UCSF

Abstract Homebound patients who could benefit from high-quality, cost-saving longitudinal home-based medical care cannot be identified easily by hospitals, health systems, or payers. Further, without a well-defined population denominator, assessments of care quality are inadequate. We addressed these gaps using the OptumLabsTM Data Warehouse, which includes more than 3 million commercial and Medicare 2014 enrollees age 65 and older. We identified two patient phenotypes who may benefit from home-based medical care: (1) patients with complex comorbid conditions already receiving home-based care (>2 in-home physician visits (N=30,251); and (2) patients receiving substantial acute care (>57% with >1 hospitalization, ER visit, or ambulance service) and low levels of ambulatory services (19.2% with <2 ambulatory visits/year); (N=171,894). This project revealed a high-need patient population for whom home-based care may be beneficial and established a method for using administrative data to identify patients who are either homebound or would benefit from coordinated home-based services.