The relative contributions of genes and environmental factors will vary significantly for different diseases, as will the needs and methods of surveillance. Surveillance for diseases and disorders where the presence of the genetic trait equals the presence of the disease, such as Huntington's Disease and Tay-Sachs, may be useful for determining the prevalence and distribution of the gene and the disease, to target services, and evaluate programs. For other conditions, such as Phenylketonuria, the genetic trait plus a particular environmental factor creates the disease or condition. Some genetic information identifies individuals as carriers who may have a greater or lesser risk of passing on various diseases and conditions to their offspring.

The diseases and conditions of greatest concern from a public health standpoint will probably be found to result from a more complex interaction between single or multiple mutations and factors in the social and physical environment. ¹ Epidemiol Rev 1997;19:175-180. For many common chronic diseases, such as heart disease and cancer, the genetic factor may be one of several modifiable or non-modifiable individual and environmental risk factors. A person's genetic composition is also thought to be a factor in the susceptibility to many infectious diseases including HIV/AIDS. Therefore, the predictive value of the genetic information will vary substantially. Surveillance activities for the particular diseases could include relevant genetic information in addition to other risk factors. In many of these cases, the presence or absence of the genetic trait may be of limited value on its own, but significant in the context of other factors, and for subsequent epidemiologic and clinical investigations.

Surveillance for genetic information is already occurring. Birth defects surveillance systems and registries are an important source of genetic information on a national, state, and local level.^{2 3} These systems vary in that some rely on available information, while others are the product of active case-finding.² Newborn screening programs collect data that can be used for valuable surveillance on the prevalence and distribution of the underlying disease conditions tested, to assess the effectiveness and need for particular screening tests in various populations, and to evaluate prevention and intervention activities. Many states have established programs to provide otherwise unavailable or inaccessible genetic services to selected populations. Data derived from the provision of such services may be critical to assess the prevalence of various conditions in the state, to evaluate whether the services offered are appropriate and effective, and in planning future health services and prevention programs.

Defining the Social Risks of Surveillance

Throughout the history of American public health, surveillance has generated occasional opposition, based on fears that the collection of data by the government threatened the civil rights, social status or economic interests of individuals. Opposition to surveillance has in some instances slowed its implementation, reduced cooperation by both reporters and subjects, and led to those at risk avoiding diagnosis and other medical services. Leading examples include tuberculosis reporting,Bull Hist Med 1975;49:169-94. 1940. J Law, Med & Ethics 1993;21:289-95. venereal disease surveillance, and named reporting of HIV disease.United States. Albany L Rev 1998;61:679-743. 1998;61:831-95. Opposition to surveillance typically reflects the deep anxieties and social divisions revealed by a specific health condition, rather than a categorical enmity towards public health data collection: people who object to HIV surveillance do not typically oppose surveillance for measles or malaria.1994;27:251-72. In such cases, successful surveillance depends upon the effectiveness of public health agencies in broadly addressing the factors engendering opposition.

Defining social risk
Part of the difficulty in assessing the impact of social risk on health-related decisions has been the lack of a generally accepted definition of the phenomenon of social risk itself. One of us has offered this definition: social risk in health behavior is the danger that an individual will be socially or economically penalized should she become identified with an expensive, disfavored or feared medical condition.¹⁸ It has two distinct components: (1) "the threat," i.e. attitudes and behavior that cause or threaten social harm; and (2) "the perception of risk," i.e. the attitudes and beliefs about the threat among those who are in some way tied to the trait or disease. Social risk will influence acceptance of surveillance whenever the social construction or economic cost of a disease creates at least the perception of a risk that, alone or in combination with other factors, outweighs the benefits of obtaining diagnostic or therapeutic care.

The threat and the perception of social risk are independent to a considerable degree. This independence has been a notable characteristic of the social risk of public health surveillance, and may be the most important characteristic of social risk for policy makers to recognize. It explains why, as in the case of HIV, public health's essentially spotless record of protecting surveillance data, and strong privacy legislation, can fail to still opposition to named reporting. It suggests that the current lack of strong legal protection against the social risks of genetic testing will not necessarily lead to wide-spread opposition to genetic surveillance.

The potential social risks of genetic surveillance
The concern that progress in genetics could create social risk is founded on historical experience with other health conditions. For the most part, scholarly treatment of the issue focuses on ethical and legal principles, stimulated by historical analogies and contemporary anecdotes. There have been only a small number of empirical investigations of the prevalence of threats or perceptions of social risk. These studies almost uniformly rely on self-report of the experiences of small, convenience samples. Several focus on attitudes or behavior under hypothetical circumstances. Many studies that touch on issues of social risk do so only tangentially in the course of studying a broader range of factors influencing health care utilization.

