Incentives for Quality Care...Is This the Future?
In the few years since it appeared in health care, pay-for-performance (P4P) has become the industry’s latest rage; a hoped-for solution to lapsing quality that might help curb skyrocketing costs. P4P programs offer bonuses to hospitals that score highly on measures for clinical quality. And as such, they create financial incentives for hospitals to provide better care.
More than 50 percent of private HMOs now use P4P, and those numbers are growing. The Centers for Medicare and Medicaid Services (CMS) have also gotten into the game. Since 2005, CMS has run a P4P demonstration project involving more than 250 hospitals, which collectively improved quality from 2.6 to 4.1%, according to a recent analysis in the New England Journal of Medicine. Charlotte Yeh, MD, the regional CMS Administrator for New England, says a federal P4P rollout covering all US hospitals will likely begin in 2009.
A New Road
P4P doesn’t come without controversy, however. It relies heavily on standardization, and can reduce the art of medicine to uniform practices that many doctors resent. Those who treat P4P-covered conditions have to follow payer-mandated guidelines, such as giving pneumonia patients antibiotics within four hours of admission. Decisions not to must be carefully documented, in order to avoid compromising the hospital's quality scores. What’s more, many hospitals will have to invest in additional infrastructure to meet the new reporting requirements. P4P makes high demands on hospital IT, which tends to be antiquated and customized for financial record-keeping. Growing needs to document clinical data electronically have already strained those systems. And quality reporting under P4P will require even more IT sophistication, and possibly additional staff, says Leslie Sebba, chief medical officer at Anna Jaques Hospital, an IHM-member hospital in Newburyport, MA.
Still, P4P could offer a way to reform the current fee-for-service system, which does little to stimulate better care. “The current system merely pays providers for doing more, not for doing the right thing,”says Karin Davis, President of the Commonwealth Fund, a private foundation that funds health care research. “Hospitals that try to offer better care almost always lose money in the process.” Meanwhile, research shows quality improvements in health care have been rising at less than half the rate of medical inflation, which is about 7% a year. That discrepancy fuels the expansion of P4P, which will ideally bring quality and cost into better alignment.
P4P's Emergence
The P4P model evolved from the financial sector during the mid-1990s; used by companies that tied executive pay to outcomes such as stock prices. Then, in March 2001, the Institute of Medicine released Crossing the Quality Chasm, a report that claimed the US was failing to translate medical advances into better care. The IOM's report galvanized enthusiasm for health care reform, and accelerated the application of incentive forces in hospital settings. Pay-for-reporting emerged first; those programs require hospitals to follow specific guidelines when treating certain conditions, such as pneumonia and heart failure. A hospital's record of compliance can be publicized, which provides information for choosy consumers. Now, hospitals that don't participate in pay-for-reporting have their annual Medicare payments docked by 2%.
P4P emerged from the pay-for-reporting framework. In fact, most of the quality measures applied to P4P have a basis in CMS pay-for-reporting schemes. They include, for instance, the percentage of acute myocardial infarction patients given aspirin on arrival; the percentage of heart failure patients assessed for left ventricular function; or the percentage of pneumonia patients assessed for oxygenation. Private HMOs have their own quality measures, which aren't generally made public. However, Meredith Rosenthal, an associate professor at the Harvard School of Public Health, says quality measures used by HMOs generally mirror those used by CMS, which have been vetted in national research.
For its current demonstration project, CMS requires hospitals to submit data on 33 quality measures for five conditions: community-acquired pneumonia, myocardial infarction, heart failure, coronary-artery bypass grafting, and hip and knee replacement. Hospitals that score highly on these measures receive a 2% bonus over their normal Medicare allotments, and while those in the second decile receive a 1% bonus. Last year, those bonuses ranged from a high of $744,000 to a low of $2,829. Participating hospitals in the lowest two deciles will be liable for 1 to 2% penalties during the project's third year, which is underway now. While the project's $8.7 million annual tab comes from ear-marked funds, the actual program—when implemented—should be budget-neutral, says Anthony Shih, a program officer with the Commonwealth Fund. That's because bonuses paid to providers who meet quality standards will be financed with cuts to those who don't.
