1
Clinical Trials in the United States

BACKGROUND ON CLINICAL TRIALS

A clinical trial is a formal study carried out according to a prospectively defined protocol. It is intended to discover or verify the safety and effectiveness in human beings of interventions to promote well-being, or to prevent, diagnose, or treat illness. Other definitions are more expansive—including even the first use of a new intervention in a human being, without a formal plan or any type of comparison—or more restrictive—including only studies comparing two or more interventions concurrently.

Properly conducted clinical trials are a necessity in health care because very few interventions produce such large or striking results that they can be evaluated by observation alone. Most often, the effects of an intervention are modest, perhaps a reduction of 10 percent in the risk of an important outcome. Such effects can be extremely important, however, especially when the endpoint is death from some common disease that kills thousands (or tens of thousands) of people each year. Differences of this magnitude cannot be detected reliably against the background of chance and other influences without a carefully planned and controlled study (Hennekens and Buring, 1987).

The common image of a clinical trial is the comparison of two (or more) interventions—new versus old (or versus placebo)—to see which one works better. Such trials are, in fact, relied upon to produce sound evidence for rational decision making in health care. To generate the most reliable information, clinical trials require certain design characteristics (particularly assignment of participants to interventions by ''randomization"), and they must include enough participants to exclude the play of chance as a likely explanation for results. Regardless of the sophistication and complexity of the design and analysis, the question of whether "a" is better than "b" is the essence of the clinical trial. The interventions that might be tested go beyond treatments (pharmaceutical, biologic, radiologic, surgical, or other procedures) to include preventive strategies,



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials 1 Clinical Trials in the United States BACKGROUND ON CLINICAL TRIALS A clinical trial is a formal study carried out according to a prospectively defined protocol. It is intended to discover or verify the safety and effectiveness in human beings of interventions to promote well-being, or to prevent, diagnose, or treat illness. Other definitions are more expansive—including even the first use of a new intervention in a human being, without a formal plan or any type of comparison—or more restrictive—including only studies comparing two or more interventions concurrently. Properly conducted clinical trials are a necessity in health care because very few interventions produce such large or striking results that they can be evaluated by observation alone. Most often, the effects of an intervention are modest, perhaps a reduction of 10 percent in the risk of an important outcome. Such effects can be extremely important, however, especially when the endpoint is death from some common disease that kills thousands (or tens of thousands) of people each year. Differences of this magnitude cannot be detected reliably against the background of chance and other influences without a carefully planned and controlled study (Hennekens and Buring, 1987). The common image of a clinical trial is the comparison of two (or more) interventions—new versus old (or versus placebo)—to see which one works better. Such trials are, in fact, relied upon to produce sound evidence for rational decision making in health care. To generate the most reliable information, clinical trials require certain design characteristics (particularly assignment of participants to interventions by ''randomization"), and they must include enough participants to exclude the play of chance as a likely explanation for results. Regardless of the sophistication and complexity of the design and analysis, the question of whether "a" is better than "b" is the essence of the clinical trial. The interventions that might be tested go beyond treatments (pharmaceutical, biologic, radiologic, surgical, or other procedures) to include preventive strategies,

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials diagnostic tests, screening procedures, devices, and other forms of medical advice or patient care. The definition of clinical trial used in this report encompasses some studies involving only a single intervention group. Such studies are included because they are often carried out before the definitive, comparative study to gather specific pieces of information about the intervention before the comparative study can proceed. To a great extent, these "single-arm," early phase clinical trials have been defined by the regulations governing approval of new drugs and are also a prominent feature of trials in cancer treatment (see Table 1-1). Their value lies in setting the stage for definitive, randomized trials. The committee believes that at least some patient care costs incurred in these trials should be eligible for Medicare reimbursement. TABLE 1-1. Phases of Clinical Trials (usually applied to drugs and devices) Phase 1 First studies in people, to evaluate chemical action, appropriate dosage, and safety. Usually enrolls small numbers of participants and typically has no comparison group. Phase 2 Provides preliminary information about how well the new drug works and generates more information about safety and benefit. Usually includes comparison group; patients may be assigned to groups by randomization. Phase 3 Compares intervention with the current standard or placebo to assess dosage effects, effectiveness, and safety. Almost always uses random allocation to assign treatment. Typically involves many people (hundreds or thousands) but may be smaller. Phase 