A fragmented health care system poses many challenges to implementing fair P4P initiatives across the board.

The concept of rewarding physicians for providing quality health care efficiently may seem straightforward, but putting that rewards system into action is no easy task. The steps for gathering practice data, determining performance criteria, and setting fair standards and rewards policies all pose unique challenges to a health care system that is fundamentally fragmented. So, it is no wonder that very few guidelines are in place to help radiologists take advantage of pay for performance (P4P) programs.

Still, P4P initiatives are creeping into health plans across the nation. Last year, Health Industry Insights, Framingham, Mass, surveyed 60 US health plans and found that 72% already had P4P programs in place. Of course, these programs are still in their formative stages—only 25% of those surveyed had P4P initiatives for 3 years or more, and 35% had them for less than 1 year. “There aren’t strong universal standards in place yet, and there’s really no timeline or calendar for these programs at all,” said Janice W. Young, program director of payor IT strategies for Health Industry Insights. “The reality is, as we look at our data, the programs change over time because they really are experimental.”

But as more health plans—especially large payors such as the Centers for Medicare and Medicaid Services (CMS)—enter the P4P arena, chances improve for more consistent performance benchmarks to solidify. In the meantime, radiologists and other specialty groups have the opportunity to shape future quality measures during these early stages. “If you have physicians at the table and they contribute to the design and implementation of a P4P program, they’re much more likely to participate willingly,” said Mark Gorden, MS, senior director of P4P initiatives for the American College of Radiology (ACR), Reston, Va.

Building Consensus

So far, there is no perfect P4P setup to serve as a model for new programs. “The ideal model is one in which there are clear expectations of what performance is, which means there is agreement about what is quality, what isn’t quality; what is a good outcome, and what isn’t a good outcome,” said Young. “[Now] we either don’t have enough information about what creates a high-quality clinical interaction, or don’t have agreement about that yet.”

To move toward that agreement, the health care community has established processes to determine quality standards. Michael J. Pentecost, MD, director of the Mid-Atlantic Medical Group radiology practice for Kaiser Permanente’s facilities in Maryland, Virginia, and Washington, DC, describes the process to develop these metrics as a funnel. Multiple sources submit a wide range of ideas, which are then narrowed down by at least three committees.

Physician groups, including the ACR, make suggestions that are discussed during quarterly meetings of the Physician Consortium for Performance Improvement (PCPI), which is under the umbrella of the American Medical Association (AMA). From there, the information filters through the National Quality Forum (NQF), which includes members from the National Institutes of Health and FDA. “They validate the AMA’s work to make sure it’s not too self-interested,” said Pentecost.

Finally, the information goes to the Ambulatory Quality Care Alliance, which consists of members from the NQF and the PCPI as well as representatives from health insurance plans, who determine how to apply these metrics to the marketplace. It can take a year or more for the metrics to funnel through these three channels, but even then, it doesn’t mean that the measurements will be applied consistently across all health plans.

“If you’re a health plan, you can create a P4P program and decide that you’re going to pay your physicians for a particular kind of quality and efficiency objective, but the health plan down the street may decide that their program is slightly different,” said Young. “One of the challenges that we have in this country is that many of our P4P programs are very different right now.”

This problem is fairly unique to the United States. In Britain, for example, the centralized National Health Service (NHS) eliminates these discrepancies, by establishing national primary care quality of care benchmarks for family practitioners. Britain’s P4P program with 146 performance-based criteria was launched in April 2004. NHS expected that providers would earn 75% of the available points, but they earned a whopping 97%—a success rate that strained the program’s coffers. Family physicians earned approximately $40,000 in P4P-related bonuses, nearly equivalent to 25% of their salaries.

“It is easy to see why so many family practitioners welcomed this bonus program, yet the vital contribution made by specialists such as radiologists goes without recognition in the British bonus program, something we are trying to allay here in the United States,” said Gorden.

Of course, chances are slim that US payouts would ever reach this level. However, Gorden, who attended the 5th World Congress Healthcare Quality and P4P conference in Boston on August 6-8, says that less than half of the British bonus would be enough to grab US providers’ attention. “When it starts being in the 5% to 10% range, that seems to be the sweet point for getting physicians to be seriously interested in participating,” he said.