These studies, while useful, can hardly be considered a sufficient empirical basis for the construction of effective policies addressing the social risks of genetics, and some commentators suggest that the risk of genetic discrimination may be overstated. Available data do not allow reliable estimation of the prevalence of discrimination, stigma, social hostility or other manifestations of social risk. As in other areas of prevention, the important and complex question of how perceptions of risk arise and influence individual behavior has been largely ignored. The psychological process of risk assessment has been well-studied, but the many insights of this body of work have been slow to permeate the analysis of health behavior. Further research on both the threats of social risk and their perceptions should be a high priority.

The experience with HIV has shown the dangers to policy-making of a lack of rigorous empirical investigation of social risk.¹⁸ Recent research in the field of HIV/AIDS has undermined the claim that reporting in and of itself inhibits testing rates generally, but also reinforced the conclusion that the availability of anonymous testing is important to reach people at risk regardless of the reporting rules in place. In the absence of sufficient data, we should be very cautious about assuming that social risk will be a problem influencing the willingness of people to be tested under conditions of reporting, or that reporting will increase social risk. We can and should avoid broad claims about the social risks of genetic technology in favor of a more nuanced and contextualized analysis of both the possible threats and the perceptions of risk posed by particular kinds of genetic information.

A first step for further investigation -- and for policy-making in the inevitable absence of adequate data -- is to define a sufficiently broad universe of potential harms associated with genetic surveillance. It is helpful to consider a range of vulnerability, which can be both economic and psycho-social. Economic harm includes, potentially, losing access to insurance, a decrease in employability, and loss of economic support (as, for example, where genetic information leads to a divorce). Psycho-social vulnerability includes stigma, loss of social status or marriageability, and exposure to social hostility because of one's condition. (Stigma, which is classically defined as a sense of spoiled identity shared by the person with the stigmatizing condition, should be distinguished from social hostility, defined as negative social attitudes towards a condition which are not shared by the person with the condition.¹⁸) We should assume that vulnerability will vary from person to person and with the type of genetic information collected.

The absolute risk of economic harm caused by surveillance is probably quite low, simply because the likelihood of the information being divulged by public health agencies is low,¹⁷ and can be further reduced by collecting the data anonymously or with coded identifiers. It does not follow that the perception of risk will also be low. People who are highly averse to the consequences of a breach may overestimate the risks of its occurrence, particularly if they have heard horror stories of lost privacy. Similarly, people's attitudes to surveillance may be a function of their attitudes towards government or professional elites generally. They may not understand that the health department is independent of the police department. Fear of government data collection may be based on bad experiences with the data collection practices of private employers or insurers. Commentators have observed that many African-Americans mistrust medical and public health authorities, and are dubious of professional claims that new medical technologies will be used for socially just ends or in a fair way. We cannot assume that people subject to surveillance know what surveillance is, what conditions are subject to reporting,³³ or who has access to the information. Even people who understand that data are being collected by a trustworthy health department may perceive a risk that a future legislature or administration will change the rules and use the data in a harmful way.

The risk of psycho-social harm from surveillance may be greater than the risk of more tangible economic losses. Even if no one else learns about a person's genetic information, the subject may feel anxiety about the government possessing personal information. If the subject feels stigma, she may suffer the agonies associated with strategies of concealment and selective disclosure.³⁵ The creation of the data makes the subject herself a potential source of disclosure, and the harm to an individual's socio-economic status resulting from a disclosure of genetic information will be no less real because the leak comes from the individual rather than the government or a health care professional.

Assessing the actual and perceived social risk of genetic surveillance is complicated by the fact that most surveillance information is created outside the public health system for other purposes. Most genetic information will be generated by physicians in the health care system, through screening tests and medical histories. Such health care records currently enjoy much less protection than public health data, are used for more purposes, and by more actors. The same information that is absolutely protected from disclosure by the health department may be readily available to an employer or an insurer from medical records. Under these circumstances, the marginal increase in psycho-social risk entailed in collecting the data for public health purposes is virtually nil. Ironically, opposition to surveillance may arise in part because public health data collection is perceived as unnecessary (compared to medical care) and preventable (as compared to collection by insurance companies and employers). Conversely, people who are highly conscious of the poor protection accorded to health records may resist surveillance because they do not see a distinction between public health agencies and other collectors of data.

If the perception of a risk from surveillance is a combination of more generalized fears of, and attitudes towards, the use of genetic information, it may often be meaningless to speak of the social risks of surveillance. A perception of high social risk associated with the information being collected by anyone has the potential to influence the acceptability of surveillance at either or both the policy level (should surveillance be conducted?) and the individual one (will people carrying the genetic information avoid the data collection process?). Moreover, individuals may be as much concerned with anticipated psycho-social as economic harms. It follows that promotion of the acceptability of surveillance is dependent upon a broader effort to address the social risks of collecting genetic information, and must consider more aspects of the problem than are necessarily encompassed in terms like breach of privacy and discrimination.

