The AHEAD Model and Global Budgets
VBC walked so AHEAD could run
No better news for health policy nerds than a new reimbursement model. At the end of last year, CMMI put out AHEAD that calls for a global budget reimbursement model. This is a change from value-based care (VBC), which has been touted for the last decade without the results we expected to see from it. This article isn’t going to get into the pros and cons of VBC (that post is coming) but rather a view into what the AHEAD model is, lessons from previous global budget models, and what changes might help this model be successful.
The States Advancing All-Payer Health Equity Approaches and Development (AHEAD) model is CMMI’s solution for state-level global budgets focusing on population health and health equity.
The mechanism is simple - hospitals receive a global budget from the government based on previous financial data for a specific patient population. Staying under the budget allows them to keep the excess funds as profit, while going over means they have to foot the bill. The incentive is to reduce unnecessary utilization of healthcare services which should lead to cost savings. Having a global budget should also slow spending growth (ideally matching inflation). If hospitals know what money is coming in through the door ahead of time, they will hopefully adjust operations and practices accordingly.
Investing in primary care and preventative services goes hand in hand with reducing costs via over-utilization. The AHEAD model will provide payments to outpatient practices, focusing on social and mental health services to prevent large future costs for patients. Unlike traditional insurance models that are not as incentivized to invest in preventative care because they may not see the benefits of reducing future costs, global budget models have more skin in the game to curb costs for the long-term. The focus on primary care is a component of the model, so much so that there is a specific arm of it called Primary Care AHEAD that is specifically for primary care practices that want to be involved.
The AHEAD model is also taking a stance on health equity measures, requiring hospitals to collect patient demographic data and screen for social needs. This data will be used to better connect patients to community resources that they need. In addition to this, there are also bonuses in place for hospitals that show improved performance on disparity-focused measures. The biggest part of health equity though is adjustment for social risk. Essentially, global budgets include adjustments for how socially risk your population. Sound familiar? (read: risk adjustment)
Global budgets aren’t a new idea. If you’ve been in the space for a while, you’ve probably seen these models implemented at one point or another. The most cited example is Maryland’s global budget system that did slow spending in some parts of the system but only had a modest impact on quality and population health metrics. A 2018 study showed that after 2 years of the global budget program in Maryland, there was no change in hospital or primary care use due to the program.
These were early numbers and these attributions are sometimes hard to make but current data suggests that global budgets can slow spending and possibly reduce it in some areas, but their impact on improving quality, outcomes, and population health metrics (a vague term that probably includes health disparity measures) remains questionable.
Even with these results, CMMI is betting on AHEAD and stating that “[w]hat separates AHEAD work from current state-based models is that CMS will implement AHEAD, with its state-based TCOC approach, concurrently across multiple states.”
It’s unclear why having multiple states will make this approach better unless they are thinking that less expensive states can offset more expensive ones which is a reasonable thought and one that could address the scale issue of just having a global budget in a single state.
AHEAD does face some obstacles that it will have to work through. The main one is that this will be a voluntary program and states can join either at the state level or sub-state level by region. After that, hospital and primary care practices will have to be recruited. The model is called “all-payer” but will have traditional Medicare, Medicaid, and a requirement of one single private payer per state. This leaves out requirements for MA plans and other commercial plans unless they opt to join. Functionally this means that provider systems can pick and choose reimbursement models. They can join AHEAD while maintaining traditional payment via commercial payers. This variability becomes an accounting game where the incentive may be to reduce utilization for patients in the AHEAD model in order to earn profits there and increase utilization or emphasis on commercial and MA patients that bring in revenue through traditional insurance models.
In this way, the AHEAD model isn’t truly a global budget but a slice of the entire revenue pie for a provider system. This is the same thing we see with VBC models where a provider system may participate in VBC for a portion of its patients. This leads to gaming the system and likely higher operational costs of figuring out which patients need specific services to meet VBC measures.
In terms of primary care, the theory is that reducing utilization will save money that can then be invested into primary care and health equity measures. The jury is still out if this will work in practice but as we saw in the study from Maryland’s early years, the global budget system didn’t change primary care use.
The health equity part of AHEAD is the most interesting but also at the greatest risk for gaming. I’m of course talking about social risk adjustment where payments to providers will change based on the social risk profile of the population they serve. We all know how this went for risk adjustment (not well) but my fear is that we’ll see a lot of coding games played by provider systems (yet another operational cost) to maximize global budget payments. It’s hard to know the risk profile of a specific population without asking the provider system but ideally this is where looking at historical revenue and numbers could help. If you suddenly start seeing that a population became “more socially risky” then something fishy could be going on that warrants investigation. The issue with all of this is that it feels like more policing instead of focusing resources on things that matter like taking care of patients.
On first pass, it may seem that the AHEAD model isn’t fully set up for success and to be fair it does need some work. However, I still appreciate what CMMI is doing. They are navigating an imperfect system and trying to find different levers that may work. It’s hard to change financial and quality outcomes in a system with multiple payment models, so targeting specific areas is a good approach. I don’t think AHEAD is a silver bullet that fixes everything but it can tie into a changing policy landscape addressing healthcare costs in our country.