Information Trading: How Information Influences the Health Policy Process (1997)

P. Pearl O'Rourke

Society is on the brink of a virtual explosion of genetic information. The discoveries are exciting not only for their scientific merit but also for the promises of improved understanding, prevention, control, and possibly, cure of human diseases. Dissecting the mysteries of the genetic code, however, raises numerous unintended legal, ethical, and social consequences that evoke issues of discrimination, confidentiality, and civil rights. The federal government has been given the formidable task of formulating policy regarding these consequences.

This paper discusses one of the first such issues brought to the U.S. Congress: health insurance discrimination based on genetic information. There is a sense of urgency on Capitol Hill to draft policy that will protect U.S. citizens from this potentially harmful consequence of genetic knowledge. This issue owes some of its prominence to the recent availability of a test for the breast cancer mutation (BRCA1) and the passage of the Kassebaum-Kennedy Health Insurance Reform Act of 1996 (S. 1028).

The BRCA1 test served as a wake-up call to the general public.

This highly publicized genetic diagnosis of a common disease forced the realization that everybody may be a test away from having a preexisting condition. No longer is genetic testing limited to rare diseases that affect small populations (Huntington's chorea) or diseases that affect specific racial or ethnic groups (sickle cell anemia and Tay Sachs disease). The Human Genome Project (HGP) at the National Institutes of Health (NIH) has nearly completed mapping the human genome, and almost every week another mutation linked to a type of cancer, heart disease, or other condition is identified. Investigators indicate that soon the DNA base pairs that determine an independent personality, or possibly an addiction, will be identified (Orkin and Motulsky, 1995). This new frontier of scientific knowledge has grown to a point where everyone is vulnerable.

Not only do people wonder what predictions could be made from the personal genome, but they also wonder who else will have access to the information. In this age of computers, information is a commodity. Vital statistics, be they financial, academic or health related, seem to be little more than datum points that can be entered into any number of banks. Is any of it confidential? Should it be? Who owns it? Who has the right to access it?

At the same time, the Kassebaum-Kennedy bill, which addressed the portability of insurance coverage and exclusion from coverage on the basis of preexisting conditions, focused attention on the reality of insurance discrimination. The fact that the Kassebaum-Kennedy bill specifically disallowed exclusion from coverage on the basis of genetic information highlighted the concept that genetic information may be equivalent to a preexisting condition. The possibility of stigmatization or discrimination is now a concern for the entire population.

Lobbying activity for the Kassebaum-Kennedy bill also provided numerous opportunities for the public airing of the ever powerful anecdote (Billings, 1993; Hudson et al., 1995).

• A woman who is pregnant with a child diagnosed with cystic fibrosis is told that if she carries her child to term, her insurance

company will not cover any medical care that is directly related to the diagnosis of cystic fibrosis.

• A woman who has a positive test for the breast cancer gene is asked to sign a waiver on her insurance policy relieving the company of any financial responsibility to cover breast cancer-related care.

Discrimination on the basis of genetic information is not new; as early as 1975, the National Academy of Sciences issued a report on genetic screening. Commissions, boards, committees, counsels, and studies (Hanna, 1995), have also excelled at identifying and massaging the problems, but the complexity of the issues has paralyzed society into relative inaction.

Now, however, Congress cannot avoid the issue; they have been requested to confront insurance discrimination on the basis of genetic information. To understand all of the issues, members of Congress need information, they need to be educated. Numerous information brokers have volunteered to provide the information and serve as teachers. Under the guise of education, these information brokers try to influence policy. The debate of genetic information and insurance discrimination has attracted "educators" from the insurance industry, the genetic research community, the biotechnology industry, and consumer groups. None of these groups are monolithic. Each has its own agenda.

One of the items on their agenda is how to define genetic information: should this be broad and inclusive or narrow and restrictive? A broad definition includes any and all patient data that suggest a genetic condition: specific laboratory tests for chromosomes, genes, or DNA; family history; physical examination; and any laboratory test that suggests an inherited disease. For example, genetic information could be a positive test for one of the breast cancer genes (BRCA1), an elevated blood cholesterol level, or a family history of cystic fibrosis. In contrast, a narrow definition is limited to a specific chromosome or gene, or to DNA or laboratory tests that have been demonstrated to be valid for the diagnosis of a genetic disorder (e.g., elevated sweat chloride levels for the diagnosis of cystic fibrosis). The use of a narrow versus a broad definition of

genetic information would have a significant impact on any potential policy concerning discrimination.