Addressing the Threat of Social Risk: Privacy of Public Health Data

The threat of social risk directly caused by surveillance can be addressed, at least in part, by strict state public health records privacy laws. Virtually all states do, in fact, protect the privacy of public health data, but the protection is uneven, not simply across the states but within them. States generally provide some protection to health data collected for public health purposes, though it is quite common for the level of protection to vary with the type of disease.

Public health privacy laws generally encompass information gathered by public heath agencies in the course of executing their duties. Disclosure is allowed for public health purposes, such as case investigation, but with an explicit or implicit requirement that release be necessary to achieve a bona fide public health goal. Disclosure is also commonly authorized for research purposes, with conditions that would preclude further dissemination of identifying information about an individual. In addition, many states authorize disclosure pursuant to subpoena or court order (e.g., Arkansas, California, Hawaii ). Other states place disclosure within the discretion of public health officials (e.g., Pennsylvania).

Some states have reformed their laws to protect all information that the government collects, including communicable disease data, under a single statute. A small number of states have adopted legislation based upon a model statute for protecting and disclosing public health data.⁵³ In most states, however, the statutes are decades old and reflect changes neither in disease nor record-keeping technology.⁴⁵ Responsible public health practices in health departments make flaws in privacy law largely a matter of theoretical concern, but reform of public health privacy laws is nevertheless an important element of an effective response to the social risk of disease.

Having recognized that the greatest risk of breach of privacy probably comes from the original collector of the data, or the insurer or employer who acquires access a payer for health care, we have to address the broader issue of medical records privacy as it applies to genetic information. It is useful to think of a genetic information infrastructure, defined as the basic, underlying framework of collection, storage, use, and transmission of genomic information (including human tissue and extracted DNA) to support all essential functions in genetic research, diagnosis, treatment, and reproductive counseling. Despite the technical problems and the cost, several governmental and private committees have proposed automation of health data, including genomic information. Several conceptual and technological innovations are likely to accelerate the pace of automation of health records: patient-based longitudinal clinical records which include genetic testing and screening information; unique identifiers and the potential to link genomic information to identifiable persons; and genetic data bases for clinical, research, and public health purposes. No matter how effective the protection of public health data, the threat of social harm arising from genetic information will depend to a considerable degree on the protection of privacy in non-public medical and insurance records.

Few states provide comprehensive protection to health care records.⁵⁵ Several states have passed specific genetic privacy legislation that dramatically restricts disclosure of and access to genetic information without the informed consent of the subject. The informed consent requirements generally apply to the conduct of a genetic test in a clinical setting. Some of the statutes further extend the informed consent requirements to any subsequent release of the information, unless covered by a statutory exception, and so in some cases could be read to prohibit reporting for public health purposes.^{62 63} Some of these laws include a specific prohibition on discriminatory uses of the information.²¹ Some include restrictions and requirements on retention, disclosure, and notice.⁶³ Some require destruction of biological samples and test results after a set amount of time, at the completion of a project or at the request of the individual. Common exceptions may include the release and retention of data used in anonymous research, genetic testing conducted pursuant to newborn screening requirements, law enforcement, and paternity testing.^{62 63}

Meaningful legal protection of personal health privacy, although a popular concept, has proven difficult to establish as law. The pressure for access arises in large part because of the usefulness of the data to insurers, employers and other commercial and official actors. Public health, too, benefits from ready access to health records. Surveillance would in many instances be enhanced by the capacity to search computerized medical records. At a minimum, it has traditionally depended upon the ability of physicians freely to report cases to the health department. While protecting medical privacy probably helps reduce the perception of social risk associated with public health data collection, it is important to bear in mind that privacy laws that do not distinguish between public health and individual medical records will place unnecessary and counterproductive burdens on public health work. Applying a requirement of individual informed consent to information reported to or shared among health departments would be particularly harmful.

Addressing the Perception of Social Risk from Surveillance

Though the actual risk of social harm directly caused by surveillance is low, perceived risks (and higher actual threats arising in other settings) can create a context in which public health data collection is politically problematic or is resisted by subjects. Public health agencies therefore have a stake in assuring that people are protected from the social risks of genetic information they carry, and are aware of this protection. Law is one important source of protection, but health officials must be conscious of its limitations.

