Rx for Medicare value-based programs, Part 1: Applying the right data

Early reflections are starting to roll in from our clients participating in CMS Medicare Shared Savings pilots and Pioneer Accountable Care Organization (ACO) programs. Based on their feedback, one thing seems certain: providers are quickly turning their focus to data as the means to achieve scalable, sustainable success in these programs.

Most of these early stage participants have expressed interest in three areas of focus for improving how data is usedMiles Snowden, MD, MPH, CEBS Chief Medical Officer, OptumHealth to manage population health: they need the right data, at the right time, in the right format. It may sound easy, but the details are daunting. In this post, I’ll discuss how data is crucial to successful population health management as we shift from a volume-based reimbursement model to one based on value.

The right data is needed for usability.

  • Claims data is important for effective population health management. Electronic medical record (EMR) systems are useful for collecting, reporting and analyzing clinical data. However, most ACOs cannot rely solely on EMR data to coordinate patient care because they use “closed systems” that only collect data on services delivered by providers within their ACO network. As a result, analyzing claims data is often the only way to learn about care a patient receives from outside providers.
  • Historical claims information is needed to support effective population health management. Many ACOs use predictive modeling analytics to forecast the cost of caring for a specified patient population. By analyzing past claims data, the ACO can manage its risk by identifying gaps in care and designing appropriate intervention strategies.  To get the most value out of predictive modeling analytics, ACOs need access to a minimum of 18 months of historical claims data. Anything less renders predictive modeling  essentially impossible for the first year of the contract, exposing the ACO to an unknown level of risk. Predictive modeling is especially important when managing the health of senior populations, where costs driven by long-term chronic disease morbidity can be predicted using newer analytic engines.
  • Flexibility of data use is important for effective population health management. ACOs should be able to access as much CMS claims data as often as they deem important for successfully managing the health of their patient population. Giving providers unlimited access to a data warehouse – a central repository of integrated claims data from a variety of sources – offers more flexibility than offering periodic access to retrospective claims data. Allowing this type of flexibility would not only enhance patient care outcomes, it would also allow individual ACOs to test new models of data use and share best practice learnings on the use of claims data for population health management.

In my next post, I’ll discuss why being able to access data at the right time and in the right format may be just as challenging and important for successful population health management as having the right data.

Miles Snowden, MD

One thought on “Rx for Medicare value-based programs, Part 1: Applying the right data

  1. Pingback: Rx for Medicare value-based programs, Part 2: Data at the right time and in the right format | Healthcare Exchange

Leave a Reply