Physicians clearly use a broad definition of genetic information for evaluation and diagnosis. Various information brokers, however, have a vested interest in specifically promoting a broad or a narrow definition. Deciding a definition of genetic information will be an important aspect of any policy.

The private insurance industry's success depends on its ability to invest wisely and minimize financial risks. This means that it must be able to calculate the cost of coverage on the basis of the relative likelihood of the occurrence of an adverse event in an individual or a community (Hudson et al., 1995; Rothenberg, 1995; Rothstein, 1993).

Although the insurance industry represents a spectrum of insurers ranging from managed care organizations (MCOs) to the conventional insurance carrier who sells individual policies, virtually all insurers would like to participate in the process of determining policy on the use of genetic information. Few insurers, however, have volunteered any specific solutions. The topic of insurance discrimination seems to be a lose-lose proposition for the insurance industry.

MCOs as a form of community-rated health insurance will have to address genetic discrimination at two levels: entry into the plan and members' access to diagnostic and therapeutic options. Some detractors are concerned that MCOs will use genetic information to discriminate, alleging that MCOs already use marketing and enrollment practices designed to "cherry-pick" or preferentially select only low-risk health clients. Despite such allegations, at present few MCOs have specific guidelines and the MCO industry is not at the table as a unified group.

Kaiser Permanente is the exception. Its entrance policy does not allow underwriting or the use of genetic information for denying plan membership. In addition, it is developing clinical guidelines to determine which of its 6.9 million patients will have the option to

receive genetic testing (Bachman and Schoen, 1996; Christensen, 1996). At present, each genetic test is considered separately, taking into account disease-specific ramifications for future health as well as proven and appropriate prevention, therapy, or cure. For example, the BRCA1 test has serious limitations: there is a 5 to 14 percent false-positive rate, and once a mutation is identified, there are no proven methods by which to prevent the cancer from developing. Hence, BRCA1 testing will be viewed differently than a more sensitive and specific test for a disease that can be easily and successfully treated. Kaiser Permanente's approach to the use of genetic information is commendable, but it will be an onerous, constantly evolving process that may be difficult for smaller MCOs to duplicate. Many argue, however, that market forces will force all MCOs to develop similar approaches. The consumer's decision to join a plan will be affected by the plan's policy regarding the definition and use of genetic information. Is the average consumer sophisticated enough to adequately discriminate among the different policies?

For the traditional third-party insurance carrier, some form of underwriting is the usual approach for setting premiums. Premiums are based on events and characteristics in an individual's history that may affect future health—hence, the insurance physical and laboratory examination. Most see the logic and accept setting higher premiums for people who have high-risk behaviors, such as smoking and drinking. Acceptance dwindles, however, when it is proposed that premiums be set on high-risk genetic characteristics that are totally outside a person's control. Although society has an embarrassing history of accepting discrimination on the basis of sex and race, genetic discrimination seems to have reawakened a societal concern about access to health care coverage.

What is the insurance industry's position on genetic information? A 1993 study done by the Office of Technology Assessment (OTA) reported that 75 percent of chief executive officers in the insurance industry felt that they should have the autonomy and authority to determine how to use genetic information (Office of Technology Assessment, U.S. Congress, 1992). No specific policies have been produced, however. The fact that the 1996

Kassebaum-Kennedy bill forbade the use of genetic information as a preexisting condition required the insurance industry to precipitously enter the discussion. Rather than being able to proactively introduce its policies, it was forced into a position of defense and reaction.

The insurance industry would like a very narrow or restrictive definition of genetic information. It is sensitive to the creep between DNA testing, family histories, and potentially, the routine diagnosis of a disease that has genetic implications. It would like to see genetic information limited to that derived from specific laboratory tests. Members of the industry nervously watch as various bills wend their way through the U.S. Senate and the U.S. House of Representatives, each one with seemingly more inclusive, broad definitions of genetic information. Insurers are fearful that if a broad definition of genetic information is accepted and they are not able to consider this information in determining premiums and coverage, they will be driven out of business.

