feature
Pay for Performance: A National Debate
Is it a quality initiative or a sophisticated means of cost control?
BY WILLIAM L. RICH III, M.D., F.A.C.S.

Washington is abuzz with talk about the newest movement in health care: Pay for Performance (P4P). Naturally, questions arise when new policies or changes occur in reimbursement. Inquiries such as: Is P4P a move to reward physicians for practicing good medicine? A revenue shift to primary care? A move to incorporate evidence-based medicine into practice and payment decisions? A way of pursuing unbiased technology assessment of new modalities of care? A sophisticated means of cost control? It is all of the above.

The following analysis focuses on the system's movement toward adoption and how potential new P4P policies could affect ophthalmologists.

Insurers and policy wonks feel strongly that the current reimbursement system fails to differentiate between physicians who provide good quality care and those who do not. They believe payment should reflect the quality and efficiency of care delivered.

Consumer advocates believe that patients should be able to make informed decisions on physician choice based on compliance with publicly reported performance measures developed by the specialty. Insurers say that there is too much regional variation in healthcare practices, leading to increased costs. Uwe Reinhardt, Ph.D., a noted healthcare economist, has written that a main reason for increased costs is variability in diagnostic testing and surgical rates from one region of the United States to another, with no rationale based on intensity of disease.

Commercial insurers have adopted sophisticated economic profiling to address variability in processes of care. Physicians who make the most efficient use of their resources tend to be paid more. However, these profiles are not reflective of quality of care and are not risk adjusted. Insurers feel the medical profession should develop measures of quality. That way, payers can link payment to the most efficient physicians who offer higher quality of care.

Healthcare policymakers are convinced that quality of care will lead to less morbidity, mortality and lower long-term costs, as diseases are treated at an early stage. The problem on the financial side is that compliance with accepted quality measures will, in the short term, lead to higher costs.

In ophthalmology, during a 7-year period in the 90s when the Medicare population grew 7%, the use of codes related to AMD and diabetic macular edema increased over 24% due to compliance with new recommended patterns of care promulgated by the American Academy of Ophthalmology (AAO) and the retinal community. Will payers accept the short-term costs of implementing good screening and quality measures? They have not shown any proclivity to do so in the past.

Ophthalmology and Quality of Care

Ophthalmology was the first specialty to develop recommended patterns of care. In the mid 1980s, under the leadership of Alfred Somner, M.D., M.P.H., former dean of the Johns Hopkins School of Public Health, AAO developed its Preferred Practice Patterns (PPP). These were consensus documents that are continually and rigorously updated through the AAO Quality of Care Committee. Sadly, they are not widely followed.

The problem is that only 52% of recommended treatments established by randomized clinical trials have been adopted 10 years after publication. This means that continuing medical education has failed to increase compliance with accepted patterns of care; and linkage to maintenance of certification has not worked. Many analysts feel that economic stimuli will lead to greater physician compliance.

With Medicare, the current fee-for-service payment method is characterized by the increased utilization of advanced imaging studies, non-evidence-based diagnostic testing and drugs delivered to patients in the doctor's office, e.g., Visudyne (QLT/Novartis), Macugen ([OSI] Eyetech) and cancer treatments. In Medicare, physicians are paid under Part B which is funded with yearly tax revenues.

Congress mandates that the Centers for Medicare & Medicaid Services (CMS), the administrator of Medicare, update the fee schedule each year and predict the growth of services. When the growth of services exceeds projections, payments are cut in the following years.

The explosive growth in advanced imaging, diagnostic testing and physician-administered drugs will lead to payment cuts of 37% over the next 9 years. Congress and the Bush administration are not inclined to fix the broken Medicare physician fee schedule until the profession agrees to adhere to reported public compliance with quality measures, which will result in payment differential among physicians.

Physician Performance

The American Medical Association (AMA) Consortium for Physician Performance Improvement was formed to evaluate the legitimacy of measures developed by specialty organizations. A consensus definition of physician quality measurement has come from the consortium: "Physician clinical performance assessment evaluates an individual physician's clinical practice behavior and adherence to objective standards of clinical process and outcomes of care, thereby holding the physician accountable for and rewarded for practicing according to those standards." Along with a definition, four physician-performance categories have been developed from an assortment of quality assessment organizations, insurers and physicians. The measures include:

■ Information technology. These measures are intended to help spur physician investment in electronic prescribing systems and electronic medical records (EMR). However, the proposed economic stimuli are not adequate to affect change.

EMRs are expensive to develop and maintain. They are financed by doctors and essentially benefit insurers. They lack a common infrastructure and are not linked to office business practices. There is no economic incentive for a large software vendor to make the huge investment needed to develop a good product. Some would say that any investment in EMR at this time is premature. I would agree.

■ Process of care. This will constitute most ophthalmic P4P quality initiatives. Process measures are the application of approved patterns of care to the patient: interval of visits, timing of diagnostic testing and indications for therapeutic intervention.

This is what we do day-in and day-out as ophthalmologists. How do we treat the diabetic who returns for an annual exam? Is the status of the retina described? Is the patient told of the importance of low HgA1C levels and the chances of retinopathy progression? Is this information documented in the chart? Are the results communicated to the primary care doctor? Is a follow-up appointment made at an appropriate interval?