4 "Post-marketing surveillance," evaluates long-term safety (and sometimes effectiveness) for a given indication, usually after approval for marketing has been granted by FDA. Brief History of Clinical Trials The formal evolution of the clinical trial dates from the eighteenth century, but the concept of comparing how well people fare after being "assigned" to different "interventions" (e.g., diets or medical treatments) has ancient historical origins. A considerable body of literature traces major developments along the way. (e.g., Bull, 1959; Lilienfeld, 1982; Meinert, 1986). The practice of randomization—randomly assigning study participants to either an experimental or a control group—was introduced by the statistician Ronald Fisher in horticultural research in 1926 and was described in his 1935 book (Fisher, 1926, 1935). Fisher asked the elemental question: How does one determine whether an observed difference in yield between fields is due to the

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials difference in the seed or fertilizer being tested, or due to differences in soil, temperature, moisture, and light? Fisher proposed dividing plots of land into narrow strips and assigning experimental treatments their place in the soil by a chance mechanism. He pointed out that "randomization relieves the experimenter from the anxiety of considering and estimating the magnitude of the innumerable causes by which his data may be disturbed" (Fisher, 1935). The British medical statistician Sir Austin Bradford Hill discussed the procedures of treatment allocation in 1937 (Doll, 1982), but time elapsed before medical researchers recognized that Fisher's ideas had applications beyond the bounds of farming research. Hill and the British Medical Research Council, in their multicenter trial of streptomycin inpatients with tuberculosis (Medical Research Council, 1948), are recognized as the first to use random sampling numbers to allocate patients to experimental and control groups. This trial also set standards for modern trials by defining, in advance, the characteristics of patients who would and would not be admitted to the trial; objectively documenting the response to treatment; and establishing a neutral committee to deliberate on the ethical concerns posed by the trial (e.g., whether it was ethical to withhold the drug from the control series, whether the physicians supervising the trial could modify the treatment schedule, and whether control patients should be given placebos that would permit the trial to be conducted in a double-blind manner). It is only in the past few decades that the randomized controlled trial has emerged as the preferred method—"the gold standard"—for evaluating medical interventions. But in that span, approximately a quarter of a million reports of "controlled trials" (though not all randomized) have been carried out (Cochrane Library Controlled Trials Register, 1999). Characteristics of Current-Day Clinical Trials Innovation is occurring continually in the design, conduct, and analysis of clinical trials, but most trials follow certain patterns and conventions. First (except for single-arm trials, discussed later in this chapter), they are comparative: two or more interventions are compared for efficacy or effectiveness.* The comparison is often between a new intervention and the current standard of care, which may be a completely different intervention, a placebo, no treatment, or the same intervention at a different dose or regimen or intensity. Trials also may *    Efficacy is defined as the extent to which an intervention produces a beneficial result under ideal conditions. Effectiveness is used to describe the extent to which a specific intervention, when used under ordinary circumstances, does what it is intended to do. Effectiveness, therefore, takes into account the fact that, in any group of individuals, some will not take the intervention as prescribed, or will take other actions that may compromise the effect of the intervention (e.g., take drugs that might interact with the test intervention). In this report, the recommendations apply to trials testing either efficacy or effectiveness.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials compare two or more "standard" interventions, all of which may be effective to some extent, to find out which works best. Participant Population and Sample Size It is generally accepted that the types and numbers of participants to be sought for a trial must be determined before the trial can begin. The study population should clearly be related to the condition under study, for example, a new treatment for Alzheimer's disease should be tested on Alzheimer's patients. But should the population be restricted further to include only a certain age range or exclude specific other diseases that patients might have? An area of ongoing debate concerns whether a trial is strengthened by patients being more homogeneous or more heterogeneous (see, e.g., Zelen, 1993). In trials of screening technologies, for example, the population may be of "average risk" or may be at higher-than-average risk due to some known characteristics (e.g., family history of the screened condition). The definition of participant characteristics directly affects how many participants need to be enrolled to answer the question addressed by the trial and how long the trial should be continued, which in turn affects the administrative structure needed to carry out the trial. The point of doing a randomized trial is to get a reliable answer, which requires avoiding undue influence of the play of chance, and this requires that sufficient numbers of "events" occur during the trial for chance to be ruled out as a likely explanation for the results. It is the number of "events"—that is, the number of participants who experience the outcome of interest during the course of the study—that drives the sample size and