Laws addressing the threat of improper release and discriminatory behavior can help reduce the perception of risk, and can do so even if they do not actually reduce the threat. Commentators on genetic issues have extensively discussed protection from discrimination, involuntary testing, and breach of the privacy of medical records.⁵⁶ Proposals such as the creation of law-enforcement databases of genetic information have been assessed for their possible impact on the perception of social risk. At least half the states have passed legislation limiting genetic discrimination by health insurers, and several have laws prohibiting genetic discrimination in employment.⁷⁵ The Federal Health Insurance Portability and Accountability Act of 1996 (HIPAA) prohibits insurers from considering a genetic trait without the manifestation of disease as a preexisting condition for coverage purposes. Both scholars and litigators have explored the applicability of general disability, race and gender discrimination laws to people who have or are perceived to have gene-based impairments. The creation of property rights in genetic information is another possible means of increasing the control that individuals have, and perceive themselves to have, over their genetic information.⁷⁵

Limits of law as an antidote to social risk
Adequate privacy and antidiscrimination legislation is justifiable from a purely moral perspective, and is the necessary foundation for a practical system of public health surveillance. For all its potential, however, law is subject to several limits to its effectiveness in influencing individual health behavior, and may have some counterproductive influence. Law does not deal directly with many forms of social risk. It does not, for example, protect individuals from current social hostility or stigma that does not take the form of actionable discrimination, though it may reduce overt expression of negative attitudes and even help change attitudes in the long run. It may not provide protection against discrimination for an action, such as having an abortion, that arises because of genetic information but is itself distinct. Even those threats that are addressed are not always addressed completely. Very small employers, for example, are often covered neither by federal nor state discrimination law. Likewise, it will be impossible even with the strictest confidentiality laws to keep genetic information out of the hands of the insurers who are paying for its creation. At least some genetic information will be considered valid underwriting information that insurers will legally be able to collect under our current system of insurance regulation.⁷⁸ Many commentators regard HIPAA as a failure, because of insurers' ability to use genetic information in setting coverage rates.

More broadly, the notion that a law will influence individuals to provide a risky piece of information to a health care provider or the state based is deceptively simple. An individual's willingness to rely on the law in making choices about future health behavior is subject to at least three factors whose presence cannot be assumed. To be directly influenced in genetic-related health behavior by law, a person must be aware of the existing legal protection. Knowledge of gene-related legal protection among those who have or may have socially-risky genetic information has not been studied, but the available data tell us that individuals are rarely well-informed about the law, and often do not take legal action when they could reasonably do so.

Second, people are unlikely to rely on legal protection in planning their conduct if they do not perceive themselves to be entitled to legal protection. There are many reasons why a person who thinks he might carry socially risky genetic information might not believe he is entitled to legal help. Stigmatized people sometimes blame themselves for the social risks they encounter, or regard them as inevitable. Some people may mistrust "the system" as a general matter, or believe it to be designed to protect the rights of more powerful people and institutions in the society. Many people will have had purely negative interactions with the system, such as being arrested or harassed by police. These factors may make the law particularly weak in dealing with the social risk perceptions of marginalized populations.

Finally, a person willing to rely on the law for protection in the event genetic information causes harm must presumably believe that it will be effective in providing the protections it promises, including adequate remedies. For many people, the prospect of a money award some months or years after losing a job or suffering public humiliation may not appear to be a good trade-off for the health benefits secured by running the risk. This factor is particularly important given that the individual can easily choose the almost completely reliable option of not creating the datum in the first place.

The dangers of "genetic exceptionalism"
Laws that treat genetic information as uniquely problematic reflect and strengthen a sense of "genetic exceptionalism," the idea that genetic information is unique and deserving of special protection. Although not all statutes distinguish genetic information from non-genetic medical information, proposed and existing legislation often treat genetics as uniquely powerful and distinct from the rest of medicine. This approach poses several problems.

Although genetic data can be regarded as distinct in its implications for family members and its rare ability to predict future illness, genetic data is more the same, than different from, other health information. One might narrowly define "genetic information" as the product of DNA-related tests, but much genetic disease is diagnosed through direct clinical observation and family history. Broader definitions may expand genetic information to all genetic-related diseases, yet many diseases are known to have a genetic component and more, if not most diseases, are likely to be found to have genetic links in the future. Even classic infectious diseases such as HIV infection have genetic components. Scientists have shown that some genetic variations inhibit acquisition of infection and it is likely that the degree to which a person progresses from HIV infection to AIDS will, in part, be genetically determined. It is also difficult to separate genetic from non-genetic information in a medical record. Are sex, eye color, blood type, and nationality genetic or non-genetic data? As we learn more about the role of genetics in disease, attempts to separate genetic information from other types of information will become even more vexing. For these reasons, some commentators have suggested that good reason exists to treat genetic information no differently than other medical information.