One effective way that insurers have presented their concerns is by providing examples of what would happen to insurance premiums if companies were not allowed to consider any genetic data. Imagine the impact if people obtain health insurance only after they have received a genetic diagnosis that portends a significant and expensive disease. If the genetic information is confidential or cannot be considered a preexisting condition, the insurance industry's ability to accurately assess risk is destroyed and it will either lose money or have to charge everyone higher premiums.

Insurance representatives are also lobbying for uniform policies between the states. Of note, 65 percent of insurance policies fall under the Employee Retirement Income Security Act (ERISA) (Rothstein, 1993). A level national playing field is needed. To this end, insurance carriers are supplying the staffs of members of Congress with the ongoing legislative debates in each state: they emphasize the difficulties that will result from huge interstate disparities. The difficulties will be for the consumers as well as the insurance carriers.

Unfortunately, an attempt is being made to retrofit this new genetic information into the existing paradigm of insurance cover-

age. The insurance company is trying to perform damage control by carefully limiting and classifying genetic information. This type of solution, however, will not likely be the final one. Once the majority of people can access the information that may predict their medical future, almost everyone may have a preexisting condition that will not be covered or covered at an extraordinary cost. Can the present system of health insurance coverage meet these needs? Will the explosion of genetic information break the existing system of risk assessment and force an alternative method of providing coverage for health care?

The biomedical genetic research community includes basic laboratory researchers in academic and nonprofit institutions. Members of this community include HGP, the College of American Pathologists (CAP), the American Society for Human Genetics, the American College of Medical Genetics, and others (College of American Pathologists, 1996; Greely, 1995; Orkin and Motulsky, 1995). Commercial enterprises that are involved not only in research but also in the development of commercial products are categorized as the biotechnology industry.

Although the members of the research community are not monolithic in their opinions, they appreciate that the American public, which although poorly versed in the science of genetics, is fearful of the potential Big Brother mentality that could result from cataloguing genetic information. The average American citizen has been exposed to enough criminal litigation on television and in the press to recognize the concept of DNA fingerprints that can link evidence to a specific person. Such genetic paranoia could hinder or even terminate support for further research.

In response to this paranoia, the entire research community is committed to the delivery of open, truthful information to the public and, specifically, to decision makers. For many years, in an effort to build support for funding, genetic researchers marketed the promise that by unlocking genetic mysteries, diseases would be better understood, new cures would be found, and all people would live

longer and healthier lives. Unfortunately, this marketing strategy created false expectations (Orkin and Motulsky, 1995). The public equated the identification of mutations with the availability of a cure. In fact, there were parents of children with known genetic diseases who decided to forego further genetic counseling or intrauterine diagnosis because ''by the time my next child is born, there will be a cure." Now these same researchers are engaged in corrective advertising. They are carefully trying to educate and reeducate the public. It is clear that genetic information is becoming a household concept, and it is imperative that the research community keep the public on its side. To this end, it is piloting major public awareness campaigns. A large part of this is directed at Capitol Hill. The authorizers and the appropriators must be convinced that the information obtained by genetic research is vital and that the scientists can handle this information with appropriate caution and responsibility.

As a result, it is quite easy to attend a genetics lecture being given on Capitol Hill by none other than the director of HGP. The lectures cover basic science presented in lay language, clinical applications, promises for the future, and comments on the ethical, legal, and social implications of this material. The goal is to portray the socially responsible scientist, the available scientist. Active dialogue is a priority. The briefings have been complete and unbiased, but for audience members who had little or no prior knowledge of genetics, the briefings are more memorable for their openness and sincerity than for their specific content.

The research community is also cognizant that the information being unraveled introduces new social, ethical, and legal responsibilities and challenges (Orkin and Motulsky, 1995). In response to these concerns, HGP created the Ethical, Legal and Social Implications Program (ELSI). The charge of ELSI includes supporting and reviewing research and setting up committees or councils to address specific legal, social, or ethical issues. However, ELSI was never given the authority to set policy. Critics of this program felt that ELSI was an impotent creation and referred to it as an "unavoidable political tax," formed to appease the public and the politicians with an affirmation of social conscience (Orkin and Motulsky, 1995).

Even though ELSI does not have the mandate to form policy, the fact that there is no other policy-forming body passively empowers ELSI. People outside of HGP state that ELSI is inappropriately seen as the single "voice of the genetic research community." No single voice exists, and independent researchers as well as biotechnology industries are frustrated by this unintended power.

