Despite the overwhelming evidence of its benefits, primary care in the United States is chronically underfunded. Decades of research have documented the benefits of robust primary care in a well-organized health care system that maximizes patient and population health, efficiency, and equity. Patients value access to and a continuous interpersonal relationship with a trusted primary care practitioner (PCP), and this relationship is associated with better outcomes for patients. Unfortunately, the way we pay for primary care incentivizes visit volume more than quality of care, and these relationships have become increasingly challenging to secure, even among well-insured patients.
In response, state and federal policymakers have implemented various primary care-oriented value-based payment (PC-VBP) models to strengthen primary care through alternative forms of payment. These models often engage commercial health plans alongside Medicare and Medicaid payers to maximize the proportion of a PCP’s patients included in these models and thus the resources available to improve care delivery. Although many PC-VBP models have been introduced since 2010, fewer than half of PCPs participate in them, and those that do are often in practices with more resources. As a result, public investments intended to strengthen primary care are not reaching some practitioners and patients who might benefit most. To understand why PCP participation has been low, we partnered with the Commonwealth Fund to conduct focus groups and interviews with 29 individuals, including frontline PCPs with no prior PC-VBP model experience and experts in primary care management, to explore the challenges to participating in PC-VBP models and potential solutions.
During our interviews, PCPs and other health experts noted that while the goals of PC-VBP models are laudable, they face two main participation challenges: underfunding and over-measurement. Participants noted that these reflect the same forces driving clinicians from primary care practice. Therefore, policymakers interested in improving model participation might develop models that address the challenges highlighted by these frontline PCPs.
Increasing funding for primary care
Despite public payers’ goals to strengthen primary care through PC-VBP models, primary care payment remains too low to support changes to care delivery or increase the financial viability of overburdened primary care practices. Due to various factors, these models are often underfunded. Study participants described those factors and proposed solutions.
Misaligned fee-for-service payments remain a problem even in most PC-VBP models. As a result, PCPs are pressured to maximize the number of patient visits per day rather than improving care so that it is more comprehensive, coordinated, and patient-centered. Decreasing reliance on fee-for-service payments and reducing misalignment in payment rates for primary care compared to other specialties would better reward primary care practices that focus on the models’ goals of providing high-quality care.
The lack of meaningful participation by commercial payers in PC-VBP models deprives PCPs of the funding necessary to make needed changes to care delivery. Experts noted that the commercial payers that do engage in PC-VBP models make relatively small enhanced primary care payments. In addition, as Medicare Advantage plans continue to gain market share, safeguards are needed to ensure that health plans and health care systems don’t game the system through patient selection and enhanced coding while those doing the hard work of transforming primary care remain underfunded.
Many current PC-VBP model participants are part of large health systems that include hospitals and specialty care. PCPs and experts noted that some health systems regard PC-VBP models as avenues to increase referrals to their more lucrative specialty and hospital facility services, rather than methods of improving primary care delivery. As a result, health systems may need to be incentivized to meaningfully invest in primary care delivery, which can achieve disease prevention and management more efficiently.
Study participants also noted that some systems participating in PC-VBP models fail to provide the resources that primary care practice sites need to meet the models’ requirements. They suggested ensuring that PCPs have a voice in allocating system resources to support staff at the practice site. Two experts suggested that models should require health systems to commit to ensuring model resources go directly toward supporting primary care practices.
Small and rural independent practices face greater financial challenges to participating in PC-VBP models. These practices generally have fewer resources and narrower profit margins than larger and system-affiliated practices, so minor changes in revenue can threaten their financial viability. Providing upfront payments at the start of model participation would help practices with limited financial reserves invest in resources to improve care delivery and make it easier for them to remain in models. In addition, participants said that PC-VBP models should avoid exposing PCPs to potential financial losses through downside risk.
Decreasing and refocusing primary care measurement
Study participants worried that current quality measures, along with coding for risk scores and the demands of providing documentation for insurers, impede their ability to deliver high-quality care. Current measures that focus on binary or stark cutoffs for treatment of individual conditions or orders of screening tests are overly simplistic and do not prioritize patient care. Worse, PCPs said that health plans and health care systems encourage maximizing the number of diagnoses listed in the electronic health record to increase patients’ Hierarchical Condition Category risk scores. This burdensome documentation enhances payments to health plans and health systems but distracts from patient care.
Study participants also raised concerns about total cost of care outcomes measures in PC-VBP models, saying they are mostly affected by other providers such as specialists and hospitalists. To reduce pressure on PCPs and increase focus on the sources of high costs, some experts suggested that total cost of care measures should apply only to large organizations such as health systems or accountable care organizations, rather than to their PCPs. As one participant said, “In terms of advice for policymakers, it’s got to be ‘Save your primary care workforce.’ It’s not, ‘Let’s see if we can squeeze a little bit of savings out of a primary care workforce that can barely make it through the day.’”
The twin challenges of underfunding and misfocused measurement divert PCPs’ attention from supporting the fundamental elements of primary care. In a statement that echoed the sentiments of other participants, a PCP said that “high-quality primary care relies on having all four Cs”—accessible (first-contact), continuous, coordinated, and comprehensive care. For example, urgent care and retail clinics may provide rapid access for minor issues, but they do not provide continuous, coordinated, or comprehensive care. Participants emphasized the importance of every person having interpersonal continuity with a PCP and the benefit of this relationship for patients’, PCPs’, and teams’ satisfaction and better health outcomes. PCPs wanted more time during visits to comprehensively address patients’ range of needs. They also wanted PC-VBP models to enhance care coordination with specialists through supports for e-consults and interoperable data exchange with other providers. In addition, they emphasized the value of behavioral health integration, noting the high prevalence of comorbidity of physical and mental health needs.
Respondents also said that if primary care practices must be measured, what matters to primary care practices must take precedence. Instead of reductionist condition-specific metrics, PCPs preferred metrics that capture patients’ access to and continuity of primary care, given the wealth of evidence connecting these to better patient outcomes. PCPs thought that existing measures of primary care access, such as appointment availability for same- or next-day visits, wait times, and contacts with patients between visits, not only reflected good care but were also within the PCP’s control. They cited measures of continuous care, such as whether patients can see the same PCP at most office visits, and measures of whether patients thought their PCP clearly explained how to prevent or manage diseases and treatment options.
Conclusion
PCPs were clear: primary care value-based payment models have potential but need to be reformed to improve primary care delivery. Enhancing financial support for primary care and reducing measurement and reporting burden are critical to engaging PCPs in these models. Furthermore, health plans and policymakers should remedy currently misaligned fee-for-service payments and further enhance primary care prospective payments. Participants noted that new value-based payment models, including Making Care Primary and ACO (Accountable Care Organization) Primary Care Flex, have begun to address some of these challenges.
Going forward, participants would like models to require health systems to commit to ensuring model resources go directly toward supporting their primary care practices. Participants also suggested that the numerous and overly simplistic condition-specific quality measures be replaced with fewer and more meaningful measures of primary care access and interpersonal continuity. They emphasized that incorporating these changes can better support a more diverse array of primary care practices to participate and can make models more responsive to how primary care practices operate. These improvements will result in more people gaining access to a PCP empowered to address most of their health care needs.
Acknowledgement
The authors thank all the participants in the focus groups and key informant interviews for their time and valuable insight. We thank the state- and national-level primary care associations that posted our recruitment notices for this study and the individuals who helped us recruit participants. We thank Corinne Lewis of the Commonwealth Fund for her feedback on an earlier version of this piece, as well as David Roberts, Dianne Rittenhouse, and Christal Stone Valenzano at Mathematica.
