Baby steps to an iffy end: pay for performance

April 2007
Feature Story

Karen Lusky

The handwriting on the wall for pay for performance for physicians got a lot easier to decipher this year. All you have to do is take a look at the Physician Quality Reporting Initiative, which some say paves the way for a future with pay for performance in place.

Authorized by a provision in the Tax Relief and Health Care Act of 2006, the quality reporting initiative allocates $300 million overall to pay physicians a 1.5 percent bonus for voluntarily reporting CMS-approved performance measures for services paid under the physician fee schedule and provided in the last six months of 2007.

The initiative doesn’t rise to the level of CMS’ conceptualization of pay for performance or what the agency calls value-based purchasing. That budget-neutral proposition would pay providers with top-tier scores on quality measures more by docking those with scores on the wrong end of the bell curve. But the initiative does assemble the necessary elements to move to that type of system if Congress approves it.

How the reporting initiative will play out is still a matter of speculative debate. Knowing just how the program will work in 2008 will require more legislation, directives from the CMS, or both, says American Medical Association board chairman Cecil Wilson, MD, an internist in private practice in Winter Park, Fla. But the prospects for it moving forward are good, in his view. “It’s clear on both sides of the aisle that Congress is interested in knowing they are getting quality care from physicians. ... The question is whether Congress will pay more for quality or just say everyone has to provide it.”

Many see pay for performance as a genie out of the bottle that could, in theory, make some wishes come true for the cost-beleaguered Medicare program and even the health care system overall.

For example, some hold out the hope that a pay-for-performance program encouraging best practices across physician specialties and the continuum of care could improve patient outcomes and produce cost savings for big-ticket chronic illnesses.

“Pay for performance is one of those ideas whose time has come,” says consultant Charles Root, PhD, president of CodeMap Inc., Barrington, Ill. “As usual, the devil will be in the details in terms of whether it buys a lot of improvement.”

For physicians, a chief part of those details is now in the hands of professional organizations, including the CAP, that are working with the AMA to develop performance measures that can be used to report quality.

The AMA-convened Physician Consortium for Performance Improvement is spearheading the effort. The consortium consists of national and state medical societies, the CMS, and the Agency for Healthcare Research and Quality. The AMA’s guidelines for pay-for-performance programs say the goal of any such program must be to promote quality patient care, “rather than to achieve monetary savings.”

AMA’s Dr. Wilson says the consortium first selected measures that “cover the largest percentage of, in this case, Medicare patients, including, for example, the major diseases,” such as heart disease, diabetes, and lung disease.

For the reporting initiative this year, the CMS approved 77 quality measures, none of which applies to pathology. The CAP is, however, advancing two measures—one for reporting breast cancer resections and another for colon cancer resections—to participate in the program in 2008.

It is the lead organization working with the AMA to develop pathology performance measures, work that was already underway within the CAP when it took on that leadership role in mid-2006, says David L. Witte, MD, PhD, chair of the CAP Performance Measures Development Working Group.

At that time, he says, the deadline for pay-for-performance measures to come forward was March or April 2007. But Congress passed the legislation in December 2006 authorizing the reporting initiative, which moved the deadline up by about 2.5 months. And “it was hard for the AMA and CAP to turn the ship and make the other port in January,” says Dr. Witte, who also co-chairs the Pathology Work Group for the AMA physician consortium, which includes pathologists, breast surgeons, colorectal surgeons, gastroenterologists, medical oncologists, radiation oncologists, epidemiologists, and statisticians.

The College also didn’t want to rush through measures for a payment process where future stakes are likely to be high.

“It’s so critically important that the College get it right,” says Patrick L. Fitzgibbons, MD, of the Department of Pathology, St. Jude Medical Center, Fullerton, Calif., and a member of the CAP Performance Measures Development Working Group and the Pathology Work Group for the AMA consortium. This is assuming that physicians will eventually have to show they have done the performance measures to “capture the full billed charge,” he says.

At CAP TODAY press time, the two CAP-proposed performance measures were posted on the AMA Web site with a 30-day public comment period (www.ama-assn.org/ama1/pub/upload/mm/370/pathologypc1.pdf).

Once the draft measures go through the public comment period, they head to the full AMA physician consortium for approval. After that, they go for approval to a separate organization, such as the National Quality Forum or AQA Alliance.

The next step would be for the CMS to include the pathology measures in the Physician Quality Reporting Initiative for 2008. The CMS will adopt additional physician performance measures for 2008 through public rulemaking. To qualify for inclusion, measures have to be “adopted or endorsed by a consensus organization, such as the AQA Alliance or National Quality Forum,” according to a recent CMS “MLN Matters” article (www.cms.hhs.gov/mlnmattersarticles/downloads/mm5558.pdf). The CMS-approved measures will include those submitted by physician specialties and that have been developed through a consensus process.

