Connecting doctors, patients and payers — MedPAC digs into MA physician networks

Nearly one in five Medicare Advantage (MA) visits with a primary care physician involves a doctor who isn’t listed in the patient’s own plan directory, according to a new federal analysis of 2023 data.

What exactly does that mean for physicians and patients? Members of the Medicare Payment Advisory Commission (MedPAC) were divided, with some calling the mismatches an error rate that health plans must fix. But others warned that missing directory listings are not necessarily a barrier to care.

This month, MedPAC staff presented the analysis at the opening session of the commission’s two-day meeting. MedPAC advises Congress on Medicare payment policy; the session was for informational purposes and did not prompt a new recommendation from the commissioners. The presenters were Katelyn R. Smalley, Ph.D., M.Sc., senior analyst; Krista Cherry, senior research assistant; and Alice Stricker, special assistant. Amol S. Navathe, M.D., Ph.D., MedPAC chair, led the session.

The staff research started with two basic premises about the importance of the directories. MA plans, which are required to maintain the directories, can use provider networks to manage care quality and costs. MA beneficiaries also consistently say they use directories because having a choice of physicians and other clinicians is important for their health care.

Danielle Pierotti, Ph.D., R.N., was among the commissioners least willing to accept the explanations for the gaps.

“There’s solutions to every single explanation in here, and I just really need to advocate that these are explanations and points for the plans to correct as performance improvement, they are not reasonable excuses that we should be accepting and normalizing. It’s an error rate. If Amazon had a 9% error rate in your deliveries, you wouldn’t be ordering from Amazon anymore. I think that this is a little bit more impactful,” Pierotti said.

Commissioner Kenny Kan, F.S.A., C.P.A., C.F.A., M.A.A.A., offered a different view.

“MA plans must be accountable for ensuring timely access, but we should be careful not to equate directory inaccuracies with barriers to care. Equally important, we should preserve the value of managed care in coordinating care, improving quality, and maintaining affordability,” Kan said.

What did MedPAC staff find?

Staff compared two sources. Provider directories are meant to show which physicians have contracts with a plan. MA encounter data, the records plans submit for services delivered, show which physicians actually treated MA enrollees. The analysis used 2023 data, the most recent year of encounter data available, and focused on physicians actively billing Medicare in local health maintenance organization (HMO) and preferred provider organization (PPO) plans.

By directory, 93% of primary care physicians and 88% of specialists appeared in at least one MA plan’s directory. By encounters, the rates were even higher: 97% of primary care physicians and 98% of specialists, counting physicians whose MA encounters made up at least 5% of their Medicare claims volume.

Staff stressed the analysis is descriptive and takes no position on the right level of participation. Most physicians, 88.4%, showed up in both sources. But 9.4% treated MA enrollees without appearing in any MA directory. If beneficiaries looked only at directories, those physicians would seem not to participate in MA at all, Smalley said.

The bigger gap appeared when staff checked whether the physician who rendered care was listed in the enrollee’s own plan directory. In February 2023, 32% of MA encounters involved a physician who wasn’t, Krista Cherry said. The share was 18% for primary care physicians, 46% for psychiatrists and 72% for specialists patients generally don’t choose, such as emergency medicine physicians, radiologists and anesthesiologists.

Is it bad data or out-of-network care?

Staff said a non-directory encounter could mean the directory was incomplete, out of date or otherwise wrong. It could also mean the physician was out of network for that patient. Evidence on how often MA enrollees get out-of-network care is limited, Krista Cherry said, citing one study that estimated 15% to 22% of MA claims were out of network.

Part of the problem lands on practices. Physicians must send updated information to every plan they contract with, and some have little incentive to do so, according to the staff presentation. Networks also shift during the year.

Commissioner R. Tamara Konetzka, Ph.D., asked whether timing could explain some gaps, such as a physician joining a network late in the year. Commissioner Gokhan Metan, M.Sc., Ph.D., NACD.DC, asked whether billing provider and tax ID data could show that a claim was in network even if the individual physician wasn’t listed.

Staff said they are working on both, plus an HMO-versus-PPO breakdown. That split matters because out-of-network care is a covered benefit in PPOs but mostly not in HMOs, Metan said.

Commissioner Cheryl L. Damberg, Ph.D., said the data themselves are part of the problem.

“But the fact that the encounter data doesn’t contain a flag that helps you easily identify that, I think, is a problem,” Damberg said. She said MedPAC needs access to in-network and out-of-network flags in MA encounter data.

Even with more data, some cases will stay unexplained, Smalley said.

