Hostname: page-component-7b7567c8c-fvjsw Total loading time: 0 Render date: 2026-11-14T13:58:05.647Z Has data issue: false hasContentIssue false
  • English
  • Français

Bioethics, Business Ethics, and Science: Bioinformatics and the Future of Healthcare

Published online by Cambridge University Press:  01 October 2008

KENNETH W. GOODMAN and
ANITA CAVA
Rights & Permissions [Opens in a new window]

Extract

The intersection of ethics, computing, and genetics plots a space not yet adequately mapped, despite its importance, indeed, its rapidly growing importance. Its subdomains are well-enough known: “Genethics,” or the study of ethical issues in genetics and genomics, is part of core curricula everywhere. Ethics and computing is an established subfield. Computing and genetics—bioinformatics—has in little more than a decade progressed from subsubspecialty to the sine qua non of contemporary biomedical research, and it bids fair to transform clinical practice. We must prepare for the complete digitization of the genomes of individual patients and the storage of millions of these genomes in very large databases.

Information

Type
Special Section: The Newest Frontier: Ethical Landscapes in Electronic Healthcare
Information
Cambridge Quarterly of Healthcare Ethics , Volume 17 , Issue 4 , October 2008, pp. 361 - 372
Copyright
Copyright © Cambridge University Press 2008

The intersection of ethics, computing, and genetics plots a space not yet adequately mapped, despite its importance, indeed, its rapidly growing importance. Its subdomains are well-enough known: “Genethics,” or the study of ethical issues in genetics and genomics, is part of core curricula everywhere. Ethics and computing is an established subfield. Computing and genetics—bioinformatics—has in little more than a decade progressed from subsubspecialty to the sine qua non of contemporary biomedical research, and it bids fair to transform clinical practice. We must prepare for the complete digitization of the genomes of individual patients and the storage of millions of these genomes in very large databases. Moreover, “ethics” here is vague, making no distinction between two of the largest components of applied philosophy, bioethics and business ethics. This is regrettable because bioinformatics is now also big business.

This ensemble of bioethics, business ethics, information technology, genetics, and genomics presents a hydra-headed challenge, the same one faced when the shoulder of moral inquiry is put to the wheel of any new science. It requires identifying the issues, suggesting first approximations of best practices, and recommending ethically optimized public policy options. This is a grand project for ethics and bioinformatics, and it should be uncontroversial to suggest that it be undertaken comprehensively. That is, however, not our mission here.

Our goal is to build upon earlier work, lay out some of the issues that seem to us most interesting and urgent, and suggest possible paths for future research. The bioethics landscape is often (re)shaped by hot topics; neuroethics and nanoethics stand as recent examples. We will succeed here just in case the intersection of ethics and bioinformatics begins to enjoy some of the attention it deserves and that the ethics community ought both to require and provide.

Ethics and Health Informatics

Interesting and important ethical issues attach to the use of computers and other devices to collect, store, analyze, and transmit personal health information and biomedical research data.Footnote 1 The main issues have tended to lie along four axes:

  • • Privacy, confidentiality, and best practices for data management

  • • Decision support, including the use of diagnostic and prognostic engines

  • • Loci of accountability and responsibility for duties ranging from database design and system evaluation to educational standards and patient awareness

  • • Use of intelligent machines in human subjects research to monitor, track, sort, randomize and otherwise perform operations on data about people.

There has even emerged what has been called the “Standard View” of appropriate useFootnote 2 of intelligent machines for decision support in healthcare. According to a chapter in the leading textbook on health informatics,

human cognitive processes, being more suited to the complex task of diagnosis than machine intelligence, should not be overridden or trumped by computers. The standard view states that, when adequate (and even exemplary) decision-support tools are developed, they should be viewed and used as supplementary and subservient to human clinical judgment. They should take this role because the clinician caring for the patient knows and understands the patient's situation and can make compassionate judgments better than can computer programs; they are also the individuals whom the state licenses, and specialty boards accredit, to practice medicine, surgery, nursing, pharmacy, or other health-related activities. . . . The standard view addresses one aspect of the question, “How and when should computers be used in clinical practice?” by capturing important moral intuitions about error avoidance and evolving standards. Error avoidance and the benefits that follow from it shape the obligations of practitioners. In computer-software use, as in all other areas of clinical practice, good intentions alone may be insufficient to insulate recklessness from culpability. Thus, the standard view may be seen as a tool for both error avoidance and ethically optimized action. (p. 381)Footnote 3

