This is an archive page. The links are no longer being updated.

Testimony on Evaluation Science & Our Health Care System by John M. Eisenberg, MD, MBA
Administrator, Agency for Health Care Policy and Research
U.S. Department of Health and Human Services

Before the House Ways and Means Subcommittee on Health
April 17, 1997

INTRODUCTION

Good Afternoon Mr. Chairman and members of the Subcommittee. It is a pleasure to appear before the Subcommittee in my first full week as the new Administrator of the Agency for Health Care Policy and Research (AHCPR). I enjoyed my work for you as Chair of the Physician Payment Review Commission, and I look forward to working with you in the yew to come. I especially want to thank you for giving me the opportunity today to address the issue of evaluation science and how it relates to coverage decisions in our health cam system.

Market forces have spawned unprecedented changes in our health cam system. The, health we market today is characterized by intense price competition, with purchasers demanding greater accountability and value. These pressures are leading health plans and facilities to give increasing scrutiny to capital investments and decisions to cover new medical procedures or services. In an atmosphere of constrained resources, it Is hardly surprising that we arc facing increasing demands to prove that we are getting real value for the health care dollars we spend. Unfortunately, decision makers often lack the scientific evidence on effectiveness that would help them judge the relative value of alternative health technologies.

One result of our shortage of scientific evidence is our health care system's wide variation in use of services. Often, this occurs when them is ambiguity on the effectiveness of the procedure or service. There is also variation in what services are covered. For example, a recent study on laser therapy demonstrated that there is substantial variation in coverage on new technologies. A 1994 General Accounting Office Report found substantial variability among Medicare carriers with respect to denial rates for services screened for medical necessity (GAO Report, 1994).

Variation also exists in medical practice among states and local jurisdictions. For example, the rate of prostate biopsy in Dallas, Texas is 29.6 per thousand of Medicare enrollees, compared to McAllen, Texas where it is 8.3 per thousand of Medicare enrollees (Wennberg, 1996). Health services research has shown that these differences can be attributed in large pall to a lack of knowledge regarding the efficacy and effectiveness of medical technologies. (Steiner, 1996 & 1997). A study examining the high variability in the m of electroconvulsive therapy found that a primary cause was the lack of knowledge regarding the effectiveness of the procedure (Hermann RC, 1995).

Another consequence of inadequate evidence on effectiveness is that old technologies linger even when they have been surpassed by better ones. in 1989, 1 published a study that examined old and new technologies in hospitals and found that new technologies did not replace older technologies, but were simply added to the hospital regimen. For example, we found only a small reduction in one older technology, oral cholecystogram, after the introduction of a new and closely better technology, gallbladder ultrasound (Eisenberg, 1989).

The lack of scientific evidence on the effectiveness of medical procedures and technologies is troublesome: it complicates the treatment decisions physicians and patients face, makes it difficult for plans to make appropriate coverage decisions, and can lead to appeals of coverage decisions in court.

Technology and the Cost of Medical Care

Mr. Chairman, let me make two points at the outset. First, when I refer to technology, I am not only referring to expensive high-technology devices such as magnetic resonance imaging, but also to the vast array of "little ticket" technologies, such as pap smears and other common medical procedures and services. These small-cost technologies can add up to big expenses. For example, an AHCPR funded recommendation for preventing pressure ulcers led to the finding that use of low-tech treatments for pressure ulcers could, in just two states, New Hampshire and Vermont, result in Medicare savings of over $1 million annually.

Second, as my earlier prostate example suggests, the need for scientific information related to the appropriate use of technology does not end with a medical coverage decision. Even when medical coverage is not an issue, clinicians and patients need the latest scientific information to determine whether the procedure or service is appropriate in each specific instance. They need to balance carefully the risks and benefits of every medical intervention and any alternatives for each individual patient's circumstances.

This is critically important because there is a growing body of scientific research that demonstrates that patients often receive services that are not clinically indicated. For example, researchers at the RAND Corporation found that up to 30% of patients whose medical records they reviewed received medical procedures that were not appropriate when measured against rigorous clinical standards established by experts. They also concluded that the appropriateness of another 10% of procedures was equivocal (RAND Studies). Other researchers have estimated that only 20% of technologies now used in the practice of medicine have been specifically evaluated (OTA Report, 1994).

When medical technology is used appropriately, it can improve health, and in some instances, reduce costs. For example, in l991, AHCPR conducted a technology assessment on laparoscopic cholecystectomy on behalf of Health Care Financing Administration (HCFA). Based on AHCPR's recommendations, HCFA covered this procedure, which has subsequently resulted in reduced hospital length of stays. This finding is borne out in a recent study funded by AHCPR, in which patients with laparoscopic cholecystectomies had hospital stays that were one-third less than those who underwent the standard surgical procedure.

Another example is AHCPR's assessment of routine laboratory tests in end stage renal disease (ESRD) for patients undergoing dialysis. In 1994, there were 200,000 individuals receiving dialysis at a cost of $7 billion annually. An AHCPR health technology review found that them was little scientific data to support the routine use of some of the tests performed in patients undergoing dialysis. Clearly, millions of dollars can be attributed to the cost of testing - some of which is unnecessary.

At the same time, technology can increase costs. Excessive and inappropriate use of now drugs, devices, and other services is driven by many causes -- economic incentives for physicians and hospitals, the practice of defensive medicine, and insurance systems that pay for technological interventions more generously. Poor understanding concerning the effectiveness and outcomes of medical technology is also a contributing factor.

Another problem is the use of ineffective technology, which can raise costs but not improve patient outcomes. Also, the rapid rate of development and diffusion of new and promising technologies often means that timely data are not available to make a scientific judgment regarding their effectiveness. A recent study found that a majority of medical directors of health plans cited the lack of timely data on the safety, effectiveness, and cost-effectiveness as the major barrier to establishing optimal coverage decisions. As a result, we need to develop innovative approaches to acquire the necessary data (Steiner, 1996).

