Rules for alternative payment healthcare models are increasingly moving the risk from government (one-sided) to shared risk (two-sided) with accountable-care organizations (ACOs). Today, for example, ACOs participating in the Centers for Medicare & Medicaid Services (CMS) Medicare Shared Savings Program have options both for one-sided (Track 1) and two-sided risk (Tracks 2 and 3).
In fact, one-sided risk options are going away sooner rather than later. Beyond Tracks 2 and 3 and Next Gen, to earn the best upside rewards under the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA) program, physicians must participate in a qualified Alternative Payment Model – one that includes two-sided risk. MACRA rules state that participants will receive a positive, negative, or neutral payment adjustment, growing from +/- 4 percent in 2019 to +/- 9 percent in 2022.
Based upon the growing belief that potential gains and losses under two-sided risk are required to transform healthcare, it’s highly likely that most or all future programs will require two-sided risk. That’s because the evidence shows that, to change behavior and transform healthcare, ACOs must face both potential upside and downside financial outcomes. As a result, we can also expect commercial payers to follow suit with their own two-sided risk models.
How should ACOs proceed?
The first step should be to understand baseline year information to assess the levels of risk they are prepared to assume based on two factors. The first is patient population information, which includes the quality and costs associated with – as well as the stability of – their population. It is far riskier, for example, if you expect your population to turn over frequently, with the potential addition of a patient cohort with more significant health issues.
The second is looking at the stability and characteristics of the provider community. Once again, you’ll want to ensure that you have a good, stable base from which you can make future assessments, since the level of risk you assume would be at least partially based upon these components.
Once you’ve looked at the data and determined the level of risk you are comfortable assuming, then you’ll need tools that help you understand the risk stratification of your population. There are two components to this: The cost drivers, so you can create the right programs that support improved efficiency, and quality compliance, to determine possibilities for improved patient management in terms of specific registries or cohorts of the patients that are high risk or rising risk. For example, consider a registry for diabetics. You’ll want to look at quality issues such as gaps in care or unnecessary variations in care so you can create programs that positively impact the quality and cost of care for them. These programs may include criteria for specialty referral, group education programs, or intense case management for the rising risk group. This will reduce your risk for this diabetic population, while improving quality outcomes. Your risk stratification tool should allow you to track results for this selected population in order to assess return on intervention.
Once you have the population risk stratified and have identified high-risk patients, you’ll also want to look at the provider side. What patterns can be shared as best practices, and what areas demonstrate the need for improvement? With this information you can improve problem areas by creating new program initiatives, provide patient-level insights for incorporation into workflows at the point of care, or identify cohorts for broader “one-to-many” healthcare campaigns.
All of these components should be part of an iterative process, as you’ll need to monitor them to see if your actions are actually moving the needle. This means watching the consistency and quality of care and noting any new patterns to obtain the impact you expect.
What else do you need?
At the highest level, ACOs will need data control, analytics, care coordination, and management systems to achieve “the first mile” of this process, with patient outreach required for the last mile. This last piece is critical, as engaging patients can extend positive results beyond what providers can do with treatment alone. It can also support scalable population health, as well as improve the patient experience of care.
More specifically, ACOs should use risk management, care coordination, and management systems to determine which individuals are likely to incur the most costs, then intervene and manage those costs. Quality improvement tools that support real-time actionable quality metrics data are essential for benchmarking and tracking quality of care. On the provider side, they should use utilization and financial analytics systems to understand which physicians and other providers in their ACO are higher cost or drive higher patient utilization of services, so they can be educated and incented to appropriately reduce costs and the utilization of services. They will need to track costs, length of stay, 30-day readmission rates, complication rates, etc. In addition, they will need to determine network leakage and steerage issues – identifying the providers that refer patients outside the network (where there is no control) for services – in order to meet their financial goals. Tools that assess both quality and efficiency of care should be used for provider performance scorecards that track performance over time as well as to create actionable data at the point of care.
Two-sided risk is an iterative population health journey that starts with a strategic baseline review and moves to the creation and monitoring of specific operational initiatives or programs based upon the analysis. Ongoing monitoring, strategic innovation, and leadership are required to address changes in quality or cost drivers. Health information technology tools should be carefully selected to meet changes in the framework of the risk bearing in alternative paying models. There are no guarantees in this new alternative payment universe, but those adept at assessing their risks and improving or creating scalable, efficient, evidence-based and data-driven approaches are most likely to succeed.