only indirectly the number of study participants. For instance, consider a hypothetical example of testing a yearlong intervention (e.g., an exercise program, or a drug) to prevent hip fractures. Enrolling people under age 40 would require far more people than would the same trial in people over age 70 because so few would be expected to experience hip fractures in the absence of the intervention. The trial of under 40s could also be done with the same numbers, but it would require several decades longer than the trial of over 70s. Real trade-offs must be made in reaching these decisions. In the example given here, perhaps the intervention has greater potential when begun at a young age, so the effect could be at least quantitatively (if not qualitatively) different than if only applied at an older age. The complexity and expense of carrying out the trial in the younger group might overwhelm any resources available (not to mention the fact that the technology might be obsolete by the time the decades-long study is complete), leaving researchers with little choice but to enroll older individuals.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials Randomization and Blinding It is widely accepted that participants must be assigned to intervention groups through randomization, a process that ensures an equal probability of getting any one of the treatments. Variations on the simplest form of the randomization process (a list of random numbers) have been developed to improve the balance of prognostic factors among intervention groups (e.g., stratification of participants before random assignment by potentially prognostic factors such as age, gender, or other medical conditions). However, the aim of the various methods is essentially the same. Another important aspect of trial design is "blinding" (or "masking")—keeping secret which intervention each patient is getting—which may be built in at several levels. The purpose of blinding is to avoid any bias—conscious or unconscious—in interpreting the effects of the interventions. In many cases, it is possible to keep the patient and the practitioner blind to which intervention the participant is receiving. This is common in pharmaceutical trials, in which dummy pills, injections, or other products can be manufactured so that all the treatments appear to be the same to doctor and participant. Blinding is often more difficult for procedures. Surgeons and others carrying out procedures must know what they are, but it may be possible to keep the patient unaware. Even when the practitioner knows which participants received which intervention, outcomes can often be evaluated by a third party without that knowledge. Baseline Characteristics and Outcomes In clinical trials, participants are generally followed from a well-defined point (e.g., diagnosis), which becomes time zero, or baseline, for the study. Usually, baseline information is recorded for each participant. This information may be as basic as age and gender, or it may include the results of diagnostic tests such as imaging, endoscopy, biopsy, cytology, or laboratory tests. The distribution of these characteristics in the different intervention groups is used as one measure of assurance that the groups are similar. The assumption is that, with a large enough number of people randomized, both the known and unknown factors that may affect outcome will be approximately equally distributed. The outcome measures may consist of laboratory test results, death or survival, a nonfatal clinical event, patients' symptoms or views, signs of disease, or quality of life. There may be various short-and long-term outcomes in a given trial, which are monitored in different ways. In a trial treating people just after a heart attack, short-term survival (e.g., one day, one week, one month) is the immediate goal, and that information can be recorded quickly in the hospital. Longer-term survival is also important, however, and tools exist in the United States (e.g., the National Death Index) and other countries through which the fact and cause of a person's death can be determined without directly contacting next-of-kin.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials Analysis The fundamental analysis of a controlled clinical trial is a comparison of the rates of important outcomes among the intervention groups. The difficult part is assuring that differences are due mainly to the different treatments, and not to chance or bias. The effect of chance is minimized by making the trial large enough, and the effect of bias is minimized by randomization, blinding, and other design features. A key concept in analysis is that the fundamental comparison is between the entire group randomized to one intervention and the entire group randomized to another intervention regardless of whether everyone in the groups actually got the intervention they were assigned. This may seem counterintuitive, but the "intention-to-treat" analysis is the only unbiased method for comparing the interventions. This makes it important to try to get as many people as possible to partake of their assigned intervention, and to maintain complete follow-up of participants. Trial Organization Clinical trials are simple conceptually, but they involve large numbers of people and require a significant infrastructure to carry out properly. In order to complete a study in a reasonable period of time, it is common to enroll participants in several different sites (in some cases hundreds) in the United States or around the world, in "multicenter" trials. One site acts as a coordinating center, usually controlling randomization of patients at all centers. That center or another may contain a central laboratory, receiving thousands of aliquots of blood or other biological samples to analyze from all centers. A critical component of a trial organization, particularly for larger randomized trials, is a "data safety