them with the most data possible is really going to drive consistency in terms of some metrics and evolve these new data sources - the health risk assessments and the biometric data information - about behavioral characteristics and readiness to change, all of that will become part of our daily information set.
Mullaney: I agree. Those are the types of things that are going to drive transformation in this industry. I think that the collection of things like biometric data from traditional sources, but also non-traditional sources. There are a lot of capabilities out there. Blood pressure monitoring for example, glucometers, all of those things can digitally transmit information.
But it's about how do you gather and integrate it with other types of information, traditional claims data, lab data, PBM data. How do you integrate that with an EHR? I think that that's what's going to drive change, the appropriate integration of all of these data sources.
In the past at least we've looked at information but we didn't have all of it together. To Amy's earlier point, so often we don't have critical lab information. Now we're getting that additional data regularly. I think it's finally coming together.
