The author of a study on pay for performance in healthcare, says there is little evidence to support that incentives for following certain processes that have been proven to add value… actually improve quality or value.
Payers have for years tried to impart a sense of accountability to healthcare providers through financial incentives, yet inefficiency and poor outcomes still plague the healthcare system.
The way the majority of healthcare is paid for is a chief culprit. Fee-for-service healthcare provides virtually no incentive for coordination of care and perversely rewards waste.
This is not a new problem, but it defies easy solutions.
Payers have tried to imparting some performance expectations and measurement tools into healthcare payment through so-called pay for performance programs that offer incentives for following certain processes that have been proven to add value, and they should make a difference, as logic dictates.
The problem, says the author of a study on pay for performance, is that this kind of logic may not apply in healthcare, at least not at the levels currently in place.
One of the more widely used tools in a thin toolbox that payers use to modify behavior among service providers has been so-called P4P incentives. Insurers have relied on them for years to better coordinate care, eliminate waste and improve processes. But do they work?
We don't really know, says the study's author. While there's widespread evidence that providers do respond to such incentives, says Andrew Ryan, an associate professor of public health with the Weill Cornell Medical School, there's little evidence that such incentives actually improve quality or value, which are, after all, the goals.
P4P Shortcomings
And although P4P seems to make common sense, and P4P based programs are now widely used by both CMS and commercial payers, Ryan says one shortcoming is that they're probably too limited in scope and in scale.
"One of the problems with the programs that have been implemented to date is that incentives aren't high-powered," says Ryan. "P4P represents less than 1% of revenue for providers in some cases."
An analogy would go something like this: Two competitors are hired to build a fence. A high-quality fence would bring Competitor A $100, and a low quality one that built by Competitor B would bring him $99. The cheaper fence could be the worst fence in the world, and fall down tomorrow, but as long as it was standing today, Competitor B would be paid the agreed-upon $99. Competitor A might build the best fence ever and still been paid $100.
What incentive does Competitor B have to improve? Not much. What incentive does he have to continue to labor intensively for his high quality fence? Again, not much.
If it now seems ridiculous that providers would make the sorts of investments that would help them improve quality and attain such meager bonuses, it is, Ryan says.
P4P Incentives Too Small
Another problem is if the incentives need to be larger, where is the money going to come from? Very few commercial insurers want to add new money so that means moving away from bonuses. After all, the larger goal is to reduce the cost growth of healthcare. New bonus money has to come from somewhere. Meanwhile, Medicare is trying to do incentives that are budget-neutral, Ryan says.
"The problem here is if we went to 5% or 10% [of revenues], that would get attention, but risk disadvantaging certain types of providers like the safety net. Some could get penalized a lot," he says. "It's a classic trade-off and I'm not sure what the sweet spot is as to the level of incentive anyway."
Another problem with getting traction from P4P on improving value is that such programs focus on process measures instead of outcomes, Ryan says. But measuring outcomes presents its own set of complications, which is one reason why P4P still gets so much attention.
"Providers often don't like outcomes because they argue they are largely out of their control," he says, adding that performance on outcomes measures is quite noisy statistically, "so you might have a rate that looks good one year and bad the next and is a result of statistical noise. There's not much of a quality signal there."
The other problem, of course, is that incentivizing outcomes also creates incentives to avoid providing care for certain types of patients.
"We have risk adjustment, and if it was perfect, we wouldn't have people contending reimbursement on outcomes is unfair, but it's not and never will be," says Ryan.
There is a move toward using outcomes more in determining reimbursement levels, which is what's happening in the Medicare programs.
"On the hospital side in the first year, incentives were just for process measures and patient experience. In the second year, outcomes will be incentivized, so over time, process measures will get less weight and outcomes will get more," says Ryan. "It will be interesting to see once we start to emphasize action on improving outcomes whether we see those unintended consequences."
Such incentives have a mixed record at best of obtaining the type of care coordination and handoff work desired by both payers and patients. Ryan says health policy for decades has tried to influence payers and providers to replicate Kaiser, a fully integrated system with both a payer and provider component under one roof. That way, he says, incentives are aligned for low costs, efficient care and high quality. Since the Clinton health plan and managed care both collapsed due to a variety of factors unrelated to healthcare transactions, he says P4P has filled the vacuum as something that could be done to influence quality on the margin.
"But the results have certainly not been transformative," he says. "We're at a place where there's almost complete recognition we need to move away from fee-for-service somehow, so we have ACOs, P4P, and CMS experimenting with bundled payment models to pay prospectively rather than retrospectively."
None of those alone has the blueprint to better, cheaper care, he says, adding that messing around at the margins is not likely to bring success.
"Providers want to be ahead of the curve, and participate in these new payment models to get some learning experience so they'll be ready, but it's not clear that these models are in their best interest to participate in," he says.
Wanted: A More Cooperative Relationship with Payers
"Anything that's entirely voluntary won't get much momentum, and ultimately, if it's not in providers' financial interest, they'll stop participating." P4P programs are a poor motivator not only because they are limited in revenue scope, but also because they have been seen by insurers as the best they could do to influence providers to improve, he says.
"That's the reason why they're doing this," he says. "Something must be done. This is something, therefore this must be done," he jokes. "Other valid reforms are more challenging."
Ryan advocates a more cooperative relationship with payers, which includes intensive technical assistance, coaching, and learning collaboratives.
"The easiest thing for payers to do is to change how they pay, so we have this generation of P4P programs," he says. "It's hard to argue with the concept but there are so many ways it can fail, so there's some opportunity cost for policymakers and payers."
Ryan says the cost of doing less than optimal programs to improve quality is that they take away momentum and enthusiasm for programs that might be more beneficial.
"I wouldn't say we should abandon P4P, but we should think about these tradeoffs."
Philip Betbeze is the senior leadership editor at HealthLeaders.