Thursday, July 21, 2005
Because it is critical for CMS payment systems to support better outcomes for our beneficiaries at a lower cost, CMS is working closely and collaboratively with medical professionals and the Congress to consider changes to increase the effectiveness of how Medicare compensates physicians for providing services to Medicare beneficiaries. I am engaging physicians on issues of quality and performance with the goal of supporting the most effective clinical and financial approaches to achieve better health outcomes for people with Medicare. At the same time, however, we are concerned and are closely monitoring the current volume-based payment system for physicians' services, which projects seven years of negative updates in physician payments. Simply adding larger updates into the current payment system would be extremely expensive from a financing standpoint, and would not promote better quality care. Under this system, there are significant variations in resources and in spending growth for the same medical condition in different practices and in different parts of the country, without apparent difference in quality and outcomes. CMS is committed to working with Congress and the medical community to remedy this situation by developing reporting and payment systems that enable us to support and reward quality.
CMS is engaged with a number of stakeholders to address this situation. For example, MedPAC has made many recommendations to improve the payment system. In addition, a number of specialty societies, including the American College of Physicians, American Academy of Family Physicians, and the Society of Thoracic Surgeons, are in the vanguard of leading change. In addition, like many private-sector health care payers and plans, CMS also is conducting a number of demonstrations and piloting various payment reforms to reward providers for better quality, better patient satisfaction, and lower overall health care costs in the Medicare fee-for-service program.
Madam Chairman, in terms of physician payment, the best ideas of physician organizations, health care experts, the private sector, and our own demonstration programs are all pointing to the same conclusion. Medicare needs to move away from a system that pays simply for more services, regardless of their quality or impact on patient health, and consequently contributes to reductions in the physician update under the current payment formulas, to a system that instead encourages and rewards efficiency and high quality care for the Medicare program and its beneficiaries. As it stands, the physician payment system does not always recognize clinically appropriate care. For example, Medicare will pay for a duplicate x-ray or blood test right before surgery if a hospital does not coordinate care adequately with the physician's office. CMS' physician payment system should support, encourage, and provide an incentive for physicians to achieve Medicare savings by avoiding unnecessary services such as duplicate tests and to use those savings in developing improvements to the physician payment system. In addition, the physician payment system also may steer patient care decisions. Oncologists, for example, are paid less for transitioning a terminal patient to palliative care and focusing on quality of life issues than for recommending and providing intensive procedures even if the side effects of those procedures are significant and the benefits negligible. In addition, currently the payment system does not reward physicians who actively prevent readmissions for patients with heart failure or diabetes.
Physician Payments Based on Statutory Formula
When actual spending exceeds targeted spending, the following year's update is modified to bring actual spending back in line with the targets. Unfortunately, actual spending has greatly exceeded targeted spending, and the formula results in negative updates to physician payments to correct this disparity. Recent rapid growth in the volume and intensity of physicians' services per beneficiary is driving the growth in Medicare physician spending and resulting in the negative updates. Presently, we project a negative 4.3 percent update to physician payment rates for 2006 and additional negative updates for the following six years. The current system eventually corrects the discrepancy and under our latest projections results in positive updates after 2012. CMS is fully cognizant of the potential implications of seven years of negative physician updates, remains concerned, and is closely monitoring physicians' participation in the Medicare program and beneficiaries access. What is especially concerning is that these reductions in the payment rates for services will occur at the same time as Medicare physician spending continues to go up. Projected increases in the volume and intensity of services for 2006 would result in increases in total physician revenue from Medicare for 2006. In other words, our current payment system has a risk in terms of access problems, yet we still are facing sharply rising increases in Medicare spending.
We need to do better. We will continue to work with Congress and physician communities. Although we have not yet seen evidence of a problem, we are closely monitoring access for people with Medicare. Thus, we need to ensure payment for physicians is adequate and appropriate. Both the President's Budget and the Congressional Budget Resolution presume this will be done in a cost-neutral manner. We have worked with Congress and physicians to understand more about why physician volume, and thus spending, is going up, and to develop better approaches to supporting physicians in providing high quality, up-to-date care.
