Berkeley-Dartmouth researchers develop taxonomy of accountable care organizations

December 16, 2014

waiting roomThere are more than 600 accountable care organizations (ACOs) in the United States working to improve the quality of care and reduce costs for a defined population of patients with associated financial rewards or losses. A major question is whether or not they share any commonalities that can be used to help target technical assistance efforts and with which to evaluate their performance. Based on the first National Survey (supported by The Commonwealth Fund - New York) of 173 Accountable Care Organizations, researchers at the UC Berkeley School of Public Health and The Dartmouth Institute have developed a taxonomy of ACOs, published in the December 2014 issue of Health Services Research.  

The taxonomy is based on eight attributes grounded in resource dependence and institutional theories of organizations and include number of FTE physicians;  the number of different organizations constituting the ACO; the scope of services provided; whether or not the ACO was part of an integrated delivery system (IDS); the percent of primary care physicians involved; whether it was physician, hospital or jointly led; the extent of performance management and accountability mechanisms used; and the extent of prior experience with payer reforms. The resulting three types included larger, integrated delivery systems that offer a broad scope of services and frequently include one or more post-acute facilities; smaller, physician-led practices centered on primary care and that possess a relatively high degree of physician performance management and accountability; and moderately -sized hybrid or jointly led groups that offer a  moderate scope of services with some involvement of post-acute facilities. 

Ninety-four percent of the integrated delivery system ACOs were correctly classified; 87 percent of the physician-led; and 76 percent of the hybrid. Preliminary predictive validity tests indicated that the large IDS ACOs were more likely to take on two-sided risk, and had higher care management capabilities, quality improvement capabilities, and electronic health record capabilities than the smaller physician-led and the hybrid-led.

"The taxonomy is a starting point for describing some of the commonalities among the diverse population of ACOs. Its ultimate utility will lie in its ability to examine and explain differences in performance over time," said Steve Shortell, lead author and professor of health policy and management and director of the Center for Healthcare Organizational and Innovation Research (CHOIR) at Berkeley. 

In an accompanying editorial, Larry Casalino, professor of health policy and management at New York Cornell Weill Medical College, highlighted importance of the taxonomy for conducting comparative organizational effectiveness research. The investigators plan to update and further validate the taxonomy as additonal ACOs are formed. For the full article and accompanying editorial see Shortell SM, Wu FM, Lewis VA, Colla CH, Fisher ES. A taxonomy of accountable care organizations for policy and practice. Health Serv Res. 2014 Dec;49(6):1883-99. doi: 10.1111/1475-6773.12234. Epub 2014 Sep 23.