Should P4P succeed nationally, the public will get more value for health care dollars spent. But many questions remain. As a nascent concept in health care, P4P is a massive experiment. Experts don't know if the bonus structure under Medicare will be high enough to lure hospitals to the program. And Davis concedes payment systems have yet to be optimized for P4P. "We need to move away from the silos of paying hospitals one way and doctors another," she explains. "Instead of fee-for-service, we want to start paying for care episodes. The problem is patients tend to see so many different specialists. If you want to write one check for a hip replacement, who do you hold accountable? Who do you reward for doing a good job, or penalize for wasteful care?"
Meanwhile, Rosenthal has found that most rewards go to organizations that were already offering top quality. And that creates a dilemma for lower-performing hospitals, which might not get bonuses despite their efforts to improve. Remarkably, though, studies so far show low-performers subjected to P4P improve at greater rates than top performers. Rosenthal discovered this during her 2005 investigation of roughly 300 physician groups in the Northwest. A review of the CMS demonstration project by Lindenauer, et al., in the February 1, 2007 issue of NEJM, showed the same thing—hospitals in the lowest quintile of baseline performance improved 16.1% compared to 1.9% for top performers. Davis attributes that finding to the hospitals' "pride, financial incentives, public reporting, and mostly the fact that they care about their patients."
But Rosenthal cautions the findings might just reflect that those hospitals are under a research spotlight. P4P still needs to find better incentives for low-performers to improve, she says. "If we really want hospitals to feel like they have a case to reorganize and invest...we have think a little more strategically."
Meeting the Challenge
In the midst of the academic and policy debates, P4P is a growing reality for hospitals around the country. "Our expectation is that this will become a bigger and bigger piece of what we have to do to get paid," says Sebba of Anna Jaques Hospital, which has an upcoming P4P arrangement with a private payer worth millions. Sebba believes P4P can improve quality, but he emphasizes there's still a lot to learn about how to make it effective. Meeting four-hour windows for antibiotics in pneumonia cases, for instance, requires doctors, nurses, and pharmacists to interact in new ways. "Everyone has to be aware of how important it is," Sebba says. "And that's where it gets hard—there are an enormous number of people that have to be coordinated to make this happen."
As an added challenge, quality monitoring requires streamlined data systems, which aren't always amenable to the action on the floor. Most physicians—particularly older, attending physicians—have grown accustomed to "free-form" note taking, whether handwritten or recorded. Electronic medical records are evolving to extract data from free-form narratives, but that's still a touch-and-go process.
Sebba says the hospital's arrangement with the Institute for Health Metrics makes adapting to P4P easier. For instance, IHM tools have text-recognition capabilities that allow some data to be collected from narrative records. And with its Automated Indicator Extraction System (AIES), IHM imports the hospital's clinical data and transforms it in ways that allow Sebba to compare Anna Jaques' performance on hundreds of quality metrics to national and regional benchmarks. "The advantage is that the data is so current," he says. "We get it from IHM once a month, as opposed to other systems where the lag times between events and data reporting can be six months, or even years." In an actual application, Sebba says he might refer to AIES for the hospital's incidence of surgical infection, which he can tie into feedback for improving quality.
Mount Sinai Hospital in Chicago, IL, which is also an IHM-member hospital, participates in the CMS demonstration project. Geoff Page, with Mount Sanai's quality department, says the AIES system offers surrogates for measures that aren't captured electronically. Antibiotic timing, for instance, doesn't have its own field in Mount Sanai's electronic medical system, so Page uses an AIES data element for blood culture timing as a proxy. "The logic is that blood cultures are taken before the antibiotics are given," he explains. “So, by looking at blood cultures we can see if our patients are way over the four-hour window.”
Page compares treatment before the emergence of core measures, pay-for-reporting, and P4P to the Wild West. Before these new approaches came along, he says, doctors used a range of approaches for treating the same conditions. There might have been guidelines for best practices, but there was no reason to follow them. Now, Mount Sinai doctors have to consult "order sets" for specific disease states, and check off their compliance with specific measures. Page acknowledges many of them don't like the intrusion. “You hear the term “cookbook medicine,””all the time, he says. But he adds the order sets also suggest procedures that might not have been considered otherwise; a key aim for quality improvement.
From within the P4P trenches, Page's view is that the new approach shows promise—not just for improved quality, but perhaps for limiting costs. "Maybe that's a naïve belief," he says. "Do I have hard evidence to support this? No. But consider this: There's a lot of hidden, un-reimbursable cost in hospital infections that aren't properly dealt with. So, if P4P helps drive out unnecessary infections, then that could lead to lower cost. And I also know that financial incentives can change behavior—if there's money involved, hospitals are going to do what they can to improve their quality performance."