Preventable acute care spending for high-cost patients across payer types

Background: Healthcare expenditures are shown to concentrate in a small percentage of individuals. Many of these expenditures are thought to be preventable. Programs have developed to target high-cost individuals with the goal of reducing cost. Two of the underlying assumptions of these programs, degree of persistence and share of preventability costs, have lacked rigorous empirical research to inform payers about the general prospects. The purpose of the study is to quantify preventable expenditures among high-cost individuals across three plan types (Medicaid, Medicare Advantage, and commercial insurance plans) in Oregon. Methods: A retrospective longitudinal analysis of claims data was conducted. Shares of acute care expenditures considered preventable were calculated for non-high cost, episodically high cost, and persistently high cost patients. The results are shown for 74,717 Medicaid, 768,865 commercially insured, and 158,503 Medicare Advantage adults from Oregon using data from 2011 to 2013 data from the State of Oregon's All Payer All Claims (APAC) database and Medicaid data from the Oregon Health Authority. Results: In 2012, high cost patients account for 61.8% of Medicaid, 69.1% of commercial, and 60.0% of Medicare Advantage inpatient expenditures. Preventable inpatient expenditures accounted for 11.8%, 4.6%, and 10.0% of inpatient spending for persistently high cost patients in Medicaid, commercial and Medicare Advantage programs. Rates of preventable ED spending for persistently high cost patients in the Medicaid, commercial, and Medicare Advantage programs were 44.7%, 38%, and 34.1% respectively. Mean reversion led to declines of 11%, 25.6%, and 30.6% in the third year of spending among persistently high cost patients in the Medicaid, commercial, and Medicare Advantage programs. Conclusions: Potentially preventable health care spending for high cost patients accounted for less than 6% of total spending. More evidence is needed to support programs that target superutilizers, as opposed to disease-conditions, as a way of reducing total health care spending.