This problem is related to a second: providing special protection to genetic information will often be unfair because it treats people facing the same social risks differently based on the biological cause of their otherwise identical health conditions. Why, for example, should a woman who has developed breast cancer of genetic origin (e..g, BRACA 1 or 2) be given greater protection than a woman who has developed breast cancer because of environmental or behavioral factors (e.g., smoking)? Even diseases that are solely genetic are not inherently special so as to give rise to unusual claims to legal protection. A debilitating, ultimately fatal condition exposes those who have it to pain, emotional distress and social risk with substantially equal severity whether it is caused by a gene, like Huntingdon's, or a virus, like HIV. Indeed, it is conceivable that genetic illnesses are less in need of legal protection than sexually transmitted diseases, which often carry a stigma arising from the perception that they are just desserts for voluntary behavior.

On a practical level, we must be cautious that the very people whom policy makers hope to encourage to take advantage of genetic testing may become more reluctant because of the heightened focus on its exceptional nature. Treating genetics as distinct from the rest of medicine may enhance the stigma of genetics testing, even as legislators attempt to remove its stigmatizing effects. This can create public fears and misapprehensions about genetics that could discourage individuals from seeking testing and treatment, and thwart future scientific progress. Conversely, by focusing only on genetics information, legislators may convey the perception that the public need not worry about the confidentiality of other kinds of medical information, fostering complacency in an area where insufficient protections may exist.

The Public Health Ethics of Surveillance

The discipline of bioethics has become influential in the development of health care policy and medical practice. Public health, although often seen as a quintessentially utilitarian enterprise, likewise raises profound moral questions that can benefit from an ethical analysis. If "public health" may be defined as the process of maximizing, within the bounds of possibility set by the resources allocated, the level and distribution of satisfactory health in a population, then its pursuit entails an endless series of choices about the definition and determinants of health, and the best means of its just and effective attainment.⁴⁵ The ethical system that has developed for analyzing health care decisions is of limited use in considering measures aimed at creating the conditions in which the population as a whole can be healthy. Fundamental questions of equality and justice in the distribution of scarce resources and the allocation of the burdens of disease control and prevention require a distinct "ethics" of public health. Such a system of thought has been slowly emerging over the past two decades,J Health Pol, Policy & Law, 1976-77; 1:338-54. taking shape in such documents as the Council for International Organizations of Medical Sciences' 1991 International Guidelines for Ethical Review of Epidemiological Studies. It was one of the interests of Jonathan Mann, whose untimely death is a loss to this endeavor as to so many others. Such a system would recognize the limits of an individualistic orientation in assessing the costs and benefits of population-oriented health measures, by, for example, focusing less on individual relative risk and more at population attributable risk. It would place on those who are at low personal risk an obligation to make or accept changes in social behavior necessary to reduce the toll of ill health on others in the community.⁹⁷ At a minimum, "[s]uch an ethic would ensure that no group or person was arbitrarily excluded from protection and that the burdens of collective action were fairly shouldered by all."⁹⁷ (p343).

In the absence of a fully developed system of public health ethics, we may still profitably ask three questions that can help us assess whether surveillance for genetic information is a just tool for public health.

1.  Will surveillance enhance the health of the population? To find a genetic surveillance measure "ethical" in the public health sense, we need to determine that it has a reasonable probability of leading to information that can be used to reduce the incidence of genetic illness or mortality in the society as a whole or in a significant population facing a special threat. If it does not, the measure, however useful to individuals, ought not be an important public health priority, and may not be worthy of significant public sector resources. By the same token, however, a measure may be compelling in its importance to public health even though it has limited value to any one person in the population.

2.  Is surveillance for particular genetic information a wise and just use of resources? Rationing in public health is a "moral imperative ... in the face of scarce resources." Ideally, each public health dollar will be spent to achieve the greatest marginal increase in the level and distribution of well-being in the population, and the socio-economic costs of any intervention -- including the social risk -- will not fall disproportionately on any group within the population. Public health ethics make explicit the competition for scarce resources that has often proven difficult to address in the patient-centered gaze of medical ethics. Measures that are of significant value to some relatively healthy patients may be ethically dubious from a public health perspective if the resources could be used instead to improve the lot of a relatively less healthy group, or to increase by even a small degree the overall health of the population. Similarly, the individual pursuit of low relative risk has no place in the ethical practice of public health when it comes at the expense of reduction in the population attributable risk. Public health measures may entail infringements on individual civil liberties, and more generally upon an individual's status in the society, but only where they can be justified by a compelling necessity.