The lack of an authorized voice has resulted in disparate views and approaches. Several formal reports on the use of genetic information determining insurance coverage have been published. In 1993, the ELSI Task Force on Genetic Information and Insurance reported the following principles (ELSI Task Force on Genetic Information and Insurance, 1993):

1. Information about past, present, and future health, including genetic information, should not be used to deny health coverage to anyone.

2. There should be universal access to basic health services appropriate for the healthy to the seriously ill.

3. Basic health services should treat genetic diseases comparably to nongenetic diseases with appropriate testing, counseling and treatment.

4. The cost of health care should not be affected by information, including genetic information, about the past, present, or future of a person's health.

5. Access to health care should not depend on employment.

6. Access to health care should not require access to information, including genetic information, about a person's past, present, or future health care.

7. Pending universal access to basic health services, there should be alternative means of reducing the risk of genetic discrimination. Health insurers should consider a moratorium on the use of genetic tests in underwriting and insurers should undertake vigorous education efforts.

This 1993 task force report was largely ignored. The principles were simply too encompassing and the report suggested no practical solution. At about the same time, the Institute of Medicine pub-

lished a report entitled Assessing Genetic Risk (Institute of Medicine, 1994). That report called for the following:

1. Keeping genetic testing voluntary, mandating informed consent, and keeping strict confidentiality.

2. Genetic tests must have a near zero chance for mistaken results.

3. Counseling and individual education must be available.

4. Genetic testing is human investigation until benefits and risks have been well identified and assessed.

5. Publicly supported population-based screening programs can only be justified when the disorder is of significant severity and frequency and when there are available interventions.

Then, in 1995, a workshop sponsored by ELSI and the National Action Plan on Breast Cancer proposed the following recommendations. Insurance providers should be prohibited from (Hudson et al., 1995):

1. using genetic information or the request for such to determine enrollment or coverage,

2. establishing differential rates based on genetic information or the request for genetic information,

3. requesting or requiring collection or disclosure of genetic information, and

4. releasing genetic information without written authorization for each disclosure including to whom the disclosure is being made. This prohibition includes any holder of genetic information.

All of these reports favor strong protection of the individual and any genetic information. They each recognize the implications of genetic information on access and costs of health care and they each propose very cautious approaches to the expanding use of genetic information. Part of that caution is to limit genetic testing to experimental status until more is understood about the specificity and sensitivity of specific tests as well as knowing what to do with the information. Both the Institute of Medicine and ELSI support the

idea of having centralized regulation of genetic tests; this would most likely be under the auspices of the U.S. Food and Drug Administration (FDA).

In short, the information brokers that represent the full-time research community are very interested in a broad definition of genetic information. The information that they bring to the Congress is complex basic science translated for consumption by lay-people. They are very conversant in the unintended consequences of genetic information and want to proactively address these issues and be a part of the problem-solving process.

The biotechnology industry is an active participant in the genetic research community. Its distinguishing characteristic, however, is the fact that the majority of its research is directly linked to the development of a commercial product. The biotechnology industry must consider the consequences of commercialization and the need to realize a profit: this makes it difficult for members of the industry to embrace all of the propositions stated by HGP. These companies want to be able to bring their products to the marketplace as soon as it is safely possible. Any suggestion from HGP that genetic testing is experimental until benefits and risks have been well identified and assessed threatens the biotechnology industry's ability to commercialize genetic tests (Benowitz, 1996). The biotechnology industry is understandably troubled by HGP's monolithic stand on how the "research community" feels. As a result, the biotechnology industry is in the difficult position of lobbying as a research community that disagrees with HGP.

