Does Medicare Advantage Affect the Way Primary Care Practices Deliver Care?

Introduction

With enrollment growing in Medicare Advantage (MA) plans, the private alternative to traditional fee-for-service Medicare, it is important to understand how they affect the health care delivered to their enrollees. While evidence largely indicates that access to care is similar for beneficiaries in MA plans and traditional Medicare, the data around quality of care has been mixed. Little information is available about how the type of coverage affects social drivers of health, care coordination, and other components of care.¹

To incentivize the provision of efficient, high-quality care, the federal government makes capitated payments to MA plans, meaning payments are not directly tied to enrollees’ use of services. In contrast, there is no limit on government spending on beneficiaries in traditional Medicare and few incentives to provide efficient care. MA plans may, in turn, either directly or indirectly pass these efficiency and quality incentives along to providers, such as by enabling care to be coordinated more easily, sharing financial risk with providers, and reducing the time they spend on administrative tasks. MA plans may be more likely to pass incentives along to primary care physicians (PCPs), who often help prevent costly, long-term health issues. PCPs also provide referrals to specialists, and MA plans that work closely with PCPs may be able to ensure that their enrollees are referred to less costly in-network providers.

Prior research suggests that MA plans’ incentives for providers may substantially spill over from MA patients to also affect the care and costs of patients in the same clinic covered by traditional Medicare.² MA plans could also decrease the amount of time PCPs can spend with all their patients if the plans’ utilization management techniques, such as prior authorization or billing requirements, are more time-intensive.

This data brief examines whether PCPs who primarily serve MA patients report differences in care delivery, care coordination, and administrative burden compared to those who mostly serve traditional Medicare patients. The analysis relies on United States–specific data from the Commonwealth Fund’s 2022 International Health Policy Survey of Primary Care Physicians, involving 453 PCPs. We also used 2021 Medicare claims data to classify PCPs into three groups (see “How We Conducted This Study” for details):

• PCPs who predominantly serve traditional Medicare patients, meaning 60 percent to 100 percent of their Medicare patient panel was covered by traditional Medicare.
• PCPs who predominantly serve MA patients, meaning 60 percent to 100 percent of their Medicare patient panel was covered by MA.
• PCPs who had a patient panel with a mix of both types of coverage, meaning fewer than 60 percent traditional Medicare but more than 40 percent MA.

We analyzed survey responses with a specific focus on questions about care practices that MA plans could be incentivizing or otherwise encouraging. We discuss how aspects of care reported by PCPs vary across these different patient panel mixes. All respondents included in this analysis had Medicare patient panels greater than 10 people to minimize the likelihood of responses reflecting PCP experiences with non-Medicare patients.

Key Findings

In our analysis, most primary care physicians saw both MA and traditional Medicare patients. One in 10 PCPs saw only MA enrollees; these likely include physicians employed directly by an insurance plan, such as Kaiser Permanente, and possibly physicians in areas with high MA enrollment, such as Miami-Dade County. One in 20 PCPs saw only traditional Medicare patients, and about one in four PCPs saw similar shares of MA and traditional Medicare patients.

Overall, PCPs who predominantly saw patients in traditional Medicare had similar demographic characteristics to the PCPs who saw more MA patients. PCPs who mostly saw MA patients were significantly more likely than those who mostly treated traditional Medicare patients to be working in community health centers or community clinics and in practices recognized as a patient-centered medical home or an advanced primary care practice (see the table in “How We Conducted This Study”).

Management of care for patients with chronic conditions was similar among PCPs who mostly saw MA patients and those who saw mostly traditional Medicare patients. However, 52 percent of PCPs treating mostly MA patients said they recorded their patients’ self-management goals in the medical record, significantly more than the 40 percent of PCPs treating mostly traditional Medicare patients.

Compared to PCPs with mostly traditional Medicare patients, a larger share of PCPs seeing predominantly MA patients reported that they received notifications when their patients were seen in other settings, including when they received after-hours care, went to an emergency room, or were admitted to a hospital. MA plans could be helping to facilitate these notifications.

PCPs with a mix of Medicare patients reported waiting significantly longer to be notified of their patients’ discharge from a hospital than those with mostly traditional Medicare or mostly MA (data not shown). One in four PCPs seeing a mix of patients reported waiting five to 14 days to receive necessary information after their patient had been discharged — nearly double the number of PCPs who saw mostly traditional Medicare patients. Overall, about half of all PCPs reported receiving the necessary information within 48 hours following discharge (data not shown).

Less than half of all PCPs said they regularly screen their Medicare patients for specific social needs, such as food security or transportation, with few differences between PCPs with mostly MA patients and those with mostly traditional Medicare patients. PCPs with more MA patients did report significantly higher rates of screening for domestic violence than those with more traditional Medicare patients, but PCPs across the board did not report high rates of screening for specific social drivers of health. No more than five in 10 PCPs reported that they “usually” or “often” screen their patients for social needs, regardless of their Medicare patient mix.

While screening for the social needs presented in the exhibit above are generally low, 62 percent of PCPs across all Medicare patient mixes reported screening their patients for at least one social need more than half the time (data not shown).

PCPs face both internal and external challenges to addressing their patients’ social needs. Internal challenges were flagged as major problems across different PCP groups, including inadequate staffing to make referrals and coordinate care as well as the amount of paperwork involved in coordinating with social services.

More than three in 10 PCPs, regardless of their Medicare coverage mix, said they experienced problems coordinating patients’ care with social services, including lack of a referral system, follow-up from organizations, or information about services in their community. PCPs who saw more traditional Medicare patients were as likely to report these problems as those with more MA patients.

The time PCPs reported spending on various administrative tasks — including coverage issues, quality reporting, visit documentation, and referral coordination — did not differ significantly by mix of Medicare coverage in their patient panel. Dealing with insurance claims issues and getting patients needed medications or treatments were cited as a major problem by most PCPs in this analysis, and more than one in four said quality of care data reporting and coordinating specialist referrals were major problems for their practices. Time spent documenting patient care or visits was cited as a major problem by more than four in 10 PCPs, with rates not differing significantly across practices’ patient coverage mixes.

Discussion

Medicare Advantage plans receive capitated payments to provide plans the flexibility to help manage patients’ care. This analysis shows that across many measures of care coordination, there were no significant differences in responses based on the mix of Medicare coverage among a primary care physician’s panel of patients. Physicians who predominantly saw MA patients reported being notified regularly about their patients receiving care in other settings. Primary care physicians who mostly saw MA patients also reported the highest rates of screening for domestic violence issues, but there weren’t significant differences by patient mix in reported screening for other social needs. Overall, the majority of PCPs did not report regularly screening for social needs such as housing or food insecurity.

The lack of significant differences between PCP groups on most of the study’s measures suggests either MA plans have limited influence on care delivery or there are sizeable spillover effects — meaning that even a comparatively low number of MA patients may be influencing how a PCP functions, impacting all their patients. The findings also show the limited influence of MA plans in addressing key challenges that providers face, as one insurer typically constitutes only a small slice of a physician’s insurance panel. It may be that only large insurers that dominate markets can influence primary care. PCP samples for each group in this study are also relatively small, which means only large differences between the groups were statistically significant.

As MA enrollment continues to grow, the effects on primary care practices will be important to monitor. If plans are providing resources to PCPs to manage their patients more efficiently, this could have positive effects for all patients in the practice. If plans are not enabling efficiencies, however, or plans are adding onerous requirements that make it more difficult to care for patients, then it could have a negative impact on physician care. A greater understanding of plans’ support, incentives, and requirements on physicians’ practices will shed light on their effects on patient care for all Medicare beneficiaries.