This year was the first year that Decision Point has sponsored the Bridge to PopHealth East conference in Boston, MA. The two-day conference covered operational performance in risk-based contracts, integrating your predictive models into your clinical workflow, innovation in clinical care delivery in a value-based world, and implementing a population health management strategy in the transitional space between fee-for-service and value-based care.
For those who could not attend, here are a few insights and stories from the two-day show.
On Changing Physician Behavior
PCPs are strapped for time. According to Dr. Mike Fischer, a speaker at Bridge to PopHealth and a PCP and researcher part of NaRCAD, “54% of PCPs say they are burned out!” It’s really not a huge surprise once you learn that according to the Health Resources & Services Administration, the demand for PCPs is growing almost twice as fast as the supply. With PCPs being shuffled from patient to patient from sun up to sun down, there really isn’t a whole lot of time for learning new systems.
However, it is not impossible. Several speakers were successful in changing PCP behavior within their organizations, but only after a series of failed attempts. At the end of the day, the successful cases met two or more of the following criteria:
- Convinced PCPs that solution would significantly improve their quality of care.
- Improved the PCP’s efficiency and gave them back more time in the day.
- Incentivized PCPs with Starbucks gift cards…! Dr. Uli Chettipally turned to this approach in order to get ER docs to leverage an in-EHR decision support tool, which advised the docs on how to build the discharge plan for pulmonary embolism cases.
On Predictive Analytics’ Role in Population Health
Just about every traditional payer & provider as well as the ACOs in attendance touted the capabilities of their predictive analytics. They showcased dashboard after dashboard on the World Trade Center’s massive projector screens, however few presenters recognized that predictive analytics are only a means to an end. They are not the solution.
Analytics are only useful if they facilitate the rapid execution of a decision. One health plan, focused on improving its predictive accuracy, demonstrated their implementation of the predictive modeling framework from OHDSI; a non-profit health data science and research organization. They explained that the framework summarized below drives their “Predictive Model Factory.”
OHDSI Model
- Define the Target Population
- Define the Outcome
- Select Variables and Execute Model
I couldn’t help but think, “this is backwards!” Plans today are best suited by first zeroing in on an outcome that they have the capacity to change, and then identifying the target population that is most likely to deliver the desired result. Without using predictive analytics for an actionable purpose, predictions are a waste of time and resources.
Here’s our take on a three-step framework:
Decision Point Predictive Modeling Framework
- Identify an outcome that you have the capacity to influence.
- Measure the capacity of the channels & resources available to impact a population.
- Select variables and execute model to identify the optimal mix of members to target using resources available.
On the Importance of Behavioral Economics in Changing Member-Patient Behavior
Here at Decision Point, we empower plans to craft effective engagement with their members by leveraging a diverse in-house library of behavioral economics driven tailored call, email, and direct mail message templates. The verbiage within each unique template varies depending on the member’s clinical complexity, previous plan engagement, and their differing barriers to care. Gone are the days of believing that a blanketed message to a mass audience is going to move the needle, drive engagement, or further member loyalty.
Big-box consumer brands have been employing these strategies for years with massive success in the same areas that we strive for in healthcare (engagement, action). Netflix, for example, drives you to binge watch TV shows by tailoring its selection to your preferences and once you’ve chosen a show, keeps you watching by automatically queuing up the next episode in the series. A behavioral economist would say, this isn’t rocket science! This a classic case of offering status by tailoring content to you, and forcing you to opt out rather than opt in to watch the next episode
Not surprisingly, Mt. Sinai Health System has begun utilizing help from the world of behavioral economics to improve their patient engagement. The New York City based provider has seen care management acceptance rates increase just by reframing their care management outreach as an invitation to their VIP service. By tweaking the messaging for those specific members with low-status personal identities, Mt. Sinai provided high-risk patients with the feeling of prestige which made them more comfortable enrolling in the program. Though, for patients who already hold higher esteems, this type of messaging did not work as well.
The Mt. Sanai presentation affirmed that healthcare marketing & engagement teams should consider leveraging behavioral economics research when crafting offers to different cohorts of people with similar social and clinical backgrounds.
We are happy to have been able to sponsor this year’s discussion and look forward to bringing new insights and observations to the table next year. See you then!