Volume and Utilization Drive Medicare's Increasing Physician Expenditures
Such a large increase in expenditures has significant ramifications for future Medicare spending, and thus, updates to physician payments. This growth has increased the cost of addressing negative physician updates. These increases strain the Federal budget and contribute to annual increases in beneficiary premiums. That is why understanding the sharp rise in these expenditures in 2004 is very important.
CMS' preliminary analysis of the 2004 increases in spending for physicians' services indicates that major contributors to growth included:
CMS has taken collaborative steps to better understand these concerning trends, including what changes in utilization are likely to be associated with important health improvements and which have limited or questionable health benefits. We have been reviewing the technical aspects of this situation in detail with health policy experts as well as the AMA and various specialty societies. For example, the AMA has provided us with some potential reasons accounting for growth. While it was not possible with available data to precisely analyze the impacts of every factor identified, we were able to assess the impacts of most of them. Generally, our results indicate that while the factors the AMA identified have contributed to higher spending, our preliminary analysis suggests that these identifiable factors do not account for a substantial part of the $10 billion spending growth between 2003 and 2004.
In a number of cases, although the rate of spending growth for a particular service may be above average, it is not significantly different than the growth in spending for other, similar procedures where no special clinical developments have yet been identified. For example, the AMA suggested several reasons for a particular increase in the use of echocardiography. However, we compared the growth in these services from 2003 to 2004 with the growth in all imaging services, and found the increases to be the same for both groups (19 percent). Therefore, although the increase in 2004 is large, by itself, the increase in echocardiography services does not account for a disproportionate percentage of the increase in imaging services overall.
For several other services where clinical factors suggested growth might be particularly rapid, we did observe large percentage increases from 2003 to 2004, but the dollar impact was relatively minor. An example is electrical stimulation for bone healing. Spending rose for this service by 35 percent in 2004, but total spending was only $50 million, although a $13 million increase from 2003. This factor would have minimal impact in explaining the overall growth in physician spending from 2003 to 2004.
Of the factors the AMA identified, the one that appears to contribute the most to overall spending growth is the drug Pegfilgrastim (Neulasta). Spending for this drug in 2004 was $518 million, up from approximately $253 million in 2003. We have not been able to confirm the clinical suggestion that by strengthening the immune system of cancer patients, this drug is preventing immune-related complications so that patients are kept out of the hospital, shifting costs from Medicare Part A to Part B, rather than simply increasing costs. But we remain interested in supporting the development of such "preventive" evidence.
We appreciate the efforts of the AMA and the many specialty societies that assisted CMS in identifying these medical trends. They have helped further our understanding of the reasons for the growth in spending. I am sure that all stakeholders involved in these critical payment issues will benefit from an ongoing, evidenced-based dialogue regarding these issues, particularly focusing on which changes in utilization are likely to be associated with important health improvements and which ones have health benefits that may be more questionable.
Options to Change the Physician Update
A second option recommended to CMS would remove prescription drugs from the services included in the SGR, either prospectively or retrospectively. We are currently reviewing the legal arguments regarding whether CMS can take this step under existing authorities, and our actuaries have estimated the payment and budget implications of such changes. A prospective approach would not provide relief to the negative updates projected for 2006 and the succeeding several years. While it would eventually help close the gap between the cumulative target and actual spending, it would not result in positive updates for several years. CMS actuaries estimate this change would cost $36 billion over ten years using the mid-session baseline.
Some have suggested a retrospective approach that would remove drugs retroactively from the services included in the SGR beginning with 1997, the year the SGR was implemented. Some health care trade associations have asked CMS to make this change administratively, as have members of this Subcommittee, most recently Chairman Johnson and Chairman Thomas. However, retrospective removal presents somewhat more difficult issues of statutory authority than prospective removal. For example, the statute requires the estimated SGR be refined twice based on actual data. The crux of the issue is that after the SGR has been refined twice, the statute does not provide for additional revisions, which would be required for at least some prior years should drugs be removed from the system retroactively. In terms of budget impact, we estimate that this approach would cost $111 billion under the Mid-Session Baseline. Moreover, even if the authority existed to remove drugs from the SGR retroactively, positive updates may not occur for physicians in 2006 or the succeeding few years. This is a notable change from estimates using the FY 2006 President's Budget Baseline where the preliminary estimate of the 2006 update would have been positive. The change from a positive to negative update for 2006 is due to higher actual physician spending for 2004 and revisions made to the 2003 data. In addition, prospective or retrospective removal of drugs would increase beneficiary premiums. And because Part B drugs use shows substantial variation across physician practices, payment reforms that provide more support for higher-quality care would probably not ignore drug use, but rather consider how drugs (and other important treatments) are used to lead to better patient outcomes at a lower cost.