The CAP chose the initial two performance measures for breast and colon cancer because “any proposed measure has to address some sort of gap in care,” Dr. Fitzgibbons says. And the CAP’s Q-Probes studies show gaps in terms of pathologists reporting all of the required elements in the breast and colon cancer checklists.

The draft breast cancer measure, for example, measures the percentage of breast cancer resection pathology reports that include the pT category (primary tumor), the pN category (regional lymph nodes), and the histological grade. The numerator for the measure equals the number of breast cancer resection pathology reports that include these three required elements. The denominator is the number of breast cancer resection reports (excluding biopsies) defined by CPT and ICD-9 codes. The clinical measure doesn’t include pathology reports for the same breast neoplasm up to six months following the index resection pathology report.

“All a pathologist needs to do to comply with the breast cancer and colon cancer measures is to use an appropriate checklist when signing out an applicable case,” says Paul N. Valenstein, MD, president of Pathology & Laboratory Associates, Ann Arbor, Mich., and a member of the CAP performance measures working group. He compares the process to a pilot running through a pre-flight checklist before takeoff.

The College will continue to take the lead role in developing more pathology performance measures for the reporting initiative, with a goal of coming up with measures that apply to as many pathologists as possible, though the first two measures the CAP proposed “only encompass a small portion of pathologists’ practice,” Dr. Fitzgibbons notes.

Measures under consideration at this point do not address the role of the pathologist in helping clinicians order and interpret tests related to areas outside of anatomic pathology, says Michael Laposata, MD, PhD, a clinical pathologist at Massachusetts General Hospital, Boston, and a member of the CAP performance measures working group. In that regard, he likens the process of developing measures thus far to “being asked to feed someone a large plate of spaghetti and you put a small carrot on the dish instead.”

“At least it’s food, and we are doing something and it’s positive,” he says. “But it doesn’t match what clinicians are begging for—a consultative dialogue about their patients’ lab tests that is like their interactions with other specialists.”

For example, he says, a clinically valuable, “patient-specific, expert-driven narrative interpretation” of a lab test might read: “This 25-year-old woman was found to have a factor V Leiden mutation in heterozygous form. The recent initiation of oral contraceptives, her leg injury from a fall, and finally, the six-hour flight from Boston to London likely explains the development of her deep venous thrombosis.”

Dr. Valenstein points out, however, that creating quality measures for pathologists’ consultation with clinicians might require developing separate measures for each clinical scenario and type of interaction. “Such measures don’t fit into the ‘report card mantra’ that dominates current government thinking and relies upon very simple measures,” he says.

Physicians who voluntarily report approved quality measures for 2007 will do so through the claims process. The measures are reported using a category II CPT code on a claim for Part B services.

Of course, reporting quality measures through the Medicare Part B claims process adds a new layer of work and costs for pathologists and other physicians.

There is a concern “that this type of approach has significant upfront administrative costs,” says Fay Shamanski, PhD, CAP assistant director of public health and scientific affairs. “A particular physician group may not be set up to report the measures, or the way they report them may not be amenable to this type of program.” There also may be personnel training costs.

Pathology is better positioned than some specialties in that electronic systems for coding are common. “But there will be some extra work involved in complying with reporting the quality measures—there’s no question about that,” says Raouf E. Nakhleh, MD, a member of the CAP performance measures working group and a professor of pathology at Mayo College of Medicine, Jacksonville, Fla.

For example, pathologists may face a learning curve in being able to produce easily auditable reports. They may have to include in a report “some pertinent negative,” which they don’t always do, he says.

Say the pathologist receives a breast re-excision as a secondary procedure where the surgeon took more tissue because of positive margins or a close margin. The pathologist would have to “not only report that there’s no residual tumor, but also include key words such as ‘staging not applicable’ or similar information,” he says. Otherwise, a Medicare contractor auditing the report might misinterpret it as the pathologist failing to comply with the performance measure.

On the prospect for and potential impact of pay for performance overall, pathologists and other experts tread cautiously, and even warn of its dangers.

To Dr. Fitzgibbons, pay for performance for physicians is a work in progress and one that will require waiting to “see how it pans out.”

“There’s a lot of concern that it will introduce a new bureaucracy for the payment process without additional payment to offset the additional work required. There could be a great deal of unhappiness about it if it takes that direction,” he says.

Stephen Black-Schaffer, MD, associate chief of pathology at Massachusetts General Hospital and a member of the CAP performance measures working group, says he’s a “little more agnostic than some advocates of pay for performance in terms of where we are going to end up.”

“Pay for performance is not simply a quality initiative designed to pay those who do better more money. It’s also intended to make sure the government pays less than it would otherwise.”

In Dr. Nakhleh’s view, a “pay-for-performance system would drive poor performers to change.” But the reduction in poor performers would skew the budget-neutral payment scheme.

“If the system is [designed] to take from the bottom to give to the top, and everyone moves to the top, where do you get the money?” he asks. Or does that measure “become a de facto standard and we move to another measure?”