and monitoring committee," which is independent of the trial investigators themselves. This committee's main responsibility is to the study participants. By conducting periodic reviews of interim data, they can determine whether, on one hand, any of the treatments have demonstrated a definitive benefit, or on the other, whether any of the interventions are clearly harmful. In either case, they have the authority to stop recruitment (and treatment, if it is appropriate) or modify the trial. Because of the complexity involved, there are no universal rules for deciding when a trial should be stopped. Protecting the Rights of Trial Participants Individuals joining clinical trials do not forfeit their individual rights to participate in their own health care decisions or to change their minds at any time, but they also may be limited in their ability to fully exercise these rights. Participants often are ill, which is why they are entering the trial, and are therefore

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials vulnerable in decisions relating to their health. In addition, they may not understand all the intricacies of the trial or their alternatives. For these reasons, considerable effort has been put into instituting mechanisms that seek to assure the rights of participants in clinical trials. The first line of protection comes from the trial protocol itself developed by the study investigators and approved by the data and safety monitoring committee (see Trial Organization, above). The protocol is a formal written document that describes the rationale for the trial, interventions, and other medical services that participants will get, numbers of participants needed, outcomes that will be measured, plan for analysis, and other derails of the trial organization. The protocol is used to develop a patient consent form that describes the protocol, the potential benefits and risks, and the patients' rights in the trial, in nontechnical language. The information on the consent form, as well as information supplied by study investigators or other health care providers involved in the trial, forms the basis of "informed consent," which must be given by individuals before they can formally enroll in a trial. The aim of informed consent is to ensure that participants understand the potential benefits and risks of participating, as well as their rights during the trial. The informed consent process and documentation follow detailed rules set out by the institutional review boards (IRBs) at each site where participants are enrolled. For trials funded by the federal government and trials involving medical interventions subject to federal regulation, these rules are dictated by the Department of Health and Human Services (DHHS) Regulations for the Protection of Human Subjects in the Code of Federal Regulations (45 C.F.R. §46). The Office for Protection from Research Risks (OPRR) in DHHS is the center for implementing the regulations and providing guidance on ethical issues in biomedical or behavioral research. All responsible host institutions require equivalent procedures, regardless of who is sponsoring the trial. IRBs are charged with protecting human volunteers in biomedical research. This involves not only ensuring informed consent, but also all aspects of the trial that bear on the welfare of the participants. Before a trial can begin, each IRB reviews the protocol, including such aspects as existing knowledge relative to the arms of study, the participant population and recruitment, potential risks and benefits to participants and society, investigator credentials, and monitoring requirements for the trial. During the trial, the IRB receives regular, periodic reports from the investigators, which it reviews and discusses. The IRB has authority to stop recruitment or take other actions necessary to protect participants. Interpreting Evidence from Clinical Trials It is rare that a single clinical trial answers a question definitively. Commonly, many trials with modest numbers of participants are carried out, asking the same or related questions, with identical, similar, or loosely related interventions being tested. The challenge for decision makers (at all levels in health care,

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials from central policy makers to the physician in the clinic) is to use all the reliable information available to determine the best intervention for a specific condition. Specific techniques can be used to express the results from multiple clinical trials in single numbers, variously called "meta-analyses" or statistical "overviews." In their most detailed form, they use the original data for each participant randomized in each trial; other techniques use only published results to estimate an average result. The increasing reliance on meta-analyses has emphasized the need to have access to all the evidence from all the trials of a particular question—not only the trials with published results, and not only those funded by certain sponsors. The reason is that trials that are published differ from those that are not. "Publication bias" occurs for various reasons: researchers may be less likely to seek publication for a trial that has negative results (e.g., a new treatment was no better than an old one), journals may be less likely to accept trials with negative results, or trials funded by the pharmaceutical industry that do not support a product may be withheld from publication. Follow-up of "cohorts of initiated trials" confirms that those trials with positive findings are more likely to be published than those with negative findings (Dickersin and Min, 1993). The impact on the evidence is clear: interventions that may, on balance, be ineffective or even harmful may be adopted because the published evidence is supportive while the negative evidence is unavailable. The idea of "prospective registration" of clinical