Assessing the Short-Term Association Between Rural Hospitals' Participation in Accountable Care Organizations and Changes in Utilization and Financial Performance.

Abstract

Although much research has been done on accountable care organizations (ACOs), little is known about their impact on rural hospitals. We examine the association between rural hospitals' participation in an ACO and their performance on utilization and financial measures. This quasi-experimental study estimates the relationship between voluntary ACO participation and hospital metrics using propensity score-matched, longitudinal regression models with year and hospital fixed effects. Regression models controlled for secular trends and time-varying hospital and county characteristics. Hospital measures were from the American Hospital Association, RAND Hospital Data, and Leavitt Partners. The initial population comprises 643 rural hospitals that participated in an ACO for at least one year during the 2011 to 2018 study period and 1,541 rural hospitals that did not participate in an ACO. From this population we created a sample of propensity score-matched hospitals consisting of 525 ACO-participating and 525 comparable non-ACO hospitals. Rural hospitals' participation in an ACO is not associated with changes in hospital utilization or financial measures, nor is there an association between these performance metrics and whether another within-county hospital participated in an ACO. A secondary analysis limited to Critical Access Hospitals provides some evidence that inpatient utilization increases in the second year of ACO participation, though the increases are not significant in year 3 and beyond. We find no evidence that rural hospitals experience substantive changes in outpatient visits, inpatient utilization, or operating margin in the years immediately after joining an ACO.