WASHINGTON — Medicare’s accountable care organizations (ACOs) are helping the program save money, but they need some tweaking, members of the Medicare Payment Assessment Commission (MedPAC) said.
“I was actually pleased with the results, given the short time period ACOs have been in existence,” commission member Warner Thomas, of Ochsner Health System in New Orleans, said at a commission meeting here Friday. “I hope people feel like we’re heading in the right direction.”
The commissioners heard a report from MedPAC staff about a specific type of ACO within Medicare, known as the Medicare Shared Savings Program. Under that program, ACOs that achieve a cost of care below a defined benchmark are able to reap a share of the savings.
The commission staff found that spending on ACO beneficiaries was about 1.2% below the benchmarks set in 2017; shared savings payments amounted to about 0.8% of benchmarks. The staff also compared the growth in spending for beneficiaries assigned to ACOs to spending growth for other Medicare beneficiaries, and found gross savings of 1.1% to 1.2% of Medicare spending from 2013 to 2015, after adjustments in spending for changes in risk scores.
However, the staff also uncovered some differences among ACO patients. For one thing, beneficiaries who either switched into or switched out of an ACO ended up costing Medicare more money. Beneficiaries who switched ACOs in 2013, 2014, or 2015 had 1.2% higher spending growth compared with the average in the Medicare marketplace, while those first assigned to a newly formed ACO in 2016 had spending growth that was 2.1% higher. And those who were assigned to an existing ACO in 2016 had 16% higher spending growth.
In addition, while beneficiaries assigned to the same ACO from 2013 to 2016 had spending growth rates 10% below the average in the market, those who stayed in the same ACO from 2013 to 2015 but left in 2016 had spending growth rates 13.8% higher than average, noted Jeff Stensland, PhD, a principal policy analyst at MedPAC.
“These beneficiaries had the benefit of care coordination the ACO provided [those first 3 years], but something happened in 2016: most likely they changed physicians they saw, possibly due to a change in health status,” said Stensland. “This tells us there’s an association between changes in assignment and changes in spending. For example, a beneficiary may fall ill and start to use a new set of physicians.”
Another thing complicating the study results is the way the patients were assigned to the ACO. Patients are generally assigned to an ACO if they have gotten primary care services from a physician in that ACO, although if they don’t have a lot of primary care services initially, they can be assigned based on having had a service from a specialty physician in the organization. The question then becomes when the patients are considered members of the ACO.
Some patients are “retrospectively” assigned — that is, the ACO is held responsible for the patient’s health costs in the same year they join the ACO, but the ACO doesn’t find out that the patient was assigned to them until the end of that year — while others are “prospectively” assigned, in which case the ACO is responsible for spending on the patient in the year after they are assigned to the ACO. The prospective assignment method has several benefits, since the ACO physician will have seen the patient already and the ACO will know at the beginning of each year which patients they’re responsible for.
Commission member Brian DeBusk, PhD, of DeRoyal Industries in Powell, Tennessee, said that retrospective attribution seemed to be “one of the serious flaws” with the ACO program. “It seems like prospective attribution solves a lot of those problems,” he said.
Commissioner Dana Gelb Safran, ScD, of BlueCross BlueShield of Massachusetts, said her organization uses a “concurrent assignment” method for its own ACOs. “[The ACO] knows all through the year who [the plan] thinks is attributed to them, but at the end of the year we ‘settle up’ on who is manifest as their patient, because of all the switching that happens,” she said.
Commissioner Bruce Pyenson, of Milliman in New York City, expressed some reservations about ACOs. “This work identifies one of the prevailing myths of population health that many hoped would be realized with the ACO movement — that if somehow we engaged physicians and patients better and got them into the system consistently, we’d be able to bring magic of better care to them and they’d be healthier and less expensive,” he said.
“We often attempt to blame the patient — if patients were only more compliant, or only indentured servants and didn’t move around, but the reality is that it is not the patient who decides on unnecessary surgery and admissions, and excessive stay in a [skilled nursing facility] and that sort of thing,” he continued. “We’re up against a failure to take the kind of steps [effective in the short term] at saving money.”
But Pyenson explained that incentives are not aligned to make that happen. “Hence we’re in this awkward situation of a program that seemed so promising but is disappointing,” he said. “There is a potential for things to change here — the ability to direct care more strongly, I think, would be a very important tool to get at underlying potential savings in the short term.”
DeBusk suggested that Medicare may have to do more to engage beneficiaries in the ACO process, perhaps a surcharge for people who insist on “unmanaged care.”
“At some point, we have to recognize that people who insist in not participating in any of the choices in front of them are costing the system money,” he said.