It has also been suggested that the SGR be revised to account for National Coverage Determinations (NCDs). The theory is that the spending for NCDs shows up on the expenditure side but the SGR target is not adjusted for them. While coverage of new medical technologies as reflected in NCDs would seem to lead to changes in spending beyond physician control, there has been substantial discussion of this theory in recent years. Much of spending for NCDs is for services covered before the NCD by local carrier discretion. Thus, an NCD does not necessarily increase spending to the extent the local carrier covered the service without an NCD. In many cases, an NCD might simply replace differing local carrier policies with a uniform national policy. In some instances an NCD might limit expenditures if the NCD has narrower criteria than applied previously by local carriers. The use of the real GDP per capita in the SGR formula was intended to be a proxy for a number of factors that might increase the volume and intensity of physicians' services (other than beneficiary enrollment growth and statutory or regulatory changes), including coverage of new services and other factors, whether within an NCD or otherwise. In addition, the NCD impacts are generally small and unlikely to change the physician update significantly. Also, we currently are reviewing the legal arguments on whether CMS can take this step under existing authorities. Nonetheless, CMS will continue to evaluate the evidence related to NCDs and how they are accounted for in physician payments.
CMS will be issuing the proposed physician payment rule for 2006 soon and we welcome comments on these issues and other issues that might affect the physician payment calculations.
Incorporating Performance Based Payments into Medicare
In the FY 2006 budget, the President recognized the potential for payment reforms to improve the value of care delivered to people with Medicare by exploring programs that promote quality in a budget-neutral manner. In its March 2005 Report to Congress, MedPAC offered several recommendations including the development of measures related to the quality and efficiency of care by individual physicians and physician groups. I would like to note that the American College of Physicians is supporting linking physician payment and performance.
CMS is already engaged with the physician community in the development and improvement of specific quality measures. CMS has worked in collaboration with the American Medical Association's Physician Consortium for Performance Improvement and the National Committee for Quality Assurance Ambulatory care to develop measures of improvement in care. This partnership resulted in a set of proposed measures that were submitted late last year for endorsement to the National Quality Forum, a voluntary private consensus setting organization. As part of the Ambulatory Care Quality Alliance (AQA), led by the American Academy of Family Physicians, the American College of Physicians, America's Health Insurance Plans, and the Agency for Healthcare Research and Quality, CMS and other stakeholders, including the American Medical Association and other physician groups, as well as representatives of private sector purchasers and consumers, selected a subset of these measures (26) as a starter set for implementation. Additional measures that assess dimensions of specialty care and efficiency will be added to this starter set. In addition, the AQA is now developing approaches for reporting results to individual patients and physicians and evaluating strategies to minimize physicians' burden of reporting.
The entire starter set of ambulatory care measures are now in the final stages of endorsement. These measures are designed to reflect performance in primary care and also apply to certain specialists, insofar as those specialists are involved in the furnishing of care to patients with common chronic diseases, including diabetes and heart disease. In addition, measures of effectiveness and safety of some surgical care have been developed through collaborative programs like the Surgical Care Improvement Program, which includes the American College of Surgeons. The goal of the Surgical Care Improvement Program is to prevent or decrease surgical complications, in an effort to improve outcomes, and decrease hospital days and unnecessary use of resources. We are also collaborating with many specialty societies, such as the Society of Thoracic Surgeons, to develop quality measures that reflect important aspects of the care of specialists and sub-specialists. For example, we are working closely with oncologists to develop measures of the adequacy of treatment planning and follow-up that oncologists furnish as part of their evaluation and management services; with cardiologists on measures of cardiac care for heart attack or heart failure conditions; and with cardiovascular surgeons on measures related to cardiac surgery. As part of this effort, on July 14, 2005, I sent a letter to a number of specialty societies, summarizing some of the work to date and requesting an update on their efforts to develop quality and performance measures. Historically, CMS has had productive exchanges with most medical specialty organizations, and if an organization has not entered discussions with us, I would encourage them to initiate a dialogue with us as soon as possible so we can work together to develop clinically valid measures and obtain our goal of improving the care we provide the Medicare beneficiaries.
