Accountable care organizations are supposed to be a new alternative vehicle for groups of providers to coordinate care better for a population of patients, improving quality and reducing costs. The Medicare program has a number of ACO variants in operation and private payers have also adopted the approach. Research is beginning to focus on the results and reality of ACOs. Researchers in a study published in Health Affairs examined the persistence of physician and beneficiary participation for Medicare ACOs. (HA Article) The authors used data regarding the Partners Healthcare Pioneer ACO, one of the largest of Medicare’s ACOs. This eastern Massachusetts health system was part of the initial three-year contract from 2012 to 2014. It had a set of physicians initially identified as participants and the authors tracked changes in this list. One of the truly dumb features of the Medicare ACO program is that beneficiaries aren’t enrolled in the ACO but after-the-fact have their care and spending “attributed” to it. But most of the beneficiaries in an ACO are there because their primary care physician participates.
748 primary care physicians were listed as affiliated with the ACO during at least one of the three contract years, but only 661 had even one attributed beneficiary. The average number of beneficiaries attributed to a primary care doctor was 70. This typically represented less than 5% of a doctor’s practice. Hard to change behavior when so few patients are affected. Only 52% of the doctors remained affiliated with the ACO for all three years. Those doctors who were in the ACO for an extended period of time were more likely to be employed by it or to have some other strong affiliation. I.e., they were likely told to participate. Attributed beneficiaries tended to follow their primary physician in and out of the ACO, particularly ones with higher spending. The fundamental problem faced by accountable care organizations is that physicians in particular are looking for pretty easy and quick ways to maintain or improve their compensation. They aren’t interested in a lot of work flow changes and additional administrative hassles. And the health systems that increasingly employ them don’t want to create a lot of dissatisfaction by forcing participation in ACOs. All these issues are exacerbated when Medicare patients who might be in an ACO form a relatively small portion of a physician’s patient panel.