One particular area of concern is the regulation of genetic tests, possibly by FDA. This concept runs against the expressed interests of not only private industry but many of the academic research centers as well; this request for regulation is seen as HGP's attempt to destroy the competition (Benowitz, 1996). The biotechnology industry feels that such regulation would hurt the advancement of this science. It states that adequate regulation can be maintained by market forces as well as by existing requirements of the Clinical

Laboratories Improvement Act (CLIA). Further regulation is seen as unnecessary and counterproductive. CAP agrees with the biotechnology industry on this issue. CAP presently accredits laboratories under the auspices of CLIA and reports its findings to the Health Care Financing Administration (HCFA) (College of American Pathologists, 1996). CAP feels strongly that CLIA requirements are adequate:

Biotechnology firms are convinced that the public will demand and should be able to obtain genetic testing as it becomes available. Market forces should be adequate to protect the quality of the available tests; federal regulation is not needed. Many have bet their financial futures on the belief that genetic research will increase the demands for genetic tests, the need for accurate tests, and the desire for genetically manipulated approaches to treatment. As a result, many of these firms are purchasing the rights to genetic processes, and they are patenting information. For example, Millennium Pharmaceuticals, Inc., in Cambridge, Massachusetts, already has $180 million in research and license agreements with three pharmaceutical firms (Johannes, 1995).

All of this information is relevant to the issue of insurance discrimination. If discrimination based on genetic information is allowed, there are a number of ways in which the biotechnology industry could be affected. Insurance companies could demand only FDA-approved laboratory tests. The insurance industry could demand tests that had been verified in clinical studies: this would relegate genetic tests to "experimental" status before these studies were completed. The term "experimental" means that the biotechnology industry would be able to charge only enough to recoup costs—a profit could not be realized. Perhaps the biggest effect

would be the fact that individuals would forego genetic testing because of the fear of losing insurance coverage.

Many biotechnology firms have created their own ethics departments to better understand the evolving problems and to position themselves to be players in the resolution of these problems. OncorMed, Inc., a genetic biotechnology firm in Maryland, is providing genetic tests to institutions, but only under research protocols that are approved by an institutional review board (IRB). Because there are so many potential consumers who are not in academic centers equipped with an IRB, OncorMed has established its own IRB to review patient protocols (Hubbard and Lewontin, 1996; Murphy, 1996). By the end of 1996, OncorMed will begin marketing tests for BRCA1 and BRCA2 (another breast cancer mutation) and colon cancer directly to physicians—these physicians may or may not be trained in genetic counseling. Their protocols will pass through OncorMed's own IRB. Myriad Genetics in Salt Lake City, Utah, will simultaneously begin marketing clinical tests for BRCA1. Marketing of these tests will begin at a time when HGP has stated that testing for BRCA1 should only be done in the confines of a research project because it is not yet known what to do with the results.

The biotechnology industry is an important player in this issue. It has been packaging its message in a number of ways. It is committed to preventing any form of discrimination on the basis of the results of genetic tests, but its message is more complex because of its desire to limit oversight and regulation of its products. Because federal regulation of these tests is a concern, the industry has been trying to demonstrate the inability of FDA to effectively approve genetic tests in a timely manner. In addition, the biotechnology industry focuses on the right of American citizens to get genetic testing if they so desire because requiring regulation and clinical testing is allowing the federal government to be too paternalistic. The industry believes that since the information is available, let individuals access it if they so desire. The present climate of antiregulation supports the concept that the government has no business regulating these products.

The biotechnology industry can effectively market itself to the

disease-phobic American public simply by advertising available tests. The industry suggests that the genetic test should be seen simply as any other laboratory test; it should not be elevated to something that it is not. It supports the autonomy of the American public as well as privacy and confidentiality. In the absence of specific legislation limiting the ability to market tests, the industry is proceeding. Today in Fairfax, Virginia, Dr. Joseph Shulman will test human blood for the presence of the BRCA1 mutation. The information belongs to the patient and to no one else. Dr. Shulman feels certain that the information is the right of the individual. But what happens when the test is positive? The false-positive rate is between 5 and 14 percent. Even if it is positive, it means that the woman has an 85 percent chance of developing breast cancer in her lifetime. Is there any way to prevent it? That is unknown. Despite the lack of information, some people are having bilateral mastectomies and oophorectomies prophylactically. Aside from the question of whether or not this is good medicine, what is the role of the insurance company? Must insurance companies be financially responsible for medical procedures that were selected on the basis of a test that the insurance company does not recognize?