We also are preparing to implement the MedPAC recommendation to use Medicare claims data to measure fee-for-service physicians' resource use and to share these results with physicians confidentially to educate them about how they compare with aggregated peer performance. We are using existing claims data to simulate and test the measurement and quantification of individual physician patterns of practice, incorporating both services they order (including facility services) as well as services they furnish. Resource use is often measured for episodes of care and periods of time (e.g., 3 months). The most widely used measure is total expenditures per episode or period of time. Other measures of resource use are possible, such as examining the percent of a physician's patients who have a particular service ordered. This can indicate potential variations in practice that may affect costs significantly without evidence-based benefits for patients. For example, MRI scans may be ordered for patients with non-specific lower back pain, a condition that often does not warrant the test. By comparing relative use of such a service among physicians, a data-driven foundation for identifying opportunities to avoid some medical costs without harming patients may be developed. As a next step, we expect to begin pilot projects to share the results with physicians confidentially to educate them about how they compare to peers in an effort to decrease the use of unnecessary services.
We also have implemented a number of demonstration projects including one to test pay-for-performance in Medicare's fee-for-service payment system for physicians. The Physician Group Practice demonstration is assessing large physician groups' ability to improve care that could result in better patient outcomes and efficiencies. Ten large (200+ physicians), multi-specialty physician groups in various communities across the nation are participating in the demonstration. These physician groups will continue to be paid on a fee-for-service basis, but they may earn performance-based payments for implementing care management strategies that anticipate patients' needs, prevent chronic disease complications, avoid hospitalizations, and improve the quality of care. The performance payment will be derived from savings achieved by the physician group and paid out in part based on the quality results, which CMS will assess. Providing performance-based payments to physicians has great potential to improve beneficiary care and ensure fair and appropriate payment in the Medicare program.
In addition, CMS will implement the Medicare Health Care Quality Demonstration. This demonstration program, which was mandated by the MMA, is a five-year program designed to reduce the variation in utilization of heath care services by encouraging the use of evidence-based care and best practice guidelines. CMS also is implementing the Medicare Care Management Performance Demonstration, a 3-year pay-for-performance pilot with small and medium sized physician practices that will promote the adoption and use of effective health information technology, i.e., health IT that actually achieves improvements in the quality of care and reductions in preventable costs for chronically ill Medicare beneficiaries. This demonstration will provide performance payments for physicians who meet or exceed performance standards in clinical delivery systems and patient outcomes, and will reflect the special circumstances of smaller practices. This demonstration project will give CMS the opportunity to provide technical assistance to small providers in adopting clinical information technology to improve quality and avoid costs, as CMS has already been working to do in limited pilots. This demonstration, required by the MMA, currently is under development and will be implemented in Arkansas, California, Massachusetts, and Utah. We are supporting an evaluation of this demonstration with AHRQ and insights from health IT implementation that produce improvements in quality and efficiency will be shared broadly through AHRQ's National Resource Center.
Quality Improvement Organizations Assist Physicians' Offices
Over the past year, the CMS California QIO, Lumetra, has been piloting CMS DOQ-IT assistance efforts for over 500 physicians and their offices in California. Many of these physicians' offices are small offices with one or two physicians and are located in rural or underserved areas of California. Lumetra staff and consultants provide consultation and technical assistance for these offices, supporting the clinical process changes resulting from the incorporation of health information technology in their offices, which in turn will allow them to utilize electronic health records, electronic prescribing, decision support and clinical practice guidelines relevant to their patient population, and electronic billing and communications. In addition, QIO staff will assist these offices in implementing office redesign to enhance patient management, and increase office efficiency. All of these efforts are designed to result in enhanced patient safety and better quality of care. Our goal is to help support such effective physician office enhancements becoming standard to all medical practices in the coming years and CMS QIO efforts will help make sure that all physicians' offices can accomplish these enhancements.
The QIOs also have implemented quality improvement projects that lead to better care in rural and underserved areas. For example, Qualis Health, the CMS Alaska QIO, has worked with the almost exclusively rural Alaska providers to increase the rates of preventive services available to rural Alaska residents. Mountain Pacific QIO, the CMS QIO in Hawaii, is working to implement telehealth services to bring care not otherwise available to rural Hawaiian beneficiaries.
