Advancing Primary Care Through Alternative Payment Models: Lessons from the United States & Canada

Advancing Primary Care through APMs in Canada

Canadian provinces have primary responsibility for health care while the federal government plays a smaller role in funding and setting some national standards. In this regard, they have much more autonomy than do states in their neighbor to the south. This independence means that there are significant differences in how they operate, but improvement of primary care is a shared concern and the decentralized model has also allowed important innovations to develop. Alternatives to FFS payments rose from under 10% to nearly 30% of total clinical payments to physicians between the years 2000 and 2015, leading to innovations in team-based delivery.¹¹ Family Health Teams, for example, are largely funded through capitated advanced per patient payment for all services provided, and cover one quarter of the provincial population in Ontario.^3,4 Early returns on this innovative payment transition found shifts in proportional provincial spending on primary care, improved physician and team satisfaction, and improved processes of care, all without evidence of adverse impact on individual patient selection.^5,6

In Ontario, payment reform has also been a key driver of empanelment, as capitation requires a defined group of patients for whom the practice is responsible. Having a defined population facilitates proactive approaches to preventive health care and chronic disease management including notifying patients of needed care, audit and feedback, and performance reporting.^11,12 Capitation allows providers to make greater use of virtual contacts through phone, e-mail, and videoconferencing. It also reduces the competition and disincentive to fully use other health care team members. Finally, capitation reduces the opportunity cost for participating in nondirect clinical care activities such as quality improvement activities, case conferences for complex patients and teaching. More than three quarters of Ontario primary care physicians practice in a model that features patient empanelment and blended capitation, albeit with only small capitation payments in some models.⁷

At the system level, Ontario's transformation remains incomplete as FFS continues to be a dominant payment model, interprofessional teams are unevenly distributed, and approximately one sixth of the population is left out of primary care reforms entirely.¹¹ As yet, Ontario's APMs have not been associated with improved timeliness of care, better equity, or cost savings. Other provinces have also implemented payment and organizational reforms. British Columbia has provided substantial chronic disease management payments for multi-morbid patients, while Alberta and Quebec have undertaken structural and team-based changes in the form of Primary Care Networks and Family Medicine Groups, respectively. These reforms have largely taken place without fundamental changes to physician FFS reimbursement and they have been associated with mixed findings with regard to better quality care and little evidence of health system cost savings.¹³ While Canadian APM and team reforms have been key to attracting providers, growing the size and diversity of the primary care workforce, and averting a health human resource crisis in primary care, their measurable health system impacts have been limited.

Advancing Primary Care through APMs in the United States

Recognition of the limitations of pure FFS in the United States is hardly new. Reforms in the 1980 seconds created the Diagnosis Related Grouping system to simplify and streamline hospital payments and the 1990s and early 2000s saw attempts to refine physician payments via the Resource Based Relative Value System and Sustainable Growth Rate.³ These programs failed to generate “sustainable growth” while radically growing the specialty to primary care income gap.¹⁴ More recently, the passage of the Affordable Care Act in 2010 ushered in unprecedented experimentation with APMs among private, federal, and state payors.^15,16 APM Implementation pilot programs impacting primary care have been developed within existing programs, both traditional FFS and managed care, in an attempt to move from volume-based to value-based care delivery.^17,18

There is evidence that high-functioning APMs already in existence, or as part of recent demonstrations, put more resources into primary care.^8,9 However, movement toward APMs at the health system level does not always change the FFS payment model for primary care practices. Some large hospital systems that own practices may receive capitated payments for the care of a population while continuing to reward their providers and practices principally for services rendered (FFS) more than for value received. As such, the mixed early returns in evaluations of the US Accountable Care Organization (ACO) experiment, which offering global payments and potential “shared savings” to taking care of groups of Medicare beneficiaries.¹⁹ Symposium experts suggested that ACO savings were significantly greater among ACOs that were independent, primary care-driven, and risk-bearing, perhaps due to their greater leverage over cost control and diminished FFS incentives. Panelists and reactors noted that powerful variation in this and other US payment experiments over recent years belies a continued lack of understanding of the value of primary care to population health in value-based payment models. They also noted the failure of APMs as a cost-bending panacea absent additional levers. For example, it was noted that primary care physicians anchoring the ACO and other shared savings and capitated payment innovations frequent lack levers to influence their patients' health costs beyond their clinic walls absent additional incentives shaping hospital and specialist behaviors. Attendees also noted that new APM investments in primary care are not always sufficient to enable robust primary care, particularly when they do not offer upfront financial support for practice transformation but burden practices with inflated expectations and measurement requirements.