trials has been proposed as a way to diminish the problems caused by publication bias. Researchers looking into specific areas could search trials registries to find out what trials have been done or are under way and whether or not their results have been published. Results of unpublished trials could be sought out, if desired. The topic of clinical trial registration is important to this report because one of the committee's recommendations requires linking Medicare claims for care received in the context of a clinical trial to a national clinical trials registry. An ongoing effort to establish such a registry is described at the end of this chapter. SPONSORSHIP OF CLINICAL TRIALS The federal government and the drug, biologic, and device industries sponsor most clinical trials in the United States. Private grant-making organizations and medical centers support small numbers of trials. Within the government, the National Institutes of Health (NIH) is the largest trial sponsor by far, but the Department of Veterans Affairs (VA) and the Department of Defense (DoD) also fund many. In addition, DHHS sponsors a few outside of NIH, through the Centers for Disease Control and Prevention (CDC) and the Agency for Health Care Policy and Research (AHCPR). Pharmaceutical and device companies mount clinical trials as a routine and necessary part of the process of securing federal approval for new products and adding approved indications to the labeling of existing products. Trials of new procedures (e.g., surgery or radiologic proce

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials dures) do not involve the approval of commercial products and are typically funded by NIH or a medical center where the trial is undertaken. Clinical trials take place in a variety of settings, including academic and other medical centers, such as comprehensive cancer centers. In attempts to expand enrollment in trials and involve more practitioners, NIH has extended clinical trial networks to include community hospitals and other communitybased providers in their trials. Trials sponsored by VA take place within VA facilities and associated medical centers. DoD-sponsored trials are mounted mostly at DoD facilities, and a small number take place in other settings. Trials sponsored by the government are usually managed by a team of academically affiliated investigators at a "coordinating center." Industry sponsors trials at the same medical centers that run governmentfunded trials and collaborates directly with VA and other government entities. Industry also recruits private clinics and individual practitioners to enroll patients in trials. Companies run some trials directly, but more often they retain one of the growing number of "contract research organizations." These organizations, like the coordinating centers mentioned above, manage the day-to-day activities of trials, including recruiting collaborators, analyzing data, and writing reports. Exactly how many clinical trials are under way in the United States is unknown, as is the number of people participating in them, but information is available for some categories. Some, but not all, NIH institutes have developed centralized lists of trials. Since the approval of an IRB is typically necessary to conduct trials, their existence is not confidential, but there is no easy way to find them all. Within the government, the Food and Drug Administration (FDA) has information about industry-funded trials for new drug and device applications, but FDA is bound to maintain the confidentiality of that information unless the sponsor chooses to make it public. NIH-Sponsored Trials Most NIH-sponsored trials are carried out under grants and contracts by researchers at universities, specialized treatment centers, and other medical settings. Some (particularly smaller phase 1 and 2) trials are conducted at the NIH Clinical Center in Bethesda, Maryland, by researchers employed by NIH. Only a few NIH institutes maintain registries or lists of the trials they are currently sponsoring. Lists of trials in cancer, AIDS, eye conditions, and rare diseases are available. Overall, NIH estimates that it sponsors about 7,000 clinical research studies at any one time—including both clinical trials and more basic studies—but NIH does not estimate how many of these are clinical trials (McCray, 1999). In 1999, NIH is sponsoring about 1,100 cancer treatment trials (about 400 are phase 3 randomized trials) and some smaller numbers of other trials (e.g., about 200 AIDS treatment trials and fewer than 100 trials in eye disease). NIH

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials has estimated that about 108,000 individuals enter into NIH-sponsored clinical treatment trials (this excludes trials of disease prevention) each year, with the largest number (an estimated 30,000) in cancer treatment trials (O'Rourke, 1999). Other Government-Sponsored Clinical Trials VA has a long-standing "Cooperative Studies Program," through which it sponsors multicenter clinical trials of particular importance to veterans. About 60 such studies, most carried out over about a 5-year period, are under way (Department of Veterans Affairs, 1999). In addition, VA researchers collaborate with academic medical centers, NIH researchers, and private industry in conducting trials at VA sites. The majority of participants in VA trials are veterans eligible for health care through VA. DoD sponsors some clinical trials for people in the uniformed services and their dependents. In recent years, Congress has allocated funds for special DoD research programs on breast cancer, prostate cancer, and ovarian cancer, and a small amount of these funds has gone toward funding clinical trials outside the DoD system. Industry-Sponsored Clinical Trials For the pharmaceutical, biotechnology, and device industries, clinical trials are part of the process of developing the necessary evidence of