Arguably, a standard view also exists regarding how best to approach data protection in a world of electronic patient records, personal health records, and hypervigilance in collecting data for outcomes research and evidence-based practice. Accepted practice is to provide ethics education, establish adequate security mechanisms, and institute appropriate policies and procedures.Footnote 4 The latest effort to harness information technology for improved healthcare—via personal health records, which patients and clinicians can use to share information and progress—elicits fascinating privacy issues even before genomic data are included.

It should be uncontroversial that bioinformatics will require new tools and insights if it is to realize its potential and do so in ways that are ethically sound. It is noteworthy, for example, that the United States’ attempt to harmonize privacy rights and protections in its Health Insurance Portability and Accountability Act (HIPAA) says little about genetic information. Indeed, HIPAA provides for only the most modest antidiscrimination protections despite a generation of research in ethics and genetics that makes clear the fact that genetic privacy is different, perhaps even special, when information is about genomes and not hypertension, hypothyroidism, or head colds.Footnote 5 Further, the literature on ethics and informatics is so far nearly silent on the kinds of business ethics issues raised by bioinformatics. The last section of this paper sketches some of them.

Bioethics and Bioinformatics

There is a burgeoning literature on ethics and bioinformatics that has laid out some of the issues.Footnote 6 We enumerate some of these concerns here, and later address those falling more properly under “business ethics” than “bioethics.” First, however, we consider a cornerstone question: Does the use of information technology add anything to what we already knew or worried about? Surely all agree that clinical and research data about individuals’ genomes raises privacy issues and, further, that such was the case before the data were digitized. What then is it about computers that changes or magnifies issues in privacy and confidentiality? Error reduction has long been seen as a virtue in clinical practice and research, but is there something ethically distinctive about errors in genetic databases and sequence annotations? We have long lamented the commercialization of medicine and the commodification of human biological material, but do proprietary databases or profit motives in personalized healthcare and pharmacogenomics require a refocus of attention?

Here we lay out what seem to be the best candidates for novel ethical issues raised by bioinformatics. Note well that many important issues in genethics are not mentioned here because (1) they are already well studied and (2) little of great consequence changes when information technology is used. We suggest, however, the possibility that an ethical issue at the intersection of genomics and computing becomes significant because of the magnitude of the effect of information technology. Simply put, differences of degree in addition to novelty of kind or type require heightened concern and focused analysis.

Privacy and Confidentiality

Bioinformatics, or “the use of information technology to acquire, store, manage, share, analyze, represent and transmit genetic data,”Footnote 7 raises interesting and, arguably, distinctive privacy issues. Indeed, for better or worse, it is the confidentiality of genomic information and its broader cousin, personal privacy, that are often itemized first in most lists of ethical issues arising from the use of computers in clinical practice and in bioinformatics. But privacy issues also inhere in a family history, a medical record note about amniocentesis, and the results of presymptomatic genetic testing for any heritable malady. Do computers have a lensing effect that magnifies these issues, or do wholly new issues arise?

To be sure, if test results are stored in a computer, then it will be computer-as-medium that might alter the privacy landscape. But computers do that for all clinical and research data. Privacy and confidentiality are then protected by means different than they would be if the data were written by hand in a hospital chart or even in the ink from crushed flowers on a papyrus scroll. There is not necessarily a new issue as such. Secondary use of genomic data might be easier with computers, but surely core principles governing such use will apply no matter the medium. Even the transmission of genetic and genomic data between computers is unremarkably and usually parallel to transmission via a postal service or telephone or facsimile, although security precautions will obviously differ.

We can nevertheless identify three ways in which information technology elicits novel issues, or those worthy of attention because of their distinctiveness. The first, an exception to the point just made about transmission, is that when entire genomes can be e-mailed and stored, the sheer volume of data raises novel privacy and confidentiality issues. The magnitude of data in transfer or storage increases the risks of inappropriate disclosure. Depending on security precautions, those risks might be greater than the risks of disclosure or discovery in other media.

