ABSTRACT
This chapter examines avoidable health care conditions in high utilizers in order to prevent hospital visits and reduce related health care costs. It also examines using data-driven approaches to identify the most impactable subpopulations of high utilizers who may not have the highest health care expenditures. The chapter addresses limitations to clinical diagnostic classification tools, such as the New York University Emergency Department (ED) profiling algorithm and 3M Potentially Preventable Events software, which identifies preventable conditions but suffer from a limited population scope and lack of transparency due to commercial considerations. It presents a novel approach to analyze variations in Medicaid health care expenditures based on using higher-than-expected values of the residuals from health care utilization adjustment models. Linear regression-based adjustment models have been used in health care capitation payments because they systematically account for spending associated with specific health care conditions. Inpatient acute care hospital and ED are some important types of health care services.