A large number of consumer groups are also lobbying Congress (American Society of Clinical Oncology, 1996; Cook-Deegan, 1994; Garber and Schrag, 1996; Lerman et al., 1996). Each group, vested to its particular problem, explains how discrimination affects it now and how the expansion of genetic information threatens to make things even worse. The groups present themselves as potential victims. Virtually all of these groups express a paranoia about the potential misuses of genetic data: for example, eugenics, selective abortions, and selective sterilization. The common mantra is, "Go slowly and be careful." They are usually satisfied with ELSI's approach to the issues and appreciate the fact that HGP has focused on insurance discrimination as an important issue. The consumer groups have a shared pessimism, however, that discrimination will

continue and that somehow, in some way, the insurance industry will still make the public pay.

Groups that have been victims of discrimination are concerned that genetic information will allow discriminators to better hone their actions and find new excuses for discrimination. Racial minorities have particular concern. They remember the overt insurance and employment discrimination that came with the diagnosis of the sickle cell trait in the 1970s. Sickle cell trait means that the person is a carrier of the abnormal gene, but that the person does not have the disease. Despite the fact that these people were healthy, they were denied insurance coverage and, in some cases, employment (Voelker, 1993). This unfortunately followed an already long history of racial discrimination. Although many advocacy groups for minorities plead for antidiscrimination in insurance coverage, their concerns go far beyond health insurance.

The disability advocacy groups are openly concerned about genetic cleansing. They present emotional testimony that as soon as some of these disabilities are linked to an identifiable genetic mutation, the bias will be to abort. Although they understand that some families may decide that a particular genetic burden is unfair to impose on a child, there will also be pressures of social responsibility to prevent or at least minimize the "burdens" on society. For example, consider cystic fibrosis (CF), a multisystem disease with primarily pulmonary and gastrointestinal manifestations. People with CF rarely survive beyond their 20s or early 30s. The clinical course is one of steady decline, with expensive daily medication and increasingly more frequent hospitalizations and possibly lung transplantation. The concern of disability advocacy groups is that society will say that CF is too expensive. People who carry this gene should not be allowed to conceive and bring such a child into this life to suffer and to drain society of its resources—resources that could better serve healthy children with normal projected life spans. The pressures may come from insurance, employment, and educational discrimination. These concerns are real; remember the case of the patient with CF described earlier in this paper. The disability and minority advocacy groups share the same concerns. Discrimination in providing health insurance is only a small part of

the bigger picture, and this is an important battle. This is a battle of civil rights.

The disability community takes little comfort in the Americans with Disabilities Act. Title V of that act disallows discrimination only if the process of stigmatizing and selection was done with "subterfuge" in mind. This language creates a tremendous legal opportunity for health insurance discrimination (Orkin and Motulsky, 1995).

Unfortunately, few of these groups have formed any coalitions. The suggested solutions and approaches are often vague, are poorly articulated, and are tailored to primarily meet the needs of each interest group. Although most have not even considered the differences between a broad versus a narrow definition of genetic information, their presentations assume a broad definition.

Well-seasoned breast cancer advocacy groups such as the Breast Cancer Coalition have produced elegant policy statements. Its early entry into this discussion resulted from the availability and commercialization of the test for BRCA1. This group demands confidentiality of genetic information in its broad definition. It should be up to the woman to decide what to tell insurers, and insurers should not be able to change premiums on the basis of this information (Breast Cancer Coalition, 1996). Groups such as this that are well positioned in terms of power brokerage are the groups most likely to shape legislation.

Congress has already taken the challenge of how to deal with genetic information in the context of discrimination. The number of briefings and hearings on HGP, genetic information, and potential discrimination attest to their commitment. Four bills in the Senate during the 104^th Congress specifically addressed genetic information and discrimination: S. 1416, The Genetic Privacy and Nondiscrimination Act of 1995; S. 1600, The Genetic Fairness Act of 1996; S.1694, The Genetic Information Nondiscrimination in Health Insurance Act of 1996; and S. 1898, The Genetic Confidentiality and Nondiscrimination Act of 1996. The Kassebaum-Kennedy Health

Insurance Reform Act of 1996 was a fifth. The House had at least four bills: H.R. 3130, Health Coverage Availability and Affordability Act of 1996; H.R. 2690, The Genetic Privacy and Nondiscrimination Act of 1995; H.R. 2748, The Genetic Information Nondiscrimination in Health Insurance Act of 1995; and H.R. 3482, The Medical Privacy in the Age of New Technologies Act of 1996. All of these addressed similar issues, but they differed in their definition of genetic information and the way that insurance companies can handle the information. However, they all shared the goal of protecting and maintaining confidentiality.