The other ways information technology affects privacy considerations are more interesting. When genetic or genomic data are aggregated in large databases and analyzed, it is possible to draw inferences with increasingly fine granularity about population subgroups. These groups might correspond, or appear to correspond, to racial or ethnic (sub)communities. The concept of “group privacy” has been advancedFootnote 8 as a way of making full the initial recognition that genomic information is about others in addition to the source of the tissue or sera. Although we usually think of family members in this regard, the correlation to racial or ethnic groups cannot be ignored. It is particularly salient with regard to behavioral genomics and other ways of identifying stigmatizing maladies, especially in an era of community-based participatory research. This issue finds noteworthy resonance in discussions of genetic epidemiology and emergency public health informatics.Footnote 9

Third, the use of data mining software, enabling what is sometimes called “machine learning” or “knowledge discovery in databases,” provides a slight twist: In addition to the risk of violations of group privacy, there is also the chance that information a patient might want to protect could be discovered without it (immediately) being recognized as such.Footnote 10 Relatedly, database matching algorithms enable law enforcement authorities to compare gene samples from those arrested, even for minor crimes, to preexisting samples. Indeed, computational analyses have also been used to prove the innocence of those accused of crimes.Footnote 11

The most significant population-based genome mapping effort is the well-studied case of deCODE in Iceland.Footnote 12 Questions of privacy and informed consent were significant at the project's inception in 1999 and continue to loom large. Correlate efforts have begun in Estonia, Latvia, Quebec, and elsewhere; the United States government has created a repository for anonymized data generated during genomewide association studies (GWAS), including rules for data sharing.Footnote 13 deCode is a private corporation, and the U.S. database is public. We have yet to decide (1) how to frame our interests in evaluating whether privacy and other human rights are imperiled more by proprietary entities than by state-controlled repositories and (2) whether both pose similar threats and therefore require the same level of scrutiny.

The question of whether bioinformatics raises new privacy issues is therefore to be answered affirmatively: In some cases it is novelty by degree and in others novelty by kind. A comprehensive ethics-and-bioinformatics research program is needed to sort out which is which.

Appropriate Uses and Users

It is no great insight to conclude that many new tools and technologies, including devices and drugs, point to important ethical issues generally related to (in)appropriateness of use and of user. Some drugs and devices remain ethically interesting even after they are no longer new. For instance, appropriate use of morphine has long continued, and will long continue, to engender debate. Bioinformatics raises very interesting questions of appropriate use and user. One could even make the case that these questions are truly novel or, if not embodying novelty sui generis, do so by virtue of the magnitude of their difference from touchstone questions. Many familiar issues in genethics will acquire new urgency in a bioinformatics world.

Consider the use of microarray technology. In the most simplistic terms, genes are strings of DNA that are transcribed to messenger RNA and then translated to proteins, which make cells, which either work well—or not. The process of DNA transcription and mRNA translation is complex and underlies all biological function. In expression analysis or expression profiling, microarrays and gene chips analyze or profile gene expression. Thousands of genes can be profiled at the same time, giving insight into an individual's genome, its functions, and, of generally greater interest to the individual, its malfunctions. These profiles constitute data for research and are increasingly contributing information used in making clinical decisions. None of this would be possible without computers and, in some cases, data mining software: “Biologists have finally caught up with the computer age out of necessity.”Footnote 14 In the other direction: Genetics is no longer possible without computer science.Footnote 15

Who, then, should use a microarray and for what purpose? At least as interesting, given the case of an inchoate science, how should we determine when the answers to either question might plausibly change, or ought to change?

In the not-too-distant future, patients will submit cells to microarray analysis and receive—perhaps through an intermediary such as a physician or genetic counselor or even through the mail—an extensive expression analysis. The software used to produce the analysis will have to assess the results of gene expression, and these results will be handed or communicated to patients. Absent any great strides in science education, these individuals will be faced with a potentially bewildering ensemble of probabilistic findings, diagnoses, and prognoses. In the more distant future, we might imagine, patients will be able to buy microarray analyses in pharmacies. Indeed, 2007 seems to have ushered in the era of the “personalized genome” or the “thousand dollar genome” as for-profit companies began offering kits ($999 plus postage and handling) to individuals interested in purchasing their genomes. A saliva sample is sent for microarray analysis and the results are then posted to a (password protected) web site where individuals can “use our web-based interactive tools to explore [their] genome.”Footnote 16

This is a carnival of ethical issues. For instance,

  • • In addition to browsing information of medical and health interest, the analyses can be used to learn about ancestry. Bring the family: After receiving the analyses, users can compare theirs with those of family members. Is this healthcare or a recreational activity? Is either appropriate?

