Population Segmentation: An Environmental Scan and Qualitative Analysis of Promising Approaches

In this project, Mathematica Policy Research will conduct a systematic scan of segmentation approaches currently used by health systems to separate high-need, high-cost patients into subgroups with common care needs. The team will start with an extensive review of the literature, building on The Commonwealth Fund’s segmentation bibliography, followed by in-depth interviews with 10 national thought leaders and 40 health care organization team members that use a range of approaches for population segmentation. While interviews will cover how entities stratify their entire patient population, most of the protocol will focus on how they segment the high-need, high-cost group and then target care management resources to patients within each subgroup.

Analysis of real world examples will help health care systems (including newer ACOs), large multispecialty groups facing down-side financial risk, and practice networks better segment their high-risk patients and effectively target resources to those subgroups. Findings will offer guidance on challenges faced by providers developing and/or implementing segmentation and targeting strategies.