Another example of QIO assistance to small physician offices is their assistive role in CMS's release of the VISTA-Office Electronic Health Record Software planned for August 1, 2005. CMS staff has been working with the Department of Veteran's Affairs' (VA) staff to develop an inexpensive and interoperable software package that will allow implementation of a basic electronic health record (EHR) in physician offices. A simplified version of the EHR used in VA Hospitals & Clinics will be stand-alone and allow an in-office EHR that contains computerized medical records, a medication formulary with refill and drug-drug interaction notifications, a reminder system for preventive services and diagnostic tests, and the potential to communicate electronically with other systems in the future. It uses the VA product base which is in the public domain and therefore affordable to small practices taking care of rural and underserved populations. It also is scalable and allows major software developers to devise add-on enhancements. The QIOs will be instrumental in explaining and facilitating the use of this product.
Medicare's Hospital Performance Based Payments Have an Impact
Evidence exists that some hospital admissions are preventable. Heart failure patients have a readmission rate of 21 percent over 30 days, yet research shows that about half of the readmissions are preventable. For example, providing angiotensin-converting enzyme inhibitor (ACEI) drugs to heart failure patients is an example of high quality care, yet ACEI prescriptions are found in only 66 percent of audited patient records. Giving beta-blocker drugs to patients with acute myocardial infarction (AMI) can reduce rehospitalizations by 22 percent, but only 21 percent of eligible AMI patients receive a prescription for a beta-blocker. Pneumonia is a very common cause of hospital admissions for Medicare beneficiaries, but many of these cases could be prevented through pneumococcal and influenza vaccinations. Studies have shown that proper adherence to vaccination protocols can reduce hospitalizations for pneumonia and for influenza by about half, with reduced diseases, mortality, and savings for the Medicare Program.
If physicians are supported in their efforts to better manage patient care, preventable and costly hospitalizations, readmissions and admissions for complications may be avoided. CMS' physician payment system should support, encourage, and provide an incentive for physicians to avoid unnecessary services such as preventable admissions.
The Premier Hospital Quality Incentive Demonstration is a demonstration project to test if providing financial incentives to hospitals that demonstrate high quality performance in a number of areas of acute inpatient care will improve patient outcomes and reduce overall costs for Medicare. We believe that creating incentives to promote the use of best practices and highest quality of care will stimulate quality improvement in clinical practice. Under the Premier demonstration, a hospital can receive bonuses in its Medicare payments based on how well it meets the quality measures. Poorly performing hospitals will face financial penalties in the third year.
Preliminary analysis of the demonstration has shown that quality of care has improved significantly in hospitals participating. The demonstration tracks hospital performance on a set of 34 widely-accepted measures of processes and outcomes of care for five common clinical conditions. The 17 measures included in Medicare's national hospital quality reporting program are a subset of these measures. The preliminary analysis shows improvement in all five clinical areas being tracked in the three-year demonstration. The analysis of first-year performance found median quality scores for hospitals improved:
Overall, these conditions account for a substantial portion of Medicare costs. If we achieve improvements in aspects of care that are proven to help patients avoid complications, patients are less likely to require more costly follow-up care for such conditions, and they are more likely to have a better quality of life. As evidenced by the early work of some of our demonstration projects, we are seeing meaningful results, which are providing a promising foundation to support the most effective clinical and financial approaches to achieve better health outcomes for Medicare beneficiaries.
Private Sector Initiatives Pave the Way for Improved Quality and Efficiency
A large health plan in New Hampshire launched a quality improvement incentive program in 1998, rewarding primary care physicians for the provision of quality care. The metrics for its quality improvement incentive program are the Health Plan Employer Data and Information Set (HEDIS) measures. The program uses claims and administrative data from its disease management program to assess physician practice performance. Incentive payments are awarded to practices scoring greater than the network average. In 2001, the average physician bonus payment was $1,183 and the highest bonus payment was $15,320. In the first year, the plan's average rates for mammography, immunization, and pediatric exams showed increases. Adult female patients receiving Pap smear tests rose from an overall rate of 80 percent in 1999 to 98.5 percent in 2000 for the top quartile of physician practices. For all performance measures for which 1999 baseline data were available, the average incentive program physician practice conformity with performance measures rose from 51.2 percent to 65.6 percent in 2000.