  • • Relatedly, is a DNA analysis any sort of diagnosis and, if so, should diagnostic tests be performed by laypeople?

  • • Is this the kind of analysis that should be directly available to consumers? Why or why not?

  • • How much about the analysis software, including its design and functionality, should professionals or laypeople know before using these services? Should the source code be open or secret?

  • • In a research context, do standards for human subjects research, including genetics research, work well enough when microarray output is stored for broader analyses?

The point is clear enough: The determinants of appropriate uses and users provide a rich source of ethical issues that arise when information technology is used to manage some bioinformatics applications. Our “ethics-and-bioinformatics research program” has its work cut out for it.

Error Management

Although aspects of patient safety and medical error have made their way into the bioethics curriculum, the kinds and frequency of errors that arise in bioinformatics should be seen as an interesting and important ethical challenge for scientists and clinicians. Unlike a gross medical error, one of which can be dramatic and almost always be prevented, automatic gene sequencing errors and infelicities in databases are reckoned to be a cost of doing business when working with billions of complex components and processes. We have previously identified a number of places in which bioinformatics errors raise ethical issues.Footnote 17 They include the following:

  • Risks to persons: If a “buggy” database is used for clinical diagnosis or decision support, patients could be harmed. False positive and false negative genome studies could cause psychological distress or lead to missed opportunities for disease prevention or management. Error rates beyond a certain threshold in research databases can waste time and resources and impede research programs.

  • Faulty linkage studies: The association of a gene or genes with a malady is in principle a great boon to science and medicine. But some linkage analyses have been found to be faulty and been recanted. This is especially a problem when there is a risk of bias or stigma, as in the case of behavioral genomics, and when there is a putative link to racial, ethnic, or other population subgroups.

  • Decision support: Clinical decision support has long been recognized as a source of exciting ethical issues, not least the question of responsibility for (1) maintaining the databases on which decision support or diagnostic expert systems rely and (2) judging the quality and reliability of any machine-generated advice. These issues are no less urgent when genomic data are digitized and used to guide treatment or inform genetic counseling.

Drug Discovery and Personalized Medicine

The completion of the Human Genome Project, the subsequent creation of the International HapMap Project, and the extraordinary growth in genomewide association studies have established for scientists the era of “systems biology”Footnote 18 and for clinicians that of “personalized medicine.”Footnote 19 Although we will discuss pharmacogenomics shortly, it will be well to linger a moment here, in part because some of the hyperbole associated with personalized medicine might actually be justified.Footnote 20

It is neither naïve nor overly enthusiastic to contemplate an era that will see the treatment and management of certain maladies utterly transformed. In cancer treatment, for instance and perhaps especially, we are already witnessing what is called “molecular substaging”—the use of expression analyses to contribute to the staging of tumors to “determine appropriate management strategy.”Footnote 21 Put differently, the information-technology-dependent tasks of genomic analysis are already finding their way into the clinic. As this science evolves, future cancer patients will be treated with customized drugs developed or titrated specifically for them based on the kind of cancer they have and findings from rapid genome analysis.

Vast amounts of genomic data from genomewide association studies and other sources are needed to fledge these clinical uses. The ethical issues raised by the HapMap Project, for instance,Footnote 22 should be familiar. Even without the storage of personal identifiers, which reduces privacy concerns, the challenges identified more than a decade agoFootnote 23 still exist, but now with the recognition that one means of meeting these challenges includes the use of “community consultations”Footnote 24 and other tools of community-based participatory research.

But the idea that we could mull data sharing, community consultations, and the concerns of traditionally stigmatized populations is perverse unless we also consider in bold face the chance that we are once again going to establish a new medical science the benefits of which will be available only to those who can afford it.