As these bills began to emerge, a number of hearings were scheduled. Most of the witnesses provided further descriptions of the problems of discrimination and underscored the need for the federal government to provide legislation to prohibit discrimination. At present a number of states have individual laws regarding genetic information. None are identical. It is a patchwork quilt of genetic information legislation. The expansion of genetic testing, however, does not respect state borders. There must be a lead at the federal level, but there must be the certainty that the federal action does not preempt reasonable state laws.

Karen Rothenberg posed the following at a recent hearing in the Senate:

Although these are desirable goals, the field of human genetics may already be too far down the road. Testing is already available for the person who is willing to pay. Genetic testing is already expanding beyond research settings in the absence of protective legislation. As genetic information continues to explode there will be numerous attempts to prohibit discrimination. However, these may be only small categorical solutions that fail to address the entire issue.

The increasing availability of genetic information will also become a factor in the health care cost debate. Genetic tests cost money. Often they must be repeated or validated by other tests. Results from genetic tests must be put into the context of a family pedigree; this requires a person trained in clinical genetics. A genetic counselor should be available to help the individual understand and digest the information. In addition, genetic diagnoses may result in new and more expensive therapies. As an extreme example: what if bone marrow transplantation became the proven therapy for persons with the genetic predisposition to a specific cancer? Bone marrow transplantation is very expensive, and in the present paradigm of financing health care, this would translate directly into an increase in health insurance premiums.

How will this play out? A most fascinating question. Although the ELSI Task Force on Insurance and Genetic Information was largely ignored, its prognostication may have been correct. Universal coverage and access to basic health care may be another one of the mysteries unlocked by the genetic revolution.

It is important to understand how members of Congress have and continue to obtain information and advice on genetic information. Each member of Congress has a limited amount of time to invest in this technical, complex, and rapidly evolving field of science. As a result, members must rely on consultants not only to educate but also to advise. Because there are so many brokers of genetic information, members of Congress have numerous choices. Most have accessed persons or organizations that they have successfully used in the past. The result is that most members hear the information distilled by agents who know the bias of their office: that is, a member may receive all of his or her education on genetic information by the insurance trade group.

It is important to augment and fill out the potentially focused education that members may arrange for themselves. Briefings and

hearings are one way to present the full story, but it is equally important for specific interest groups to visit individual offices to offer their own synopsis of the information.

Anecdotes are personal, graphic, and often emotional; as such, they grab attention and demand action more quickly than does more comprehensive and broadly representative information. Although one anecdote provides interest, two anecdotes become fact. The topic of genetic information and insurance discrimination is laden with anecdotes. Although anecdotes provide illustrative examples, caution must be given whenever the lesson from an anecdote is generalized and accepted as fact.

The complexity of this topic makes any potential simplification attractive. For example, rather than providing a solution for the use of all genetic information, perhaps a carve-out solution for the use of breast cancer-related genetic information would be an answer. After all, the availability of the test for the BRCA1 mutation makes this a timely issue, the breast cancer coalition is very strong, and the women's vote is important for reelection. Similar arguments can be made for a number of specific, limited concerns, but such approaches may ultimately be destructive. A comprehensive approach is needed, not an approach that will address breast cancer mutations alone.

This is a particularly difficult problem. Unfortunately, members' enthusiasm for a cause is too often motivated by a personal interest such as a family member or a powerful group of constituents with a specific disease. It will be exceedingly important for various advocacy groups to form coalitions that can shape policy for the larger population. If every interest group tries to splinter the policy in a self-serving way, there may never be any reasonable progress.

Discussions over concerns about insurance discrimination are presently the most public arena for the discussion of genetic information. Many related concerns have been catapulted into the same arena for the purpose of gaining public awareness. One example of this is the question of regulation of genetic tests and genetic test material. While this is an extremely important issue, it is presently further complicating and confusing the discussion of genetic information and discrimination.

Members of Congress must take time to stand back and define the questions. Careful identification of the questions may prevent contamination by related but separate concerns, and it may prevent simplification of an issue that simply cannot be simplified.

American Society of Clinical Oncology. 1996. Statement of the American Society of Clinical Oncology: Genetic testing for cancer susceptibility. Journal of Clinical Oncology 14:1730-1736.

