Some policy researchers are optimistic that the five-times-larger CPC+ successor program includes enough tweaks to avoid CPC’s failings.
June 1—Improving the practice of medicine at the primary care level by deploying massive, publicly funded incentives is complicated, but doable on the margins.
That’s the thin upside from “The Comprehensive Primary Care Initiative: Effects on Spending, Quality, Patients, And Physicians,” a report from Mathematica Policy Research published in the June issue of Health Affairs.
The fatter downside is that the federal government, after spending hundreds of millions of dollars on the Comprehensive Primary Care (CPC) model, failed to find the secret sauce that can produce an ROI.
In other words, the care and incentive paradigms of this first iteration of CPC, kicked off in October 2012 and lasting four years through 2016, cost Medicare far more money than it saved in promoting lower costs of care, one of the key goals of the CPC initiative. And yet, CMS is more than doubling down with its larger successor, the Comprehensive Primary Care Plus (CPC+) model, launched in 2017.
“Results from the CPC illustrate how hard it is to reduce Medicare spending,” the study’s authors observed.
Their findings were unsurprising to some policy researchers.
“This fits in with the other studies out there on team-based care and the patient-centered medical home model,” Michael Barnett, MD, assistant professor of health policy and management at Harvard University’s T.H. Chan School of Public Health, said in an interview. “It has been pretty consistent—modest-to-no-effect on spending and quality and largely stable physician satisfaction.”
Medicare and other insurers paid providers at nearly 500 practices $479 million in care management fees through CPC, with more than half (58 percent), coming from the Medicare fee-for-service (FFS) program, according to Mathematica. In 2016, the final year for CPC, the program paid an average of $50,189 per clinician and $179,519 per practice, roughly an average of 10 percent of their revenue that year.
Shared savings payments by Medicare FFS, available in the final three years of CPC, totaled about $24 million.
Model Problems
One problem with ROI under the program was its exclusive focus on primary care practices.
Among hospitals, during the study period, “Their incentive was still heads in beds,” Deborah Peikes, a senior fellow at Mathematica and lead author of the CPC report, said in an interview.
Peikes pointed out that any CPC physician group that might have wanted to join a Medicare accountable care organization to better align itself financially with other healthcare entities would have been required to drop out of CPC.
“Until the other providers have some skin in the game, I’d be surprised whether primary care [alone] could make a difference” in cost containment, Peikes said.
Other possible barriers to better results—financial and otherwise—included the sheer burden of reporting, the lack of interoperability of health information technology systems, and the “limited actionability of data feedback,” the report said.
Still, researchers highlighted several hopeful findings.
By the final year of the program, CPC groups provided more timely follow-ups with patients after their emergency department (ED) visits or discharges from the hospital than occurred in comparison groups.
CPC groups also slowed the growth in their patients’ ED visits and hospitalizations by 2 percent each compared to their peers. CPC changes did not appreciably worsen (but neither did they improve) physician or patient experiences, or practice performance on several quality measures, such as frequency of eye exams and tests measuring HbA1c and urinary protein levels.
CPC+ Impacts
The collaboration between CMS and private payers, initiated under CPC, should pay greater dividends moving forward with its successor, CPC+, Peikes said.
“It’s almost a revolution for CMS to sit down with a bunch of other payers and align on payments, data sharing, and care delivery requirements,” she said. “In CPC+, they’re pushing to align on quality measurement. The more aligned the signal and the higher percentage of patients involved, the stronger the signal.”
At the start, CPC practices were saddled with software tested and certified to meet the Department of Health and Human Services’ (HHS’s) rudimentary 2011 Edition criteria, which didn’t match well with the population health, interoperability, and customer relationship management functions that CPC practices needed. The result was predictable complaints from CPC providers about IT barriers.
CPC+ has addressed some of those complaints, according to David Dorr, MD, chief research information officer and a professor of informatics at Oregon Health & Science University.
In 2017, 87 percent of practices in CPC+ could access data about their patients’ hospital admissions, discharges, and transfers within one day of occurrence, according to a four-page, “Year in Review” of the program by the CMS. Eighty-five percent of PCP+ practices had IT systems that helped groups with patient risk stratification, the agency claimed.
“CPC+ has a lot more emphasis on getting the right information in front of the right people, either through HIE [health information exchange] or their local system,” Dorr said. To do it, CMS extracted commitments from “a long list” of EHR vendors, which pledged to configure their systems to meet the technical challenges posed by the program, he said.
As a result, “the health information technology has improved, but it’s not without issues,” Dorr said.
In both its review and a one-page financial sheet on CPC+ outcomes thus far, CMS made no mention of ROI.
Barnett, the Harvard policy researcher, is unsure whether CPC+ will do any better at net cost reductions than its CPC predecessor did. And he’s okay with that.
“I’d like to move the goal posts away from, ‘If we’re not saving more money, we’re a failure,’” Barnett said. “We’ve sold that as the primary outcome, and I don’t think it’s going to happen. We’ve tried too many things that didn’t work. Can we do practice transformation that doesn’t cost more, but improves quality of care? That’s still a big win.”
While CMS is betting a bundle on CPC+—the program involves 2,876 primary care practices, five times more than in CPC, and CMS paid out $518 million in the first year, 2017, which is more than it did in four years with CPC—the agency already is looking beyond CPC+ for payment model ideas.
The Center for Medicare & Medicaid Innovation continues to solicit models for primary care, Dorr said. “We’re still in an area where people are trying to figure out the best approach,” he said.
Someday, some way, we’ll get the balance of incentives and savings right, Barnett said.
“Culture change is so hard; that’s why pay-for-performance has been so lackluster,” he said. “I’ll remain optimistic we’ll find the right set of models. For one reason, we have to change.”
Joseph Conn is a freelance writer based in Chicago. Follow Joe on Twitter at @JConnHIT.
