TRENDS: Show Me the Money

February 1, 2006

Like it or not - and we don't blame you if you don't - Medicare and private payers are enamored with so-called pay-for-performance models. Here's the scoop on how these programs work, and how to make the best of them.


During his 28-year health care management career, Ray Fredette has survived the onslaught of managed care, weathered the ravages of reimbursement cuts, and faced the challenges of the information technology age.

Yet he was caught off guard by a recent conversation with one of the primary care physicians in CentMass Association of Physicians, the Independent Physicians Association of which he is chief executive officer. The physician told him that he was basically firing five patients who had not gotten their mammograms as recommended.

“He said, ‘I am sending 30-day notice letters to these five patients, who, despite my best efforts of two phone calls and two letters, will not get their mammograms. These people are costing me money and wasting my time. I am tired of their names showing up,’” on a list of patients who were not getting necessary tests, Fredette recalled the physician saying.

“This was a first, and what drove him to it was that those patients meant that he didn’t meet his target to get his 50 cents per member per month bonus from one of the health plans,” Fredette says.

Welcome to the world of pay for performance, also known as P4P. Fredette understood the physician’s frustration, but he was deeply troubled and had to wonder: Is this really the way P4P is supposed to work?

For all their good intentions, P4P programs can have some unexpected consequences, Fredette and others are learning. And their lessons are important for you. Although the P4P movement is still in its infancy, more payers are offering these programs every day.

Perhaps the most important payer in terms of setting direction for the nation - Medicare - is also embracing P4P. Medicare was scheduled to begin a voluntary program of collecting patient data on some 36 measures this month.

The Centers for Medicare & Medicaid Services, which runs Medicare, has asked physicians to use G-codes to indicate, for example, whether a heart failure patient with atrial fibrillation received warfarin therapy. There are no financial incentives for participating in this program, and many medical groups have refused to endorse it.

Eventually, you will have to decide whether to join P4P programs, if you haven’t already faced this question. Fredette and others on the front lines say it’s time for practices to pay closer attention to these initiatives and to become involved in how they are structured. The steps you take now, they say, will ultimately pay off.

The role of the patient

In some areas of the country, such as Massachusetts, P4P programs are more entrenched. Fredette’s 200-member association, based in Leominster, is participating in P4P programs with all of its major payers through a coalition of payers and providers called the Massachusetts Health Quality Partners, which collects and publishes data on 16 measures.

One plan pays the docs directly while others give grants or awards based on performance. The partners organization collects and distributes data, but doesn’t have any role in the actual payment.

On average, Fredette’s physicians count on bonuses equal to about one-half of one percent of their total practice revenue from these programs.

All the plans pay for reporting on the Health Plan Employer Data and Information Set (HEDIS) measures, a program of the National Committee for Quality Assurance, with one offering 50 cents per member per month for each measure. Some also collect data on unique measures of their own; one payer, for example, gives a little extra for e-prescribing using a proprietary device.

Fredette’s group has long rewarded its primary care physicians internally for their HEDIS performance in risk contracts with payers. Fredette has generally had to rely on claims data from his payers to ascertain physicians’ performance, and the data was often incomplete and untimely.


To get better information on which patients needed following, the group recently purchased the services of a data warehousing and mining company, MedVentive Inc., of Cambridge, Mass. CentMass uses MedVentive’s Quality Focus to better track patients who need important HEDIS tests. “We can control the distribution of information to our PCPs and improve patient compliance by using more timely data from this system,” he says.

Participation in P4P programs has not only paid financial dividends, but public relations ones as well. Harvard Pilgrim Healthcare, one of Fredette’s payers, touts his group on the plan’s Web site as being among its “honor roll.” This means the group is among “practices whose performance exceeded the national 90th percentile on a majority of quality scores such as cervical and breast cancer screenings, asthma management, diabetic testing and management, and the number of well-visits a child has.”

Among the biggest issues that have emerged as the reporting program has matured is the role of the patient. Fredette echoes a concern of other physicians and their representatives, who argue that P4P measures must somehow take into account patients’ refusal to undergo a test.

“Your PCP cannot make you get a mammogram if you don’t want to get one,” he says. “It is unfair to penalize physicians for patients who are noncompliant. We put the burden on the providers of care...we don’t put any burden on the patient.”

“There is no personal responsibility built into this system,” Fredette adds, pointing out that health insurance premiums are not risk-adjusted; patients who are overweight or smokers pay the same premium as people who do not engage in unhealthy habits. Yet, for physicians in P4P program, “there are huge incentives to discharge uncooperative patients.”

Medicare’s big gamble

In late October 2005, CMS announced that it was rolling out a P4P program for physicians called the Physician Voluntary Reporting Program, or PVRP. Physicians could begin reporting data on the 36 measures as soon as Jan. 1, 2005, CMS officials said, adding that the program was modeled after the agency’s hospital program, “Hospital Compare,” in which hospitals “voluntarily” report data for a variety of measures.

When Hospital Compare first began in 2004, data were collected for 10 measures; the number grew to 20 in 2005. All of the measures were developed by the Hospital Quality Alliance, an independent organization that included the American Hospital Association and CMS.