Business Ethics and Bioinformatics

Although access to care has long been a component of the core bioethics curriculum, we address it as one in a suite of business ethics issues in bioinformatics. There are several justifications for this, not least among them that business interests beyond the control of clinicians govern the availability of expensive new technologies. It is not appropriate that physicians be required to resolve at the bedside problems rooted in corporate boardrooms or stock markets. Indeed, the case is often made that society has the duty of addressing macroeconomic challenges and clinicians and their institutions that of managing microeconomic ones. This division of moral labor continues to leave us with a chasm that we can only metaphorically bridge here—by identifying issues at the intersection of business ethics and bioinformatics.

Pharmacogenomics—Collaboration and Competition

Scientists can identify a significant number of genetic variants that serve as pharmacogenomic markers. These in turn correlate in varying degrees with drug response and outcome.Footnote 25 Pharmacogenomics is the study of “variation in drug response … related to genetic variation,”Footnote 26 a concept also embraced by the FDA, which has approved a number of commercially available tests.

The availability of designer drugs based on genetic markers is about to create a market in which pharmaceuticals are tailored to individual needs, maximizing effect while minimizing error, toxicity, and other adverse events. The potential is enormous, but a caveat is obvious: “Available in the marketplace” does not always mean “available to patients.” Many barriers are already in place, including the selection of target drugs and maladies largely on the basis of market potential, the regulatory framework designed to protect patients, and the economic allocation of resources to move these drugs into patients’ hands.Footnote 27 The healthcare crisis in the United States today, with approximately 45 million uninsured and many millions more underinsured, many of whom are members of minority groups, raises an enormous obstacle to the development and distribution of these promised benefits.Footnote 28 Indeed, “[d]riven by advances in genomics, emerging insight into each individual's unique susceptibility to disease promises to transform patient care. However, such advances will also compel a fundamental restructuring of the way medical care is delivered in the United States. … There are many reasons to pursue a rational, just, and workable system of health care.”Footnote 29

The ability to profile genetic information through microarray technology opens the doors to finer-grained analyses of patient care. Mapping gene expression offers guidance in diagnosing and treating specific manifestations of disease. Research and development in this area is both highly collaborative and highly competitive, signaling the social and ethical implications of the results.

National and international collaboration is evident in the HapMap project; the federally funded mouse Brain Gene Expression Map (BGEM) developed at St. Jude's Children's Research Hospital, which offers a potent gene expression map and is available to scientists worldwide for the purpose of studying the genetic origins of brain cancer (http://www.stjudebgem.org/web/mainPage/mainPage.php); and the Cancer Genome Atlas (TCGA), a collaborative, multiyear, $100 million project to test the feasibility and utility of a vast database of cancer genomic data that “will have consequences for how scientists study cancer worldwide and how benefits are distributed both inside and outside the United States” (p. 1962).Footnote 30

Competition, on the other hand, is evident in the movement of stellar researchers lured from one institution to another; the promise of equipment and support obviously linked to the promise of financial success; the jostling for intellectual property rights and control; and, of course and most generally, the entrepreneurial engine that has driven pharmaceutical research for generations. Cancer, for instance, is a major healthcare concern as well as very big business, and so it offers an excellent case study of the tensions in play:

Cancer prognostics will be among the first major test cases for a genomic medicine paradigm, given that all cancer is caused by genomic instability, and microarrays allow assessment of patients' entire expressed genomes. . . . Success will require planning robust prospective trials, analysing health care economic and outcome data, assuaging insurance and privacy concerns, developing health delivery models that are commercially viable and scaling up to meet the needs of the whole population.Footnote 31

The genetic/genomic information being collected and stored by computers holds the possibility of better healthcare as well as the risk of being denied healthcare in some systems, including that of the United States. As genetic markers are decoded and designer drugs are developed, the possibility of linking details to an individual poses significant challenges. Access to essential healthcare may be denied by insurance companies that obtain the data, even if only in aggregate, and employers may similarly take adverse action. Contrarily, access to unwanted information may be forced upon those whose genetic markers provide more information than is desired. The well-known and putative “right to an open future” in genetics testing takes on added significance when large-scale computing resources generate patient and group-specific data—whether the patients or groups want it or not.