Bachman, R.P., and E.J. Schoen. 1996. Managed genetic care in a large HMO. HMO Practice 10:54-58.

Benowitz, S. 1996. Scientists struggling with concerns raised by genome project progress. The Scientist 10:1-75.

Billings, P.R. 1993. Genetic discrimination. Healthcare Forum Journal. October: 35-37.

Breast Cancer Coalition. 1996. White Paper on Genetic Information. Philadelphia, Pa.: Breast Cancer Coalition.

Christensen, K. 1996. Advances In Genetic Research And Technologies: Challenges for Public Policy. Testimony of the Kaiser Permanente medical care program before the U.S. Senate Committee on Labor and Human Resources. Washington, D.C., July 25.

College of American Pathologists. 1996. Response to Federal Drug Administration's March 14, 1996 Federal Register proposal to regulate analyte reagents as Class I medical devices exempt from the premarket notification (510k) process (Docket No. 96N-0082). June 9.

Cook-Deegan, R. 1994. Private parts. The Sciences 34:18-23.

ELSI Task Force on Genetic Information and Insurance. 1993 Report. Bethesda, Md.: Ethical, Legal and Social Implications Program.

Garber, J.E., and D. Schrag. 1996. Testing for inherited cancer susceptibility. Journal of the American Medical Association 275:1928-1929.

Genzyme Corporation. 1996. A proposal for regulating predictive genetic testing. Fact Sheet. July. Cambridge, Mass.: Genzyme Corporation.

Greely, H.T. 1995. Conflicts in the biotechnology industry. Journal of Law, Medicine and Ethics 23:354-359.

Hanna, K.E. 1995. The ethical, legal and social implications program of the National Center for Human Genome Research: a missed opportunity? Pp. 432-454inR.E. Bulger, E.M. Bobby, and H.V. Fineberg, eds., Society's Choices: Social and Ethical Decision Making in Biomedicine. Washington, D.C.: National Academy Press.

Hubbard, R., and R.C. Lewontin. 1996. Pitfalls of genetic testing. New England Journal of Medicine 334:1192-1194.

Hudson, K.L., K.H. Rothenberg, L.B. Andrews, M.J. Ellis Kahn, and F.S. Collins. 1995. Genetic discrimination and health insurance: An urgent need for reform. Science 270:391-393.

Institute of Medicine. 1994. Assessing Genetic Risks: Implications for Health and Social Policy. Washington, D.C.: National Academy Press.

Johannes, L. 1995. Detailed map of the genome is now ready. The Wall Street Journal, pp. B 1, B11, Dec. 22.

Lerman, C., S. Narod, K. Schulman, C. Hughes, A. Gomez-Caminero, G. Bonney, K. Gold, B. Trock, D. Main, J. Lynch, C. Fulmore, C. Snyder, S. Lemon, T. Conway, P. Tonin, G. Lenoir, and H. Lynch. 1996. BRCA1 testing in families with hereditary breast-ovarian cancer. Journal of the American Medical Association 275:1885-1892.

Murphy, P.D. 1996. Testimony for OncorMed Inc. before the Senate Labor and Human Resources Committee hearing on Advances in Genetics Research and Technologies: Challenges For Public Policy. Washington, D.C., July 25.

Office of Technology Assessment, U.S. Congress. 1992. Genetic Tests and Health Insurance. Results of a Survey. Publication No. OTA-BP-BA-98. Washington, D.C.: Government Printing Office.

Orkin, S.H., and A.G. Motulsky. 1995. Report and recommendations of the panel to assess the NIH investment in research on gene therapy. December 7. Bethesda, Md.: National Institutes of Health.

Rothenberg, K.H. 1995. Genetic information and health insurance: State legislative approaches. Journal of Law, Medicine and Ethics 23:312-315.

Rothenberg, K.H. 1996. Testimony before the U.S. Senate Committee on Labor and Human Resources hearing on Advances in Genetics Research and Technologies: Challenges for Public Policy , Washington, D.C., July 25.

Rothstein, M.A. 1993. Genetics, insurance and the ethics of genetic counseling. Pp. 159-177inMolecular Genetic Medicine, vol. 3. New York: Academic Press.

Voelker, R. 1993. The genetic revolution: Despite perfection of elegant techniques, ethical answers still elusive. Journal of the American Medical Association 270:2273-2274.