In 2003 a large health plan in Massachusetts launched a group practice incentive program for groups of specialists. Group practices are measured in three categories: patient satisfaction and access, quality of care, and cost. Group practices that perform better than average on the quality measures earn a bonus that could total up to fifteen percent of the regular fees paid to that physician group.
An Illinois coalition of employers initiated a program in 2000 that provides incentives to physicians for monitoring diabetes patients. Compensation is awarded to physicians in the program who meet annual goals in diabetic treatment thresholds. To gain physician buy-in into the program, a committee of physicians developed the performance goals. The coalition and medical group administrators negotiated the amount of the financial incentives a medical group could receive if they met the goals. Results reveal that diabetic care for patients in the program is significantly better than state averages and cost trends for diabetics are better than trends for all other conditions.
A Hawaiian medical association launched a voluntary practitioner quality and service recognition program. Practitioners who enroll share in a multimillion dollar budget earmarked to recognize practitioners for adhering to recognized standards of quality and clinical practices proven by research to improve clinical outcomes. Each program participant receives an award based on his or her scoring in each of the program components – quality indicators, patient satisfaction, and business operations. Practitioners are measured on a total of 68 clinical measures. Analysis of data on key clinical quality indicators over the six years of the program demonstrates statistically significant improved performance.
In Minnesota a health partner's program recognizing outcomes offers annual bonus awards to primary care clinics that achieve superior results in effectively promoting health and preventing disease. Eligible primary care groups are annually allocated a pool of bonus dollars that is awarded if a group reaches specific comprehensive performance targets. Since 1997, bonus awards have totaled over $2.5 million. The impact on quality of care has been substantial. The proportion of diabetes patients meeting optimal care standards nearly tripled since 1999, and the rates of optimal coronary artery disease patients reaching all treatment targets doubled. The rate of members receiving all preventive care doubled. Tobacco use assessment at all visits increased from 45 percent to 85 percent over four years and more patients are routinely provided assistance to quit. Diabetes eye and kidney complications rates dropped by nearly 50 percent, and costs are trending significantly below costs for all other patients. Tobacco use rates dropped ten percent to an all time low. In Minnesota death from heart disease dropped to the lowest rate in the nation and continues to decline.
A health care leadership association of health plans, physician groups, and health systems in California recently implemented coordinated, state-wide pay-for-performance initiatives. Based on a comparison of data from the first year (2003) and test year (2002) nearly 150,000 more California women received cervical cancer screenings, 35,000 more California women received breast cancer screenings, 10,000 additional California children received two needed immunizations, and 18,000 more Californians received a diabetes test. The program paid an estimated $50 million to 215 California physician groups in the pay-for-performance program in 2003 (paid out in 2004), and an estimated total of $100 million to the same physician groups under all of the association's quality programs.
The American Society of Clinical Oncology's Quality Oncology Practice Initiative (QOPI) is an oncologist-led, practice-based quality improvement initiative. QOPI's goal is to promote excellence in cancer care by helping practices create a culture of self-examination and improvement. The process employed for improving cancer care includes measurement, feedback, and improvement tools for medical oncology practices. Practicing oncologists and quality experts developed the QOPI quality measures, which are derived from clinical guidelines or published standards, adapted from the National Initiative on Cancer Care Quality (NICCQ), or consensus-based and clinically relevant. Although the measures not linked to financial reimbursement yet, QOPI is an example of a specialty society-driven quality initiative that can be easily linked to a pay-for-performance program.
Results of these and many more physician-led initiatives lay a sound foundation for CMS to move forward collaboratively with you and with leading physician and health professional organizations with performance based payments for physicians in Medicare to improve quality and efficiency. These approaches also are aligned with emerging requirements from medical specialty boards for maintenance of certification. While recertification has traditionally involved demonstrating cognitive knowledge only, all boards are moving to link maintenance of specialty certification with demonstrated efforts to improve clinical care quality and performance. We recognize that physicians need to be actively engaged in establishing this new direction and will continue close consultation and collaboration to assure improved quality and reduced burden for busy practitioners.
Last Revised: July 22, 2005