This problem is especially acute when data and databases are proprietary. We know very little about how to answer the question of whether ordinary sources of genomic data—patients and other people—care if their privacy is violated or they lose health benefits as a result of actions by a public health agency or by a for-profit corporation. An intuitively plausible and testable hypothesis is that ordinary people are more likely to oppose the latter. Consent for research, for instance, is rarely but increasingly informed by disclosure of who or what will profit from the research. Our hypothesis is that individuals will tend to support a study in molecular epidemiology more than one that generates proprietary results to be used to develop expensive designer drugs. The moral success of the bioinformatics, pharmacogenomics, and systems biology revolutions might very well depend on how and how well we address intellectual property issues.Footnote 32 This is another domain urgently in need of detailed and systematic inquiry.

Corporate Social Responsibility

To be sure, businesses have duties to shareholders as well as employees, to society as well as individual customers. It would be irresponsible—and at least potentially unethical—for a business not to attend to this basic tenet of success. The challenge of managing role conflicts and multiple loci of responsibility is well established as a nexus of thorny issues in business ethics.

Accordingly, is it good business to refuse to hire or to terminate people with costly genetically mediated maladies deduced or otherwise inferred from increasingly fine-grained gene expression analyses stored in large and perhaps proprietary databases? Insurance companies, which are in the business of risk stratification, might argue that it would be an irresponsible breach of fiduciary duties to shareholders not to take person- and group-specific genetic data into account in making personnel decisions. Indeed, insurance and other companies are in some respects in the business of discrimination, at least on the basis of age (consider auto insurance for young males), behavior (history of alcohol and drug use), occupation (skydivers), and so on.

Is there something different about these decisions and one based on genetics and affecting basic opportunities such as employment and insurance? A template is available by considering what is responsibly required versus what might be legally required. For example, until the enactment of the Genetic Information Nondiscrimination Act of 2008, U.S. law did not specifically address genetic discrimination by employers, despite many opportunities for Congress to have enacted appropriate laws.Footnote 33 The Americans with Disabilities Act might provide some protection against those who have an impairment, have a record of an impairment, or are regarded as having one. ADA language does not on its face address genetic maladies or predispositions to them—so it might or might not. And, as above, although HIPAA offers protection against discrimination on the basis of genetic information by health insurers, it does not reach down to the level of the affected individual.

Yet protections against genetic discrimination in employment exist in many individual states as well as in the European Union, a sign of the existence of some consensus that health rights and business ethics can be aligned. The inquiry is further complicated by the very nature of access to healthcare. The general consensus that such access is a right has led to widespread opposition to laissez faire capitalism in health markets across the globe. Pharmacogenomics and bioinformatics will force this issue again and again as society attempts to find stasis between market-based imperatives and the idea that it is morally offensive to profit excessively from the misfortunes of sick people. Corporate social responsibility requires an examination of accepted metrics that allow businesses to do business while simultaneously demanding that the open society that fostered them must also get its due.Footnote 34

Transparency and Accountability

Going beyond the preliminary analysis of the responsibility of business in the larger social framework, consideration of other familiar tools and principles of business ethics will also serve us well here. Transparency, veracity, and accountability, hallmarks of responsible business practices, must become part of the bioinformatics calculus as well. At the least, imagine a system in which businesses made public their policies regarding hiring, insuring, and firing based on genomic data; database owners opted for an open source stance to intellectual property questions; and scientists and clinicians who enable or perform microarray analyses disclose their interests—all their interests—in particular algorithms, companies, or maladies.

Just as public/private partnerships to develop genomic prowess are being reported almost daily, the hallmarks of business ethics must become part of the analysis. Against a background of corporate greed and malfeasance, a renewed focus on transparency and accountability is or seems to be permeating the business world. In response to widespread scandals, Congress mandated that accurate accounting and financial disclosures be certified by the very top level of management. Although the Sarbanes-Oxley Act of 2002 (Pub. L. No. 107-204, 116 Stat. 745) applies only to publicly held corporations, the premise upon which it is based has broad reach, and adherence to both its letter and spirit have become synonymous with best practices in business today, not only in the private sector but in the nonprofit world as well.Footnote 35

At the intersection of bioethics and business ethics, microarray technology and its successors and accoutrements present enormous public policy and ethical challenges. The demands of fairness to investors, patients, and society suggest that any system that benefits only a privileged few is inadequate and unsustainable. It would be morally preposterous to fail to seek to realize the potential of bioinformatics, and it would be morally indefensible to pursue these benefits without regard to the patient- and society-centered values that make the search worth conducting in the first place.