Experts and discussants alike celebrated recent US efforts in the United States to “learn at scale” about advancing primary care through APM experimentation. At the state level, Rhode Island was noted to have phased in a near doubling of payor spending on primary care between 2007 and 2012.²⁰ The State Commissioner of Health estimates that a 43% increase in primary care spending ($18 million more annually) was associated with a 14% reduction in total spending ($115 million reduction annually)—a more than 6-fold return on investment.²¹ Illinois did not mandate a primary care increase, but between 2006 and 2010, Medicaid increased primary care funding by more than 30%. Over that period, the state saw total spending for the Medicaid population fall by nearly the same percentage, also a multi-fold reduction in total spending.²² Oregon, compelled by Rhode Island's results, recently legislated a similar increase in primary care spending by all payors.²³ There is an ongoing need to evaluate US APMs with an understanding about how they support primary care.

In addition lauded were large-scaled payment & delivery experiments enabled under the Affordable Care Act creation of the Center for Medicare and Medicaid Innovation (CMMI). Their efforts helped the of the outgoing administration achieve a goal by late 2016 of having 30% of traditional Medicare payments now flow through APMs, versus essentially none in 2010.¹⁵ Early returns for these grand primary care experiments have been mixed, with some states in the Multi-Payer Advance Primary Care Practice Demonstration showing positive savings and others none.²⁴ Similarly, the multistate, multipayor Comprehensive Primary Care Initiative revealed progress in transforming primary care delivery and generating positive provider/payor feedback, while failing to demonstrate significant savings in expenditures or quality by the end of second and third year reporting.²⁵ This outcome was felt to be promising enough to support a Comprehensive Primary Care Plus (CPC+), which offers primary care practices in 18 states or regions a risk-adjusted, monthly Care Management Fee and a Performance-Based Incentive Payment, in addition to traditional FFS. CPC+ also supports practice transformation with the goal of enhancing key comprehensive primary care functions: 1) access and continuity, 2) care management, 3) comprehensiveness and coordination, 4) patient and caregiver engagement, and 5) planned care and population health.²⁶

Lessons from the Binational Symposium for Policy Makers, Payors, Researchers, and Practitioners

A number of essential themes and conclusions emerged from the Symposium, which we have parsed according to stakeholder audience in Table 1. Discussants offered clear enthusiasm for APM experimentation in the United States and Canada, and its promise for advancing primary care, but cautioned policy makers that shifting toward APMs alone would fail to capture the positive primary care effect on creating healthy populations, satisfied patients and lower costs (the US “Triple Aim”). Many suggested that increasing overall spending on primary care was far more important than any of the payment experiments underway.

Payors were reminded that both the United States and Canada have learned from but also suffered an epidemic of small payment demonstration projects that were designed neither for sustainability nor scaling up at a health system level. This condition was pejoratively labeled “pilotitis” by 1 panelist, who noted this to be a particularly concerning problem in Canada. Admiration for the seminal efforts of the US CMMI in driving payment innovations was widely shared among expert discussants from both nations, who called for continuing to “learn at scale” and to grow investment in APMs. One bold proposal that resonated widely among attendees called for Canada to create its own CMMI as a remedy to its epidemic “pilotitis.” Attendees noted that a Canadian Innovation Center could draw on lessons learned from its successful US counterpart, while having data and structural advantages of the single-payor provincial health system to allow successful interventions to scale up even faster than in the United States.

Researchers were challenged by attendees to craft more robust methods of evaluation for the growing array of large APM innovations. There were repeated calls to improve methods for crafting better comparison groups that could overcome the challenge of detecting outcome changes, given the “pollution” and “spillover” effects from intervention to control practices inherent in large-scale payment experiments such as the US Comprehensive Primary Care Initiative. Attendees also noted the critical need to better understand how APMs can best serve health equity across US and Canadian populations. Calls were repeated throughout the symposium for evaluations of best practices in adjustment of payments to accommodate social determinants of health without excusing poor quality for vulnerable populations.

Discussants repeatedly highlighted the potential importance APMs to reducing provider burnout and returning “joy in practice” to primary care, a finding of considerable importance to providers and delivery systems. Attendees on both sides of the border connected best practices in APM global payments and increased flexibility in care delivery to reductions in administrative burden, enhanced team function, and less provider burnout. They also noted the absence of these connections in peer-reviewed literature, and called for richer evaluations of APM impact on these outcomes. Others noted the need for population health or risk-based payment models to accommodate primary care practices' perpetually thin financial margins and the restrictions these place on meaningful and radical transformation. One presenter noted that, absent at least some degree of upfront payment, ongoing transformation becomes impossible, and evolving practices remain in jeopardy of closing as the result of one bad risk-bearing contract or contract year.