Alternative payment models (APMs) are the primary strategy of Medicare and payers to bend the cost curve. Early data suggests that these models can reduce costs; participation in the Alternative Quality Contract or Medicare Accountable Care Organizations (ACOs) has been associated with improved health care utilization and lower health care spending.
However, both the Alternative Quality Contract and ACOs have produced only modest reductions in cost. In a health system where as much as 30% of spending is wasteful, why have they not saved more money? These models fall short on leveraging behavioral economics to maximize the impact of incentive dollars on physician change. They offer relatively small rewards with delay and uncertainty while being complicated by too many metrics. Though APMs may bring meaningful improvements to the status quo, it is possible to do better.
In the first years of the ACO program, savings were greatest among independent primary care groups. Drawing on our experience in using behavioral economics principles to motivate change among independent physician practices and our experience navigating over a dozen commercial and Medicare APMs, we worked with a payer to co-design a model primary care reimbursement plan. This plan, referred to as the Primary Care Outcomes Model, better aligns incentives for PCPs and can maximize gains in achieving the Triple Aim.
Pillars of the Primary Care Outcomes Model
The five pillars of this model involve focused engagement with primary care physicians, timely provider payment, enhanced primary care pay rates, meaningful quality measurements, and sharing accountability for total cost of care with the plan.
Design together with PCPs. One approach to engaging providers involves focusing them on narrow quality measures and tasks. We prefer the alternative of aligning primary care providers with goals and supporting them as they solve problems creatively.
In the Primary Care Outcomes Model, 10 to 20 practices would voluntarily group together in human-scale ACOs, assuming accountability for total cost of care. Transparency on performance indicators selected by the practices would create friendly competition. Over time, accountability would shift into shared downside as well as upside risk. (To motivate but not cripple, the downside risk would be capped as a proportion of practice revenue.) In later years, lower-performing practices may be asked to leave the ACO, encouraging every clinic’s engagement.
Pay clinicians in real time. The Medicare Shared Savings Program rewards ACO providers who sufficiently reduce total cost of care by sharing savings; however, providers do not learn about their results until 9 months after the end of the performance year and do not receive a check until a month later. This time lag abandons the immediacy principle: people respond most strongly to immediate feedback. A decision a clinician makes today might be punished 20 months later.
Some Medicare Advantage plans have moved toward a faster feedback system by allowing quarterly reconciliation with prospective benchmarks. The Primary Care Outcomes Model goes further to close the gap between the time a provider takes action and then sees the fruit of it. A provider’s bonus would be estimated monthly using the best available evidence by the payer. The provider could receive part of the estimated bonus immediately and the entirety at the performance year’s closing. The reward of shared savings for improving quality and reducing costs would no longer be an intellectualized concept for a 20-month span; it is a monthly, material reality that motivates because of loss aversion.
Pay more for primary care. Existing fee-for-service plans undervalue primary care, particularly when delivered by independent practices with little market leverage, which ironically can increase market consolidation and prices. In the Primary Care Outcomes Model, an outpatient office visit with participating PCPs for a complex patient would be reimbursed at a higher rate based on the complexity of the patient, as much as 50% higher for the most complex patients.
Health plans increasingly create reimbursable codes for services that go beyond office visits. For example, the Annual Wellness Visit prioritizes investing in prevention and the patient relationship. But a flat fee incentivizes the delivery of these services to those who are most accessible rather than those who would most benefit.
The Primary Care Outcomes Model would reimburse more when PCPs invest in delivering these high-value services to their highest-risk patients. Our experience prior to the Primary Care Outcomes Model suggests that receiving an Annual Wellness Visit is most strongly associated with reduced health care costs and improved quality for these higher-risk beneficiaries. Similar principles are applied for other nontraditional visits, such as Transitional Care Management and Chronic Care Management.
Restore meaning to quality measurement. APMs typically involve extensive quality measurement. Many providers view quality measurement requirements as burdensome, with targets too numerous and focused on less relevant areas. Payers increasingly find that allowing providers to pick from dozens of quality measures does not permit provider comparisons. Focusing on a handful of broad-based, more clinically meaningful measures with easier electronic reporting processes allows providers more time to care for patients.
The Primary Care Outcomes Model would partner with PCPs yearly to choose three to five meaningful measures, reducing choice overload. The selected measures would encompass a substantial portion of total spending and hospitalizations. In year one, PCPs could report on blood pressure control, diabetes control, and appropriate statin use among those at high risk. These conditions are common enough that even smaller primary care practices can have adequate sample size. Critically, none of these measures require digging through charts; they can be reported with data already being collected in routine care.
Provider performance is traditionally evaluated relative to a threshold (e.g., 70% of diabetics under control). The Primary Care Outcomes Model instead would pay for the numerator. A provider would be rewarded for every additional patient with diabetes who achieves well-controlled HbA1c. In the standard approach, providers who reach some arbitrary target threshold can stop trying to improve and clinicians well below may perceive they cannot feasibly reach the threshold, inducing a threshold effect. By contrast, paying for the numerator would encourage providers to excel on quality measures for all patients.
Foster plan-provider coordination. Building on these shifts in reimbursement, the Primary Care Outcomes Model would shift how plans and providers work together. With PCPs aligned on total cost of care, plans could engage them as partners rather than adversaries in utilization management.
The Primary Care Outcomes Model would streamline prior authorization to make higher-value choices the easy ones. When a patient needs an MRI, authorization would not be needed to see a preferred provider at a non–hospital affiliated center. For a provider who has prescribed the same medication 9 times out of 10, the Primary Care Outcomes Model would assume credibility on the 10th time when a different medication is prescribed; no prior authorization would be necessary.
The Primary Care Outcomes Model also would use value-based benefit design to align patient incentives with providers in addressing underuse of high-value services and overuse of low-value services. Finally, the PCP would not only be kept in the loop, but they would also coordinate the prescription of all clinical services, including those that insurance plans add, such as medication therapy management, home visits, or behavioral health treatment.
The Promise of Primary Care Value
The Primary Care Outcomes Model offers the possibility of substantially reducing costs compared to conventional ACOs. Performance on quality measures could improve dramatically with focused improvement efforts.
Delivering effective primary care is critical to resolving the health care cost crisis. Health care leaders need to engage PCPs in managing the cost and quality of the medical care their patients receive and permit them to stay independent. This is best done by offering PCPs an ethical and motivating way to earn more of the value they create.
Government and private payers would be smart to work with PCPs to design better aligned primary care reimbursement models. While no model is perfect, our experience collaborating with a payer to design this model demonstrates the promise of such collaborations in advancing the Triple Aim for patients, providers, and society.
Disclosures: Adam Beckman, Travis Broome, and Farzad Mostashari are employees of Aledade. Mostashari also serves as a health policy partner to NaviMed Capital. Bob Kocher is a Board Member of Aledade and Partner at the venture capital firm Venrock, which is an investor in Aledade. Co-author Laura Chmar was Vice President of Medicare Advantage for Aledade at the time of the writing.