Footnotes

6
7
6
6
19

References

1 Szolovits P, Pauker SG. Computers and clinical decision making: Whether, how much, and for whom? Proceedings of the IEEE 1979;67:1224–6; Miller RA, Schaffner KF, Meisel A. Ethical and legal issues related to the use of computer programs in clinical medicine. Annals of Internal Medicine 1985;102:529–36; de Dombal FT. Ethical considerations concerning computers in medicine in the 1980s. Journal of Medical Ethics 1987;13:179–84; Goodman KW, ed. Ethics, Computing and Medicine: Informatics and the Transformation of Health Care. New York and Cambridge: Cambridge University Press; 1998.

2 Miller RA. Why the standard view is standard: People, not machines, understand patients’ problems. Journal of Medicine and Philosophy 1990;15:581–91.

3 Goodman KW, Miller RA. Ethics and health informatics: Users, standards, and outcomes. In: Shortliffe EH, Cimino JJ, eds. Biomedical Informatics: Computer Applications in Health Care and Biomedicine, 3rd ed. New York: Springer; 2006:379–402.

4 Alpert SA. Health care information: Access, confidentiality, and good practice. In: Goodman KW, ed. Ethics, Computing, and Medicine: Informatics and the Transformation of Health Care. New York and Cambridge: Cambridge University Press; 1998:75–101.

5 Here we defer an account of the debate over “genetic exceptionalism” or the contestable idea that genetic information is or should be regarded as different than other kinds of health data. At ground, those who argue against genetic exceptionalism contend that assigning special status to genetic information perpetuates bias and discrimination. This parallels the debate over HIV exceptionalism. See Juengst ET. FACE facts: Why human genetics will always provoke bioethics. Journal of Law, Medicine and Ethics 2004;32(2):267–75.

6 Goodman KW. Ethics, genomics, and information retrieval. Computers in Biology and Medicine 1996;26:223–9; Tavani HT. Ethics at the intersection of computing and genomics. In: Tavani HT, ed. Ethics, Computing and Genomics. Boston: Jones and Bartlett; 2006:5–26; Goodman KW. Bioinformatics: Challenges at the frontier. In: Tavani HT, ed. Ethics, Computing and Genomics. Boston: Jones and Bartlett; 2006:317–21; De Bouvet A, Deschamps C, Boitte P, Boury D. Bioinformatics: The philosophical and ethical issues at stake in a new modality of research practices. Medicine, Health Care and Philosophy 2006;9:201–9.

7 See note 6, Goodman 2006:317. A fuller definition would expand on each of these properties; for instance, “analyze” is vague and ranges across a number of discrete functions and tasks. A number of different definitions have been proposed. Compare Fenstermacher D. Introduction to bioinformatics. Journal of the American Society for Information Science and Technology 2005;56:440–6.

8 See, for instance, Alpert SA. Privacy and the analysis of stored tissues. In: Research Involving Human Biological Materials: Ethical Issues and Policy Guidance, Volume II: Commissioned Papers. Rockville, MD: National Bioethics Advisory Commission; 2000:A1–A36.

9 Goodman KW. Ethics, information technology and public health: Duties and challenges in computational epidemiology. In: O'Carroll PW, Yasnoff WA, Ward ME, Ripp LH, Martin EL, eds. Public Health Informatics and Information Systems. New York: Springer-Verlag; 2003:251–66.

10 Goodman KW. Moral foundations of data mining. In: Wang J, ed. Encyclopedia of Data Mining. Hershey, PA: IDEA Group Reference; 2006:832–36.

11 Privacy International. Databases and evidence, 8 Sep 2004; available at http://www.privacyinternational.org/index.shtml (accessed 31 May 2008).

12 Arnason V. Coding and consent: Moral challenges of the database project in Iceland. Bioethics 2004;18(1):27–49; Merz JF, McGee GE, Sankar P. “Iceland Inc.”? On the ethics of commercial population genomics. Social Science and Medicine 2004;58(6):1201–9.

13 Here is how GWAS are described at the government site, http://www.genome.gov/20019523:

To carry out a genome-wide association study, researchers use two groups of participants: people with the disease being studied and similar people without the disease. . . . Each person's complete set of DNA, or genome, is then purified from the blood or cells, placed on tiny chips and scanned on automated laboratory machines. The machines quickly survey each participant's genome for strategically selected markers of genetic variation, which are called single nucleotide polymorphisms, or SNPs. . . . If certain genetic variations are found to be significantly more frequent in people with the disease compared to people without disease, the variations are said to be “associated” with the disease. The associated genetic variations can serve as powerful pointers to the region of the human genome where the disease-causing problem resides.