efficacy and safety for bringing new products to market. For the most part, the trials these companies conduct are prescribed by the laws and regulations governing approval of new products, under the regulatory authority of FDA. Companies may also conduct trials that involve only approved products, and some of these do not require notifying FDA. CLINICAL TRIAL PROTOCOL REVIEW AND MONITORING Clinical trials sponsored by the federal government undergo review before they are funded and allowed to proceed. The review consists of both examination of procedures by an IRB to ensure the rights and protections of participants, and some type of "peer" evaluation of the scientific design and technical aspects of the study (scientific and technical review also considers risks to human subjects). Although each sponsor develops scientific evaluation criteria independently, they are all similar. DHHS Regulations for the Protection of Human Subjects in the Code of Federal Regulations (45 CFR 46) establish standards for human subjects' protection in all clinical research funded by DHHS. OPRR implements the regulations and provides guidance on ethical issues in biomedical and behavioral research.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials FDA does not conduct clinical trials, but it has a major review function. Clinical trials using pharmaceuticals, biologics, or devices not yet approved by FDA always requires that research protocols be filed with FDA. In some (but not all) cases, protocols filed with FDA are also required when already-approved drugs are being used experimentally inpatient groups or in ways that are substantially different from those for which the drug was approved. These rules apply to all sponsors, government or private. FDA's authority derives from the Federal Food, Drug, and Cosmetic Act and is codified in Title 21 of the Code of Federal Regulations. All research undertaken under this authority also must adhere to the applicable federal regulations regarding the protection of human subjects. COSTS OF PATIENT CARE IN CLINICAL TRIALS Only evidence from pilot studies exists comparing the relative cost of patient care in clinical trials with the cost of treating similar patients in nonexperimental settings. These pilot studies were sponsored by the National Cancer Institute (NCI) in preparation for larger projects. Together, the studies compared the costs for about 260 patients in NCI-sponsored phase 2 and 3 cancer clinical trials with the costs of care for patients with similar diseases treated in the same health systems (Mayo Clinic, Group Health of Puget Sound, Kaiser Permanente of Northern California). Data from similar studies of patients in cancer trials have been presented at briefings and meetings, but written versions have not been available, and the methodology has not been described as thoroughly as it has for the three NCI-supported studies. Results from a larger NCI-funded study by the RAND Corporation should be available in a few years, but that study will also be limited to analyzing costs in NCI-funded cancer trials, excluding industry-funded trials. No studies of the costs of treating patients with other medical conditions in clinical trials are known to exist. The NCI-Funded Pilot Studies* All three studies relied on information from health system computerized databases, augmented by chart review. Efforts were made to match patients participating in clinical trials to patients receiving standard community care who were similar in all relevant respects and who would have been eligible to participate in the respective trial. *   This section is adapted from a summary prepared by Martin Brown (Applied Research Branch, National Cancer Institute), based on preliminary results of the studies presented at an NCI-sponsored meeting on July 7, 1998. The data should not be considered final until published in peer-reviewed journals.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials Mayo Clinic (Wagner et al., 1999) Cases were selected from local Minnesota residents who participated in cancer treatment trials at the Mayo Clinic from 1988 through 1994. From a pool of 176 candidates, it was possible to match 61 to similar nontrial patients, with 5 years of follow-up. Cost data were obtained from the Mayo Clinic Multi-Year Population-Based Data Warehouse of Standardized Medical Costs. Costs for trial patients were modestly higher (3.5%–13% after adjustment for censoring) over follow-up periods ranging from 1 to 5 years. Most of the additional cost for trial participants was incurred during the first few months after enrollment, and the observed cost differences decreased over time. Group Health Cooperative of Puget Sound (GHC) (Barlow et al., 1998) Patients in this study were members of GHC who enrolled in Southwest Oncology Group (one of the NCI Cooperative Groups) trials for breast and colorectal cancers from 1990 to 1996. Twenty trial participants with colorectal cancer and 49 with breast cancer were matched to nontrial patients of similar age, time of diagnosis, and initial stage of disease, using an automated matching procedure. Twenty-six of the trial participants with breast cancer were further matched to nontrial patients by medical record review (on co-morbidity and trial eligibility criteria). Cost data came from the GHC automated cost accounting system. Cumulative costs at 2 years of follow-up were essentially identical for trial and nontrial patients matched by computer. For the 26 pairs of breast cancer patients using closer matching, the mean cost was about $25,000 for nontrial patients and $30,000 for trial patients at 2 years from diagnosis (significant at p = .04). Kaiser Permanente of Northern California (KPNC) (Fireman et al., 1998) Patients were members of KPNC and participated in any of 10 breast and colorectal cancer trials from 1994 to 1996; 135 were matched to similar nontrial patients. Cost data came from the KPNC automated cost accounting system. The 1-year mean cost was 10 percent higher (about $1,500) for patients in trials. When the 11 patients enrolled in bone marrow transplant trials were excluded from the analysis, the 1-year mean cost for trial patients fell and was slightly lower than, but not statistically significantly different from, costs of patients receiving standard care.