■ Outcomes. These are the results of our therapeutic and surgical interventions. This is what most ophthalmologists think of when quality of care is mentioned. What is my rate of dropped nuclei, vitreous loss, cystoid macular edema and visual outcomes associated with cataract surgery?

However, there are no validated, risk-adjusted ophthalmic outcomes measures. These would require huge risk- adjusted databases, which are expensive to develop and maintain and labor intensive to document. True outcomes measures will require several more years of development.

■ Efficiencies. This is intended to look at a doctor's actual costs of fulfilling a quality measure compared to the anticipated or estimated costs. Those whose actual costs exceed anticipated costs will get paid less. Those whose costs are less will be paid more. Some scenarios would divide the savings between the physician, the employer and the insurer. Obviously, the big problem is how are anticipated/expected costs determined and how do ophthalmologists adjust the anticipated costs to meet the increased resources needed to provide the care to a sicker patient population. If this is not done properly, physicians will withdraw from participation. It could also lead to adverse selection where physicians are given strong incentives not to care for sicker patients. Either outcome would lead to P4P failure.

Measure Development and Implementation

The anticipated sequence for P4P involves physician-measure development, measure validation and implementation. Once measures are validated, voluntary reporting for primary care physicians could conceivably begin late this year, as requested by CMS. The voluntary reporting program will expand to most medical specialties in 2007. Once data is collected, planned public reporting of physician "scorecards" and the instituting of efficiency measures will come in 2008. This is an ambitious timeline.

The AMA consortium made a pledge to Congress to work with specialties in the consortium to develop 140 measures by the end of 2006.

The AAO, working with ophthalmic subspecialties societies, has formed an Eye Care Work Group chaired by Paul Lee, M.D., J.D., from Duke University. After reviewing the AAO PPPs, the work group has developed 11 measures, including ones for glaucoma, cataract, AMD and diabetic retinopathy, which were submitted to the consortium in the spring of 2006. They were sent out for public comment and hopefully will be approved this summer.

To obtain approval, a proposed measure must: address a demonstrated deficiency in care; have meaningful health impact; have a strong level of published evidence; and be practical enough to collect data using claims rather than chart review.

After measures are developed through the AMA consortium, they are validated by the National Quality Forum (NQF), a multi-stakeholder group, which includes payers, patient advocacy groups, CMS and physicians.

I have great concern about whether the NQF has the infrastructure in place to meet this exploding demand. The group only meets twice a year, has a cumbersome structure with a hierarchy of people with whom to communicate and only recently hired a new director.

Validated quality measures are then sent to the Ambulatory Quality Alliance (AQA) for an assessment of the feasibility of implementation. The AQA is composed of insurers, physician groups, CMS, the Agency for Health Care Research and Quality and employers. The AQA will design the software for the measures and ensure that the measures are practical to implement, allowing easy collection and collating of results for public reporting and the development of efficiency measures.

Efficiency Measures

Is the whole P4P effort all about improving the health care of patients and furthering patient informed choice? Hardly. Never forget the old saying, "When they say it is not about the money, it is always about the money!"

In the 1990s, commercial insurers used proprietary software to develop crude economic profiling of physicians. They lumped all ophthalmologists together and developed expected costs and resources used in providing care for a patient with a disease over a defined time period. For example, a referral cataract surgeon is expected to perform more surgeries per 100 patient encounters than a generalist. Yet, these surgeons would get threatening letters from insurers implying overutilization. Glaucoma specialists were always singled out for their higher use of visual fields when compared to optometrists.

While that program failed, there is the real possibility that economic profiling will be reintroduced when P4P is adopted.

AAO's Role in the Fight

AAO has expressed concern with the advancement of efficiency/cost-of-care measures by the AQA prior to the development of linkage to quality measures. Until recently, this process has been dominated by insurers.

Insurers have gotten the strong message from the AAO that this is unacceptable economic profiling and has nothing to do with quality improvement efforts. As a result, the AQA has made a strong commitment to tying quality to efficiency before cost-of-care measures are implemented. There are huge problems in the other key issue related to efficiency measures — risk adjustment. The AAO and other specialties will be monitoring this to ensure that any linkage of payment to quality has enough consideration to patient disease status.

The AAO has established several goals in this rush to P4P. First, to ensure that ophthalmologists can "play" in the P4P arena, we must have validated, simple quality measures developed by the profession and not insurance industry. The AAO work group is representative of all ophthalmic specialties and has been successful in measure development. In addition, we want to have representation on the NQF and the AQA, which will be ultimately responsible for measure validation and implementation.

This is a political process. The AAO has been successful in recruiting a cadre of hard-working volunteers to staff these committees to make sure the profession is treated equitably. Finally, the AAO will fight to ensure that efficiency measures are truly quality linked, risk adjusted and that their development is not co-opted by the insurance industry.

William L. Rich III, M.D., the AAO medical director of Health Policy. He can be e-mailed at hyasxa@aol.com.