AHCPR's technology assessment on lung volume reduction surgery (LVRS) is a case in point. This technology assessment concluded that there was insufficient evidence on which to make a scientific judgment regarding the effectiveness of LVRS. AHCPR recommended that coverage be granted within the scope of a clinical trial, which is now being conducted by the National Institutes of Health. AHCPR is supporting the cost-effectiveness component of that trial. It is significant that this study will also include measures of patient preferences and of quality of life, because the role of these factors are an important component of the scientific knowledge we hope to derive. It is our hope that the collaborative efforts between the NIH and AHCPR will yield the information needed to make an Informed coverage decision on LVRS.

Another mechanism that should be explored u a complement to clinical trials to evaluate technologies is the "Centers of Excellence" approach that was used in coverage of cardiac transplants. These Centers are defined by having a demonstrated expertise in performing a given procedure or technology. Additionally, patients are registered and outcomes and effectiveness data are collected throughout the episode of care to help further refine the technology or procedure being performed.

Working Together

Approaches involving cooperation between an insurer (in this case, Medicare), the scientific community, and advocates of new technology will become increasingly essential in the years ahead. Unless we address the need for solid. scientific evidence on the benefits and costs of medical technology, we run the risk that future efforts to contain health cue spending will inhibit innovation. This could mean that better technology -- even more cost-effective technology -- will not be developed, and American leadership of a global industry could be lost.

The role of the government is pivotal in this regard. h is the largest purchaser and provider of health cm services, accounting for at least 45% of all health care spending. For some procedures federal dollars account for the lion's share of expenditures. For example, Medicare and Medicaid expenses for coronary stay bypass graft procedures exceed $8.7 billion and represent 56% of the total expenditures for this procedure (AHCPR, 1992). As a purchaser of medical care, the federal government has a compelling interest in technology assessment and those decisions need to be based on the best scientific information available.

Through the work of the National Institutes of Health and AHCPR, the federal government also helps to build the underlying research, the basic science for technological innovation and its assessment. In addition to conducting technology assessments for HCFA and the, Department of Defense, AHCPR supports work that is critical to the field of technology assessment, such as:

• the effectiveness and cost-effectiveness of alternative treatments and services provided in community settings (including stroke and cardiovascular disease, prostate disease, diabetes, and pneumonia and other common conditions);
  
  
• research on the functional outcomes of treatments for patients and the importance of involving patients in treatment choice. For example, an AHCPR research project developed a "Shared Decision Making Program," which included an interactive videodisc for use in helping patients with benign prostatic hyperplasia make informed choices among treatment options. Pilot studies have shown that patients who view the videodisc choose surgery less often Ow other patients;
  
  
• the development of medical information technologies such as computerized patient records, computerized decision support systems, hospital and ambulatory care data systems that provide the ability to conduct this research while providing additional resources to clinicians. For example, an AHCPR-supported study found that AIDS patients who use the Comprehensive Health Enhancement Support System (CHESS) had fewer visits to doctors office, or spent 17% less time in doctors' offices. These patients have lower health care costs and fewer and shorter hospitalizations. AIDS patients using CHESS had 20% lower treatment costs overall; and
  
  
• research to improve the methodology underlying technology assessment. For example, AHCPR supports the development of outcomes and effectiveness measures such as the "VF-14," an index of functional impairment in patients with cataracts.

Our research demonstrates that technology assessments, done scientifically, I can provide the infrastructure to help reduce practice variation and to allow for informed decision making. This summer we will implement our new Evidence-Based Practice Center program through which we will provide public and private sector organizations with scientific information to improve the quality, effectiveness, and appropriateness of clinical practice. The Centers' evidence reports will provide critical evaluations of the available scientific literature regarding clinical interventions and technologies. The topics for the evidence reports will be selected from nominations by the private and public sectors. With the support of the Appropriations

Technology Assessment Criteria

• Use of in explicit and defined process
• Careful selection and definition of the topic
• Public participation
• Comprehensive search for the highest quality evidence available
• Review and rating of the evidence
• Use of analytic techniques appropriate to the level of evidence found
• Review of the draft assessment both by experts and potential users
• Update of the assessment when appropriate
• Evaluation of the process and its results

Committees, we have set aside funds for this effort and will fund about 10 evidence based practice centers this summer. The Centers will better enable AHCPR to serve as a "science partner for both the public and private senior. These reports will also assist HCFA, DoD, VA, states, and private sector purchasers to make informed decisions on the effectiveness or appropriateness of specific health care technologies.

Next Steps

Mr. Chairman, from my work in the area of technology assessment and the experience of AHCPR in technology assessment I conclude that we face a number of important challenges in this area:

• Both the public and private sector need to devote more resources to technology assessment. Choices are being made everyday in the health care system. These choices run the risk of being determined solely by cost considerations, unless we assure that the scientific evidence is available to make informed decisions. Failure to address this will put. medical innovation at risk. It is interesting to note that, as a proportion of national health care expenditures, total funding for technology assessment in other nations is 18 to 160 times higher than in the U.S. For example, technology assessment expenditures in France we 22 times higher, (Holohan, Lancet 1996) despite the fact that its population is 2 times greater than the Medicare population alone.
  
  The results of technology assessments need to be In the public domain so that a potential users of the technology or service can be informed. The important methodological details of publicly funded research are by definition available to all those interested in bow results were reached. Interested parties may include patients/consumers, providers, plans, and providers. The same acres may not be granted for research done for proprietary reasons. Additionally, small firms such as biotechnology start-ups, may not have the resources to conduct effective outcomes studies alongside their basic science and clinical investigations.