By contrast, CMS, which launched its Physician Quality Reporting Initiative (PQRI) on July 1, offers providers up to a 1.5% bonus for reporting on its current quality measures. While this may look small, Gorden says it can actually translate to a substantial figure. “They’re not just paying the 1.5% bonus for the patients that fall under an individual physician’s reported quality measures,” noted Gorden. “They’re paying the 1.5% bonus on all Medicare allowable physician fee schedule billings during the 6 months the PQRI is in operation in 2007.”

The PQRI is primarily a data-gathering, pay-for-reporting effort, which will help CMS—and, by extension, the rest of the health care industry—establish performance benchmarks for later P4P implementations. “With large volumes of data, you can achieve a statistical validity for benchmarking not possible on an individual physician basis,” said Gorden.

This is important because most physicians work with a variety of health plans, which can make it difficult for individual carriers to determine how well a physician is doing overall. “If you have a small number of interactions with patients on a particular health plan, that’s probably not representative of your overall practice pattern, and yet, that’s all that health plan has to look at,” Young says.

This could translate to skewed data that may inappropriately punish a high-performing provider. “Say you treat only a small number of patients in one health plan. For example, if you perform an MRI on only five patients, and the physician fails the performance metric one out of five times, you receive an 80% score,” said Gorden. “Is that fair, or is that just random variation of small numbers?”

To address these concerns, some groups, such as the Integrated Health Association (IHA), Oakland, Calif, have formed consortiums to collect and analyze data from different health plans that serve the same region. The same standards apply to all plans within the consortium, and pooling the data means having more accurate metrics to judge physician quality achievements. There is also the potential to pool incentives for physicians. Young says that collected data from organizations like these will influence the structure of other health plans down the road. “Health plans will begin to adopt the consistent strategies that come from some of the bigger programs, so over time, the variability will decrease,” said Young.

In the meantime, the industry is paying close attention to CMS’s P4P efforts. Currently, the PQRI includes two diagnostic imaging quality measures related to imaging for stroke, as well as quality measures for interventional radiologists and radiation oncologists. The ACR plans to expand the number of quality measures available for radiologists for the PQRI in 2008. “We’re currently working with the AMA PCPI through its radiology workgroup on a number of new measures, including one on increasing patient safety through radiation dose minimization,” said Gorden. “The goal is to get as many valid measures out there as possible so that all radiologists can participate in the CMS bonus program.”

Janice W. Young, program director of payor IT strategies for Health Industry Insights

Addressing Concerns

The main objectives behind P4P initiatives are rewarding providers for high-quality care and reducing the number of unnecessary tests and procedures. However, P4P brings with it a variety of challenges that pose difficult questions for the health care industry.

“[Health insurance companies] face a philosophical conundrum in the whole principle of paying you more to do a good job when they are still paying you pretty well to do an average job,” said Pentecost. “Physicians see the same conundrum. Does this mean they’re going to pay me more for this good job, but maybe not pay me at all for this other job? There’s a fair amount of suspicion, of course, between physicians and the medical insurance industry, and that sort of sets the tone for some of these conversations.”

A major concern for radiologists is whether bonuses will be fairly distributed among the entire health care team or if they will go directly into the primary care physician’s pocket instead, which is the case in Britain and for some CMS-coordinated care demonstration programs, according to Gorden. Currently, most P4P programs focus on rewarding primary care providers. “Any program that limits bonuses to primary care physicians ignores the fact that perhaps one major reason a physician did so well is because they got excellent guidance from a radiologist,” said Gorden. “One of our goals is to ensure that payment bonus systems are designed to equitably share the savings with all of the providers participating in an individual patient’s care—and the radiologist should definitely get a slice of that.”

There are other variables, such as patient case mix and severity of illness, to consider as well, which may require risk-adjustment analysis. For example, in some areas, patients are noncompliant with care regimens, which doesn’t necessarily reflect on the physician’s performance. “Another example is when a referring physician receives a radiology report with critical findings, but fails to initiate appropriate treatment in a timely fashion,” said Gorden. “For all of these reasons, it’s not always easy to differentiate an outstanding performer from an average performer. You cannot understate the complexity of establishing technically sophisticated P4P programs—not setting the bar so low that everybody passes, but at the same time allowing true stars to shine.”