Pathologists also worry that where Medicare goes with pay for performance, private payers will follow—but potentially without the rigorous process that the federal payer is following in creating and vetting measures.

Private payers “could take a performance measure and decide that it’s a standard of practice for payment purposes,” says Dr. Nakhleh. As an example, he says, many private payers have “grasped onto a recommendation” that requires surgeons to provide at least 12 lymph nodes in colectomy specimens for cancer. But that standard was studied thoroughly only after it was in place, he says. And “we are finding out that you cannot guarantee 12 lymph nodes in 100 percent of cases.” That’s because many factors affect the number of lymph nodes, such as the specimen size and location and the surgeon’s technique. And “some patients have fewer lymph nodes than others.”

E. Randy Eckert, MD, of the Department of Pathology at North Austin (Tex.) Medical Center and chair of the CAP Council on Government and Professional Affairs, sees “so many pitfalls” to pay for performance that it’s a “dangerous game to play in many ways,” whether implemented on the private pay or federal level.

“You can find yourself in scenarios where you are doing things for the payer that don’t have any benefit for the patient—you have to do them to get paid,” he says. “There are also things that physicians may not be doing that they should do for patients but the evidence doesn’t exist yet to show that those are best practices.”

Dr. Eckert points out that some large medical practices, such as Kaiser Permanente, and some academic centers, are looking at patient outcomes for physicians who treat comparable patients. But “the approach is not to say Doctor X is doing better than Doctor Y so let’s pay Doctor X more. It’s to try to identify what Doctor X is doing so Doctor Y can do the same,” he says.

But “some of what Doctor Y is doing may be intangible or not measurable, for example, communicating more effectively or being more empathetic. That’s the art of medicine.” The aim of pay for performance is to reduce what physicians do to pure science. “But there are so many unmeasurable aspects of medicine that contribute to positive outcomes. Trying to quantify them all is fruitless,” Dr. Eckert says.

Yet pay for performance may bear some fruit for pathologists and other providers. For example, Dr. Laposata believes private payers might be interested in a performance measure that creates an incentive for pathologists to help clinicians order the right tests and reach a diagnosis more quickly. If the payers realized that approach made it possible for them to save money on the total care of the patient, they “might be willing to spend more money on lab tests and pathologists who interpret them…,” he says.

Another plus for some: Pay for performance could drive up the volume of the laboratory tests required to meet performance measures, such as doing A1c monitoring for diabetes. CodeMap’s Dr. Root says, “Entities that insert themselves into this pay-for-performance equation, by providing a test or something to sell people or they have to buy, will profit from it.”

In fact, “pay for performance will be a system that people will learn to play and use to their advantage if they are smart,” he predicts.

A major question on the payment table is whether pay for performance tied to best practices for disease management could eventually improve clinical and fiscal outcomes for chronic illnesses, potentially leaving more money in the pot for providers.

If pay for performance for Medicare “ends up being coordinated with long-term care chronic disease management standards of practice, where the two are seriously connected, then you could give practitioners more money because you are presumably saving more. You would then possibly force out some of the low-quality practitioners,” Dr. Root says.

But for the private pay sector to profit from a disease management approach tied to pay for performance, it would have to become a standard of care for many services. Otherwise, Dr. Root says, health plans “will rationalize, ‘Why keep a middle-aged person healthy with testing and preventive services when that person just leaves the plan and goes to another one next year?’”

Ultimately, the CMS would like to ensure that patients with chronic diseases receive the care that meets best-practice standards across settings, Dr. Eckert says. “That’s really where you can make the biggest improvement in patient outcomes.” And “in aggregate, you can save money.”

But there are perils to using a population-based approach to payment that ignores the individual needs of patients, Dr. Eckert says. A patient with severe diabetes, for example, needs the tight glucose control, including A1c testing and other services to detect complications. But a patient with borderline diabetes doesn’t require all that intervention, which not only is expensive but also poses risks.

Internist Rodney A. Hayward, MD, says implementing evidence-based measures could improve patient outcomes if the evidence-based measures of performance “target the highly important care where interventions make a difference.” Some of those interventions will involve prevention, and others will apply to the high-risk patient who already has complications, says Dr. Hayward, professor in the Department of Internal Medicine at the University of Michigan and director of the Veterans Affairs Ann Arbor Health Services Research and Development Center of Excellence.

He says, “To get pay for performance right, we have to ask what is really quality in health care and what are the things we do in medicine that help decrease mortality and morbidity, and are we willing to invest in the cost required to measure those things reasonably well?”

The biggest challenge in pay for performance, he says, “is that it could be well worthwhile if it’s done right and worse than doing nothing if we do it wrong. It’s very easy to do more harm than good in medicine, and this is a very powerful intervention—people pay attention when money is involved. We know from a wealth of data in business and economics that people assume the incentives you give them are the right incentives.”