3.  Will surveillance be acceptable to the population being observed? This principle has roots in both practicality and morality. Public health depends largely on voluntary compliance with its guidelines. Resistance can raise costs enormously, the more as prevalence of the condition rises. The success of a public health intervention, and its cost, therefore depend significantly on the degree to which its targets perceive it to be more beneficial than costly.

Genetic information could often be useful to the public's health. Genomic data can help track the incidence, patterns, and trends of genetic carrier states or disease in populations. Carefully planned surveillance or epidemiological activities facilitate rapid identification of health needs. This permits reproductive counseling, testing, health education, and treatment resources to be better targeted, and points the way for future research. Even the collection of data about low prevalence condition or a small population could have long-term benefits to population health through the advancement of biomedical knowledge. Surveillance typically provides its benefits at a low economic cost, and the marginal cost of adding information to the list of already reportable data is probably slight. The main concern about collecting genetic information is its potential social risk.

We have already suggested that the social risk to individuals remains largely unexplored by empirical research, and may well have been overstated. To properly assess genetic surveillance, we need more information about both the perception and the threat of social risk. Given the potential variability of the phenomenon, it would be most useful to gather this data as specific types of information are proposed for surveillance. Whatever the level of risk, both the threat and the perception may be reduced through such devices as anonymous or coded collection of data, and strong legal guarantees of privacy, nondiscrimination and access to medical care.

Although we consider surveillance to pose a low social risk to individuals, a more diffuse but equally compelling risk of genetic technology has been identified and may be implicated by surveillance: an ethical public health policy would look at broader social questions arising from how we construct genetic health, in part through surveillance. A new technology providing new forms of information about health and disease can have dramatic effects on how these concepts are understood in society.¹⁶ In a society that sometimes seems enamored with the power of genomic information, genomic data are often thought capable of explaining much that is human: personality, intelligence, appearance, behavior, health, and disease. A widespread belief that health is "in the genes" may have the effect of bolstering a cultural conception of health as individualized and dependent upon medical management. Genetic screening will add yet more "risk-factors" to the individual's medical health inventory, and enhance the view that each individual is responsible for his own health. While this conception of health as some advantages in disease control, it also can lead to victim-blaming and a neglect of communal provision for health.

Genetic concepts of illness may also promote group blame and eugenic ideas in new guises.³¹ Susan Wolf has warned of the rise of what she calls "geneticism," which, "like racism and sexism, ... is a long-standing and deeply entrenched system for disadvantaging some and advantaging others" based on their genetic traits. Proponents of surveillance for genetic information will have to recognize and take some responsibility for addressing the social meaning of genetic information.

Recommendations

Our analysis of ethics, law and social risk suggest this list of needs. 

From researchers:

• More research on the social risk of genetic information, focussing on specific kinds of information, and examining both the prevalence of socially harmful behavior and the perception of social risk.
• More research on to what extent, and how, individuals rely on law in making health choices.
• Research to determine what measures are effective in reducing perceptions of social risk.

From legislators:

• Effective protection of medical privacy, and separate and even stricter protection of the privacy of public health records.
• Effective protection of all citizens against discrimination based on genetic information. Liberally construed disability laws may be sufficient for this purpose, but it is not yet clear whether the ADA and state laws will provide this protection. Race and sex discrimination laws may also play a role, wherever conditions are sex linked or associated with racial or ethnic groups.
• Efforts to reform or ameliorate the effects of our current system of health insurance on access to care for people with heightened risks.

From public health agencies:

• Careful attention to the complex relationship between overall genetic social risk and the particular social risks of surveillance.
• Use of appropriate public health ethical criteria for assessing public health uses of data.

In considering new legislation to address concerns about the use and misuse of genetic testing and genetic information, it is important particularly at the state level to distinguish between public health and other uses of genetic data, and to review and consider the existing state framework for public health and medical research. Public health authorities and researchers should be afforded the opportunity to define their needs, including access to information to conduct public health functions. Having a clear understanding of the various programs that may involve genetic testing and information, states can then assess the potential impact of proposed legislation in the areas discussed above. Any potential impact on public health programs and research, should be carefully balanced against the underlying goals of the legislation. In some cases the goals of the legislation may be promoted by specifically addressing deficiencies in the public health framework, particularly the legal ability to protect participants, or provide necessary confidentiality protections.