“I think the short answer is that it’s still really tricky,” she said. “There will be a certain number of these things that we’re not going to be able to explain the source of the discrepancy.”

Why does primary care stand out?

The 18% mismatch for primary care was lower than for many specialties. Commissioner Joshua Liao, M.D., M.Sc., said it may matter most, because patients choose their primary care physician and that physician coordinates much of the rest of their care.

“So even if you rank the mismatch rates, 18% is not near the top. But to me, it’s very consequential for that reason,” Liao said. “And actually, if you turn the prism any which way, it matters anyway.”

Whether the cause is a missing listing or care outside the network, both are worth knowing, he said.

Commissioner Robert A. Cherry, M.D., M.S., cautioned against reading too much into high participation rates.

“I would use some caution in saying that, well, because the participation rate is high, it doesn’t mean in that local catchment area that there is network adequacy,” Robert Cherry said. Physicians may sign up heavily with one plan that pays better and not with others in the same market, he said.

What about NPs and PAs?

MedPAC included primary care nurse practitioners (NPs) and physician assistants (PAs) in an earlier analysis in March 2026, but staff dropped them this time. Inconsistent recording of NPs and PAs in directories and other data, including through incident-to billing, which lets their services be billed under a supervising physician, would likely skew the primary care picture, Smalley said.

Several commissioners pushed back on leaving them out. Konetzka said the work is ultimately about beneficiary access to care and asked staff to address the issue prominently in the final chapter. Pierotti said NPs and PAs are a growing share of the primary care workforce.

“Increasingly, the proportion of primary care provided in rural geriatric and other safety net environments is being provided by nurse practitioners and physicians assistants,” Pierotti said.

MedPAC staff noted the commission recommended in 2019 that Medicare require direct billing by advanced practice registered nurses and PAs. At that time, MedPAC estimated roughly a third of their care was billed incident-to, the staff member said.

Commissioner Gina Upchurch, R.Ph., M.P.H., said beneficiaries need to see those clinicians on Medicare’s online comparison tool.

“We need the NP and the PA listed on the plan finder tool, so people can understand whether providers are going to be in network with them,” Upchurch said.

Can plans keep directories current?

Commissioner Thomas Diller, M.D., M.M.M., said directories for hospital-based physicians are especially hard to maintain because of turnover and locum tenens staffing.

“So trying to keep up with the directory is an almost impossible task,” Diller said. He suggested safeguards ensuring that physicians at in-network facilities are treated as in network. Pierotti agreed, describing surprise bills from out-of-network radiologists or pathologists at in-network facilities as a “bait and switch.” Commissioner Benedic Ippolito, Ph.D., cautioned that tying physician network status to facilities has been contentious in federal surprise billing debates.

Other commissioners said technology removes old excuses. Metan called the current level of directory accuracy “mind blowing” in the age of artificial intelligence (AI). He suggested MedPAC study policy options, including aligning contract cycles with plan years, centralizing provider directories, counting directory accuracy in plan quality ratings and using AI to maintain directories.

Plans historically said they relied on physicians to keep them informed, but that is no longer so much the case, said Commissioner Scott Sarran, M.D., M.B.A.

“I think there are probably AI enablement approaches that would help plans do a much better job, should they should they choose to or be regulated to do that,” Sarran said.

Damberg pointed to California, where she said a newly signed law, AB 280, directs the state to designate a statewide central utility for provider directory data sourced from providers.

“Consumers often just want to look up their clinic or the post acute care rehab facility, not the individual provider,” Upchurch said.

What happens when networks change midyear?

Liao asked what happens to patients when a physician leaves a plan midyear. MedPAC staff said plans must notify affected enrollees, who can ask CMS for a special enrollment period (SEP) to switch plans. CMS decides whether the change is significant enough to warrant one. Staff said they don’t know how often that happens but suspect it is not often.

Separate continuity-of-care protections exist, but staff know little about that transition period, they said in response to Sarran.

Commissioner Stacie Dusetzina, Ph.D., said midyear network changes seem “completely unacceptable” and are “a low-hanging fruit thing to fix for beneficiaries.” Upchurch noted that switching plans midyear resets a patient’s out-of-pocket spending, which she said can reach $9,850 for in-network care in 2027. She called for two SEPs: one when a contract dispute first becomes public and another after the outcome is known.

What comes next?

Staff plan to return in the spring with refined estimates and analyses of other provider types, aiming for a chapter in MedPAC’s June 2027 report to Congress. Navathe said the goal is to help beneficiaries make informed choices and be protected in those choices, not to issue recommendations yet.

“This is still kind of a fact finding mission workstream for us,” Navathe said.