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials Conclusions Taken together, these three pilot studies indicate that there may be a modest excess in medical costs for patients enrolled in NCI-sponsored clinical trials, compared with similar patients not enrolled in trials. While these analyses constitute the best data currently available on this question, they were based on small numbers of patients in a few types of cancer treatment trials. A notable finding across the studies was the great variability in treatment costs for patients with the same diagnosis. For these and other reasons, generalizations to other treatment settings, populations, and diseases are not warranted. It remains to be seen whether costs for trial patients are higher, lower, or the same as those for patients outside of a trial, and whether the cost implications can be generalized to other trials. CLINICAL TRIALS REGISTRY There is no comprehensive listing of clinical trials in the United States. NIH maintained a clinical trials registry from 1974 to 1979, but now there are only separate NIH registries for trials involving cancer, AIDS, eye conditions, and rare diseases. There are also some small non-NIH registries. As a result, most clinical trials are not listed in any publicly accessible format. This situation is changing, however. There have been calls for a national clinical trials registry from both members of the public and health care providers who want to find out about trials for purposes of enrollment. Further pressure has come from researchers who review evidence from trials and need to know the trials that are ongoing, as well as those that have been completed. These forces led to legislation mandating a U.S. clinical trials registry for serious and life-threatening diseases. The mandate, which appears in Section 113 of the Food and Drug Administration Modernization Act of 1997, calls for the creation of a registry of clinical trials for drugs being carried out under Investigational New Drug (IND) Exemptions. With an additional legislative mandate and funding, this law could become the core of a comprehensive national registry of clinical trials. A national clinical trials registry is of particular interest to the committee because it would enhance the Health Care Financing Administration's (HCFA's) ability to track the effect of a change in reimbursement policy for clinical trials, as well as audit records to evaluate compliance with the reimbursement rules. The following sections describe the status of the registry as it is developing currently, and the ways in which it should be expanded to serve both as a tool for Medicare and for broader purposes.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials Registry of Clinical Trials of Drugs for Serious and Life-Threatening Diseases The Food and Drug Administration Modernization Act of 1997 (P.L. 105–115) provided for the establishment of an ''Information Program on Clinical Trials for Serious or Life-Threatening Diseases" (Section 113). The heart of the program (referred to as a "data bank") is a registry designed to make information about clinical trials of interventions in these diseases widely and easily available to all interested parties: individuals with serious or life-threatening conditions, physicians, researchers, and others. Both publicly funded and industry-funded trials are covered by the mandate, but there is a provision for sponsors to petition for a trial not to be included if they can provide evidence that registration would substantially hinder enrolling participants in the trial. The registry is being developed and will be maintained by the National Library of Medicine (NLM), under the general guidance of the Director of NIH. A decision about whether to include trials of devices in the registry has been deferred until a report, detailing the potential public health benefits and possible adverse impacts of including device trials, is submitted by DHHS to the Senate Committee on Labor and Human Resources and the House Committee on Commerce. The report is due later in 1999 (two years after passage of the legislation). Current Status The registry has so far been carried out as an internal NIH activity, with no formal input from outside NIH. NLM is now establishing the registry, focusing first on NIH-sponsored trials, which are scheduled for complete registration by the end of 1999. They will then take up industry-sponsored trials, as well as the remaining trials funded by public sources (e.g., VA, DoD, CDC), which NLM hopes to enter by the end of the year 2000 (McCray, 1999). However, procedures for identifying and registering industry-sponsored trials have not yet been developed. Each trial will be assigned a unique identifier. The record will contain the date the trial enters the registry and dates of subsequent modifications (e.g., if eligibility criteria change, or a drug dosage is changed, these changes would have to be reported to the registry), as well as core information items. The legislative mandate requires four pieces of information about each trial: description of the purpose of each experimental drug, eligibility criteria for participation in the clinical trials, location of trial sites, and point of contact for those wanting to enroll in the trial.

OCR for page 15
Extending Medicare Reimbursement in Clinical Trials An NIH working group has developed a somewhat longer (though still brief) list of data elements, but the list is not yet final. Some data items will be required and others optional.