Conclusion

Surveillance (more data) is usually good, and poses little risk for people whose data are collected. The actual risk of surveillance can be minimized by rigorously protecting public health data from being used for other purposes, such as law enforcement, and by promoting social conditions in which people are safe from mistreatment and discrimination based on genetic make-up. No matter how well we succeed in reducing the threat of social risk, surveillance will still often be controversial, because of concerns about the social risks posed by the collection of sensitive data. For surveillance to be ethical from a public health point of view, the data have to efficiently promote the health of the population, their collection should be acceptable to the population, and the data must be protected. To protect public health data, we need effective laws protecting public health records, but we must also bear in mind that legal protection may not always be sufficient to reduce perceptions of social risk, particularly among those who are suspicious of government or alienated from the legal system. Implementing surveillance programs about which there are serious questions of acceptability requires a solid foundation of laws prohibiting arbitrary discrimination based on health status, but also political and other efforts to address perceptions of social risk that cannot be eliminated by law.

References: 

1. Centers for Disease Control and Prevention. Translating advances in human genetic into public health action - a strategic plan. Atlanta, GA, October 1, 1997

7. Holtzman NA, Watson MS eds. Promoting safe and effective genetic testing in the United States: final report of the task force on genetic testing. Baltimore: Johns Hopkins University Press, 1998.

8. Khoury MJ, Beatty T, Cohen B. Fundamentals of genetic epidemiology. New York and Oxford: Oxford University Press, 1993.

9. Khoury MJ. Genetic epidemiology and the future of disease prevention and public health.  Epidemiol Rev 1997;19:175-180.

10. Iowa Code Ann.§136A.4-6 (1997).

11. Ky. Rev. Stat. Ann. §211.665 (1997).

12. Minn. Stat. Ann. §13.385 (1997).

13. Mo. Rev. Stat. §191.323 (1997).

14. Fox DM. Social policy and city politics: tuberculosis reporting in New York, 1889-1900.  Bull Hist Med 1975;49:169-94.

15. Rothman S. Seek and hide: public health departments and persons with tuberculosis, 1890-1940. J Law, Med & Ethics 1993;21:289-95.

16. Brandt AM. No magic bullet: a social history of venereal disease in the United States since 1980. New York and Oxford: Oxford University Press, 1987.

17. Gostin LO, Hodge J Jr. The "names debate": the case for national HIV reporting in the United States. Albany L Rev 1998;61:679-743.

36. Hall MA. Insurers' use of genetic information. Jurimetrics J 1996;7(Fall):13-22.

37. Tversky A, Kahneman D. Availability: a heuristic for judging frequency and probability. In D. Kahneman, P. Slovic & A. Tversky (Eds.), Judgement under uncertainty: heuristics and biases. Cambridge: Cambridge University Press, 1982:163-178.

38. Allen AL. Genetic testing, nature, and trust. Seton Hall L Rev 1997;27:887-892.

39. Roberts DE. The nature of blacks' skepticism about genetic testing. Seton Hall L Rev 1997;27:971-79.

40. Burris S. Driving the epidemic underground? A new look at law and the social risk of HIV testing. AIDS & Pub Pol J 1997;12:66-78.

41. Mansergh G, Marks G, Simoni J. Self-disclosure of HIV-infection among men who vary in time since seropositive diagnosis and symptomatic status. AIDS 1995; 9: 639-44.

42. Lerman C, Peshkin BN, Hughes C, Isaacs C. Family disclosure in genetic testing for cancer susceptibility: determinants and consequences. J Health Care L & Pol'y 1998;1:353-372.

43. Gostin LO, Lazzarini Z, Neslund VS, Osterholm MT. The public health information infrastructure: a national review of the law on health information privacy. JAMA 1996; 275:1921,1923-1925.

44. Gostin LO, Lazzarini Z. Childhood immunization registries: a national review of public health information systems and the protection of privacy. JAMA 1995; 274:1793.

45. Gostin LO, Burris SC, Lazzarini Z. The law and the public's health: a study of infectious disease law in the United States (in press).

46. N.Y.C.L.S. Pub. Health §§ 2306 and 2782 (1994).

47. Cal. Civ. Code § 1798.24 (1995).

48. Ark. Stat. Ann. § 20-15-904 (1994).

49. Cal. Health & Saf. Code § 121070 (1996).

50. Haw. Rev. Stat. § 325-101 (1994).

51. Pa. Stat. Ann. tit. 35 § 521.15 (1998).

52. Minn Stat. Ch. 13 (1996).

53. Mont. Code Ann. § 50-16-603 (1994).

54. Rev. Code Wash. § 70.02.005 et seq. (1994).

55. Gostin LO. Health information privacy. Cornell L Rev 1995; 80:451-507.

56. Gostin LO. Genetic privacy. J Law, Med & Ethics 1995; 23:320-330.

57. Congressional Office of Technology Assessment. Protecting privacy in computerized medical information. 1993; OTA-TCT-576.

58. U.S. Department of Health and Human Services. Final report of the task force on the privacy of private-sector health records. 1995.

59. Donaldson MS, Lohr KN (Eds). Health data in the information age: use, disclosure, and privacy. Washington: National Academy Press, 1994.