Pentecost also worries that P4P may not encourage care for rare conditions that do not fall under the program’s incentive system, or that it may stifle the progress of new procedures. “P4P systems don’t compensate very well for innovation,” he noted. “They do more for conventional studies.”

To address this concern, Gorden says that registries, such as the National Oncology PET registry, allow for evidence-based data collection that will encourage Medicare and health plans to cover breakthroughs in imaging technology, such as CT colonography and coronary CT angiography. “To encourage treating physicians and radiologists to submit data showing the impact of PET on patient cancer care management, CMS reimburses for the PET procedures performed and submitted to the registry for all qualified patients,” said Gorden. “The use of evidence-based registries is how you build confidence with all of your payors, and particularly the government, that you have something worthwhile.”

The Digital Divide

To meet P4P efficiency criteria, information technology systems will play a major role in all segments of health care. The ultimate goal, of course, is to implement nationwide electronic health records. But with the variety of software program implementations available, not to mention data-gathering techniques and even procedure definitions, achieving this type of standardization will take a long time. In the meantime, providers should pay attention to technology trends that could improve efficiency down the road.

Young says that health plans have given providers mixed messages about the technology infrastructure needed to meet P4P criteria. This is a difficult problem, because not all providers invest in the same technology, and some health plans assume that providers have access to information systems when they do not. “There is a bit of a technology investment gulf,” said Young.

Pentecost is concerned that technology requirements will price indigent hospitals out of P4P participation. “It requires a fair amount of sophistication to participate in these kinds of programs—information support, information technology, etc,” said Pentecost. “The hospitals that are participating in these things tend to be the ones that are already doing quite well. And I’m sure they’ll do better. But if the money comes at the expense of indigent hospitals, then where is the fairness in that?”

Michael J. Pentecost, MD, director, Mid-Atlantic Medical Group/Kaiser Permanentes

The good news for radiology is that many practices have already invested in efficiency-related technology such as RIS and PACS, which will go a long way toward meeting future P4P criteria. Pentecost cautions radiologists against investing too heavily in P4P-related technologies just yet. “Being able to pull information up on a computer rather than flipping through charts will make it financially a lot more feasible to do,” he said. “But I’m not sure that enough money is in the system yet to make people invest in it because of P4P.”

Still, Gorden says that the number of P4P initiatives is increasing every day, which means radiologists need to be vigilant. “Radiologists need to keep abreast of what is going on in the marketplace, and to be active players in assuring that the vital role they play in the diagnosis and treatment of patients is fully recognized and rewarded,” said Gorden. “ACR is making sure that radiology is well represented in national P4P development efforts, with active participation in the AMA PCPI, the Ambulatory Care Quality Alliance, and NQF, as well as frequent input and consultation with CMS.”

Future Implications

P4P programs have the potential to reward radiologists not only with direct bonuses, but also, as Pentecost points out, through increased referrals from primary care physicians who may need to order tests to meet their own P4P criteria. It is likely that more consortiums like IHA will emerge in the future as well, which will expedite the quest for better, consistent standards. However, it is unclear how P4P will evolve in the next few years or even how it will impact health care in the long term.

“There’s a little bit more caution, or even perhaps more skepticism, than there was a couple of years ago when everyone thought this might be the new real solution to a lot of health care problems,” said Pentecost. “I don’t think that P4P is the solution to a lot of health care’s broad-based ills, but it certainly could help in certain areas.”

In the meantime, ACR continues to lobby for more recognition for radiologists’ contributions to quality patient care. “We have to work with the government and the private payors to get the message out that radiologists are a vital, indispensable element of the patient care process,” said Gorden. “We need to make sure that any programs that involve setting up incentives for primary care physicians as care coordinators fully recognize how substantially radiologists impact the course and quality of care, ensuring that they are appropriately rewarded as critical members of the patient care team.”

Ann H. Carlson is a contributing writer for  Axis Imaging News. For more information, contact .

P4P Resources

Several health care organizations provide information on the Web to keep physicians current on P4P initiatives. These resources include:

For additional information, check the Web sites for your local and regional health plans.