14 See note 7, Fenstermacher 2005:441.

15 Marturano A, Chadwick R. How the role of computing is driving new genetics’ public policy. Ethics and Information Technology 2004;6:43–53.

16 Hayden EC. Personalized genomes go mainstream: Companies prepare to offer a very personal service. Nature News, 30 Oct 2007; available at http://www.nature.com/news/2007/071030/full/450011a.html (subscription required) (accessed 31 May 2008).

17 See note 6, Goodman 2006.

18 Kafatos FC, Eisner T. Unification in the century of biology. Science 2004;303:1257.

19 Swen JJ, Huizinga TW, Gelderblom H, de Vries EG, Assendelft WJ, Kirchheiner J, Guchelaar HJ. Translating pharmacogenomics: Challenges on the road to the clinic. PLoS Medicine 2007;4(8):e209.

20 But compare Grossman I. Routine pharmacogenetic testing in clinical practice: Dream or reality? Pharmacogenomics 2007;10:1449–59.

21 Jurisica I, Wigle DA, Wong B. Cancer Informatics in the Post Genomic Era: Toward Information-Based Medicine. New York: Springer; 2007:4. Note the authors’ assessment that cancer research using high-throughput genome analysis tools has increased nearly 20 times in less than a decade.

22 The HapMap web site is a masterpiece of accessible science and introductory ethics: http://www.hapmap.org/index.html.en.

23 See note 6, Goodman 1996.

24 See, for instance, http://www.hapmap.org/ethicalconcerns.html.en:

Since the samples include no personal identifiers, the risk to individual donors that privacy might be compromised is minimal. However, because each sample is identified as coming from a particular population, . . . group stigmatization and discrimination may occur from studies that use the HapMap. For example, researchers may find that a genetic variant associated with a higher-than-average risk of a disease is more common in one population than another. This information may be misinterpreted to mean that every member of a group has a higher-than-average risk of the disease, even though the higher risk may apply only to those individuals, inside the group or out, who have that variant. . . . Also, genetic findings could undermine established cultural or religious traditions or legal or political status. Many groups have firm beliefs about the origin of the group or about the relationship of the group to other groups, and these beliefs may be challenged by findings built on the use of the HapMap. In addition, genetic findings may conflict with the social and cultural methods that groups have developed to determine who is a member of that group.

25 Court MH. A pharmacogenomics primer. Journal of Clinical Pharmacology 2007;47:1087–103; McCarthy AD, Kennedy JL, Middleton LT. Pharmacogenetics in drug development. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 2005;360:1579–88.

26 See note 19, Swen et al. 2007.

27 Phillips KA. The intersection of biotechnology and pharmacogenomics: Health policy implications. Health Affairs 2006;25:1271–80.

28 Malinowski M, Rao R. Legal limitations on genetic research and the commercialization of its results. American Journal of Comparative Law 2006;54:45–65.

29 Evans JP. Health care in the age of genetic medicine. JAMA 2007;298:2670–72.

30 Foster MW, Mulvihill JJ, Sharp RR. Investments in cancer genomics: Who benefits and who decides? American Journal of Public Health 2006;96:1960–64.

31 Willard HF, Angrist M, Ginsburg GS. Genomic medicine: genetic variation and its impact on the future of health care. Philosophical Transactions of the Royal Society of London. Series B, Biological sciences 2005;360:1543–50.

32 D'Souza S. Gene meets machine: Intellectual property issues in bioinformatics. Health Law Review 2004;12(2):34–43.

33 Moyer TJ, Anway SP. Biotechnology and the bar: A response to the growing divide between science and the legal environment. Berkeley Technology Law Journal 2007;22:671–731.

34 Werther WB, Chandler D. Strategic Corporate Social Responsibility: Stakeholders in a Global Environment. Thousand Oaks, CA: Sage Publications Inc.; 2006.

35 McLaughlin N. Welcome to accountability. Modern Healthcare, 11 Jun 2007, Issue 24:19.