60. Gostin LO, Turek-Brezina J, Powers M, et al. Privacy and security of personal information in a new health care system. JAMA 1993; 270:2487-2493.

61. Confidentiality of Medical Information Act, Cal. Civil Code §§ 56 to 56.37 (1982 & Supp. 1998).

62. Ga. Code Ann. § § 33-54-1 to -8 (1997).

63. N.J. Stat. Ann. §§10:5-43 to -49 (1997).

64. Ill. Comp. Stat. §§ 513/5-30 (1998).

65. Or. Rev. Stat. §§659.700-659.720 (1997).

66. La. Rev. Stat. Ann. §213.7 (1998).

67. Bornstein RA. Genetic discrimination insurability and legislation: a closing of the legal loopholes. J L & Pol'y 1996;4:551-610.

68. Rothenberg K, Fuller B, Rothstein M, et al. Genetic information and the workplace: legislative approaches and policy challenges. Science 1997;275:1755-1757.

69. Ind. Code §27-8-26-5 (1998).

70. Kan. Stat. Ann. §40-2259 (1997).

71. Tex. Ins. Code Ann. §21.73 (1997).

72. Cal. Health & Safety Code § 1374.7 (1997).

73. Vernon's Tex. Code Ann., Labor Code, §21.401 to 21.405 (1997).

74. Or. Rev. Stat. §51- 659.710 (1997).

75. Yesley MS. Protecting genetic difference. Berkeley Tech L J 1998;13:653-665.

76. Mehlman MJ, Durchslag MR, Neuhauser D. When do health care decisions discriminate against persons with disabilities? J Health Pol Pol'y & L 1997;22:1385-1408.

77. Rothstein MR. The law of medical and genetic privacy in the workplace. In genetic secrets: protecting privacy and confidentiality in the genetic era. Rothstein M (ed.). New Haven, CT: Yale University Press, 1997: 281-288.

78. Kass NE. The implications of genetic testing for health and life insurance. In genetic secrets: protecting privacy and confidentiality in the genetic era. M Rothstein (ed.). New Haven, CT: Yale University Press, 1997:299-316.

79. McEwen JE. DNA data banks. In genetic secrets: protecting privacy and confidentiality in the genetic era. Rothstein M (ed.). New Haven, CT: Yale University Press, 1997; 231-51.

80. Pub L No. 104-191, 104th Cong, 2d Sess (1996).

81. Blanck PD, Marti MW. Genetic discrimination and the employment provisions of the Americans with Disabilities Act: emerging legal, empirical, and policy implications. Behavioral Sciences and the Law 1996;14:411-32.

82. Gin BR. Genetic discrimination: huntington's disease and the americans with disabilities act. Colum L Rev 1997;97:1406-1434.

83. Norman-Bloodsaw v. Lawrence Berkeley Laboratory, 135 F.3d 1260 (9th Cir. 1998).

84. Lessig L. The regulation of social meaning. U Chicago L Rev 1995;62:943-1045

85. Sunstein CR. Social norms and social roles. Colum L Rev 1996;96:903-968.

86. Cocco M. The right medicine for health care. Newsday 1998; Jul 14: A30.

87. Ellickson R. A critique of economic and sociological theories of social control. J Legal Studies 1987;16:67-99.

88. Weiler PC, Hiatt HH, Newhouse JP, Johnson WG, Brennan TA, Leape L. A measure of malpractice: medical injury, malpractice litigation, and patient compensation. Cambridge, MA: Harvard University Press, 1991.

89. Coates D, Penrod S. Social psychology and the emergence of disputes. Law and Society Rev 1980-81; 15:655-680.

90. Musheno MC. Legal consciousness on the margins of society: struggles against stigmatization in the AIDS crisis. Identities 1995;2:102-22.

91. Me Rev Stat Ann tit. 18-A, § 9-310 (1996) .

92. Ohio Rev Code Ann. § 3729.46 (1997)

93. N.J. Rev. Stat. § 2a (1996)

94. NIH-DOE Working group on ethical, legal & social implications of human genome research, genetic information and health insurance 1993.

95. Rothstein MR. Genetic privacy and confidentiality: why they are so hard to protect. J L Med & Ethics 1998;26:198-204.

96. Markel H. The stigma of disease: implications of genetic screening. Am J Med 1992; 93:209- 215.

97. Beauchamp DE. Exploring new ethics for public health: developing a fair alcohol policy.  J Health Pol, Policy & Law, 1976-77; 1:338-54.

108. Wolf S. Beyond 'genetic discrimination': toward the broader harm of geneticism. J L Med & Ethics 1995; 23:345-49.