Hospital Compare data are reported quarterly and are collected retrospectively in most cases. Hospitals have a financial incentive to participate (and nearly all the nation’s 4,000 hospitals are doing so). Hospitals that did not report in 2005 saw their annual payment increase reduced by about half of one percent. On April 1, 2005, CMS began posting some of the data, by hospital, on a public Web site.

CMS’s announcement of PVPR took physician groups by surprise. Many, including the Medical Group Management Association, the American Medical Association, and the American College of Physicians, had been working behind the scenes for months in hopes of delaying - or outright scrapping - the program.

Their objections are many, and include everything from the lack of incentives and the use of G-codes to the selection of the measures themselves. The AMA estimated that one physician treating a 70-year-old women for osteoporosis, diabetes, and heart disease could report on 12-13 measures for her alone, and have to consider 36-39 G-codes for the measures.

“Our issue, which we have been raising since last summer, is that by coupling the reporting to the Medicare claims process, you run the risk of greatly disrupting cash flow to the practice,” says William Jessee, MD, MGMA president.

Claims submission would likely be delayed if the physician has to hunt down a specific G-code for an appropriate patient, Jessee says. In addition, many private payers do not recognize G-codes, so if a claim was being sent to both Medicare and the payer, the payer would most likely reject it, Jessee adds.

MGMA has proposed that physicians, like hospitals, be able to submit data quarterly, and through some means other than G-codes. For their part, CMS officials say the G-code method is only temporary - once electronic health records (EHRs) are in more widespread use, the G-codes won’t be necessary.

Of course, that day is unlikely to come soon, as less than 15 percent of practices are believed to have EHRs. And those that do, according to Jessee, say they will need six months to a year to reconfigure their software to capture these codes.


That reconfiguration was just one issue raised at a focus group of 23 practice administrators MGMA convened to talk about the P4P program. Administrators and physicians came from around the country and represented a number of specialties, but they were united in at least one way: All of them said they did not plan to participate in the program, Jessee says.

The focus group members all thought the program was too complex, would result in additional staff time, and believed that many of the measures simply did not relate to their practices. “There are just so many gigantic questions that have not been resolved,” Jessee says.

Trent Haywood, MD, CMS’s deputy chief medical director, says the agency has tried to be responsive. “We recognize some of the issues that MGMA has raised, and we want to work with them [and others] to address the issues, but we want to move forward,” Haywood says.

“There is an opportunity here to work with physicians and start to see how we are doing on reporting this information. What all this should do is help us start reforming the payment process. Keep in mind that this is voluntary - it is a pilot to establish a system for reporting.”

Haywood says the incentive for physician participation is to help them prepare for the future. A physician “would have a risk-free environment” to report data, and “see how my results look relative to the benchmark. “This allows me [as a physician] to see how my results look relative to this benchmark, allows me to determine whether I need to make procedural redesigns or data submission redesigns.”

He notes that private payers are increasingly adopting P4P programs, and members of Congress have put forth bills that would base Medicare payments on such systems in the future. “The momentum is obviously there,” Haywood says. “I think it’s fair to say that’s where everything is headed.”

The California experience

For at least five years, Ronald Bangasser, MD, a family practice physician with the Beaver Medical Group in Redlands, Ca., has been active in P4P initiatives. He serves on a number of national organizations developing reporting measures, and his practice is among the 225 physician organizations currently taking part in a P4P program sponsored by the Integrated Healthcare Association, a California coalition of health plans, hospital systems, employers, and consumers.

The physician groups report on 10 measures and a report card is published with outcomes for each organization overall - data is not physician-specific.

Since 2003, Beaver Medical Group has earned “between half a million and a million” extra from the seven major health plans in the P4P program. Half of the bonus is pumped back into the medical group to fund IT expenditures, while the balance is distributed among the 100-plus physicians in the practice. On average, this has meant an extra $5,000 per physician.

Bangasser calls the P4P program a success, but notes that there are crucial features that have made it so. For one thing, all the health plans agreed on the same measures, so groups did not have to scramble to collect different data for different plans. Starting this year, the groups will report on 15 measures.

Secondly, the program offered financial incentives from day one; and thirdly, no physician is hung out to dry on his or her own.

Bangasser warns physicians against passively accepting P4P programs that are put before them.

“What I try to tell physicians is they have to get involved to bring these [programs] together, because there is no way even with technology that we could report separate measures to each plan.” When considering whether to participate, he says the “most important thing is to have confidence in the organization putting the program together.”

You should also believe in the validity of the measures, and if possible, the reporting should be to an independent organization (such as IHA in California) or the National Committee for Quality Assurance, he adds.


Bangasser is also highly critical of Medicare’s proposal. “It has too many measures, and some of the measure sets are hospital sets. Some of them have no connection to what I do [as a PCP]. It has too quick of a ramp-up. Our recommendation is five years, with consistent dollars behind it,” Bangasser says.

Theresa Defino is an editor of Physicians Practice. She can be reached at tdefino@physicianspractice.com.

This article originally appeared in the February 2006 issue of Physicians Practice.