Bioethics, Bench, and Bar:
Selected Arguments in
Landry v. Attorney General
© 2000 DH Kaye. A final version of this paper appeared in Jurimetrics: The Journal of Law, Science, and Technology, vol. 40, 2000, pp. 193-216.
Abstract: In Landry v. Attorney General, the Massachusetts Supreme Judicial Court upheld a statute that requires individuals convicted of a wide range of felonies to submit to the extraction of samples of their DNA for the analysis of individualizing features and for the inclusion of that data in a computerized database. Various organizations submitted amicus briefs to help the court understand the underlying science or technology or to appreciate the bioethical issues in using the data or samples in subsequent research. This article reviews portions of two of these briefs for their accuracy and completeness. It concludes that they are no less adversarial than those of the parties. It suggests that the arguments about invasions of a right to genetic privacy suffer in the translation from medical genetics to law enforcement identification databases. It also contends that whether research uses of data or samples should be allowed without the consent of the offenders is a question of public policy that cannot be resolved by absolute and sweeping claims that information on people can never be used without their consent. It urges that the developing norms of research on human subjects be examined with greater recognition of the differences between clinical or research uses of genetic data and law enforcement databanking.
Citation: D.H. Kaye, Bioethics, Bench, and Bar: Selected Arguments in Landry v. Attorney General, 40 Jurimetrics J. __ (2000)
Landry v. Attorney General (1) is an exceptional case. It is the only instance in which a court invalidated a statute establishing a system of DNA databanking for convicted offenders. In 1998, a Massachusetts Superior Court determined that compelling felons to submit to blood sampling violated the Fourth Amendment's prohibition on unreasonable searches and seizures. (2) The court reached this conclusion on particularly unforgiving grounds. Observing that the United States Supreme Court never has upheld a physical intrusion into the body to uncover evidence of a crime without "at least individualized suspicion, if not probable cause, of criminal activity and a clear indication that such intrusion will provide probative evidence," (3) the court held that the Fourth Amendment leaves no room for balancing the interest in law enforcement against the interest in bodily integrity and informational privacy. It impounded 1,200 DNA samples and enjoined the state from enforcing the law. (4)
The decision generated consternation among law enforcement officials (5) and jubilation among the opponents of DNA databanking. (6) As the Supreme Judicial Court of Massachusetts prepared to review the decision, law professors and students at Boston College, Harvard University, and the Illinois Institute of Technology submitted briefs as amici curiae. (7) Two of these briefs focused on bioethical concerns. (8) The Boston College group argued that the statute violates international law because it permits medical experimentation on prisoners without their consent. (9) The Illinois Institute of Technology's Institute for Science, Law, and Technology argued that modern genetic technology and the range of possible uses of the samples allowed under the statute contravened "basic ethical principles" and constituted an unconstitutional invasion of privacy. (10)
The Massachusetts Supreme Judicial Court vacated the preliminary injunction. It made short shrift of the Superior Court's reading of the Fourth Amendment as a per se rule against taking blood without probable cause. (11) Although the supreme court purported not to rely the "special needs" exception to the probable cause and warrant requirements of the Fourth Amendment, it adopted the view that "the high government interest in a particularly reliable form of identification outweighs the minimal intrusion of a pin prick" (12) and it followed the U.S. Court of Appeals for the Ninth Circuit in reasoning that "[o]nce a person is convicted of [a serious crime], his identity has become a matter of state interest and he has lost any legitimate expectation of privacy in the identifying information derived from the blood sampling." (13)
In addition, the supreme court brushed aside the description of the dangers of the DNA technology that the IIT brief provided and the bioethical concerns that the two amici emphasized. In the court's view, the statute was suitably circumspect and protective of privacy, and the provisions that the amici found "reprehensible" (14) and violative of "accepted human rights standards" (15) posed no imminent or realistic dangers. (16)
This article considers the arguments about ethics and technology propounded by the amici. Although the court's interpretation of the Massachusetts law makes it far less of a threat to privacy than the briefs suggested it was, that law is but one of 51 DNA databanking statutes enacted in this country; (17) as science and technology evolve, there will be pressure to amend and extend the statutes to give law enforcement authorities an ever more powerful investigative tool. As the data banks grow in conjunction with knowledge of how the 100,000 or so human genes function, it is possible that biomedical researchers will request access to these repositories of human DNA. At least, this is one fear that animated the briefs and to which critics of DNA databanking have appealed in their efforts to mobilize public opinion against the practice. Consequently, it is appropriate to examine the extent to which the norms governing experimentation with human subjects might apply to research with DNA samples or data derived from convicted offenders, and the Landry briefs are useful for this purpose. In addition, the briefs indicate the perils of ignoring context and nuance in applying bioethical principles that developed in one setting to a distinctive setting. For all these reasons, the arguments propounded by the amici in Landry and the manner in which they were advanced deserve attention.
Part I of this article critiques the IIT Brief's view of "the evolution of DNA technology." (18) It shows how the brief obscures the science and technology of forensic DNA typing, making the procedure seem much more invasive than it actually is. Part II isolates two arguments intended to show that the Massachusetts law invades genetic privacy. It reveals how certain privacy concerns that are important in genetic counseling are only marginally implicated in the law enforcement context. Part III turns to the argument that the norms of medical research and international human rights law make it impermissible to use biometric information or tissue samples without the specific consent of the individuals who provided the data or samples. It suggests that these claims are, at best, highly controversial, and it identifies some of the factors that must be considered to reach a well reasoned conclusion as to the ethical propriety of various kinds of research on DNA samples taken from convicted offenders.
I. The Science and Technology of DNA Databanking
Because the possible threat to privacy posed by criminal DNA databases depends on the genetic information stored in them, an assessment of the reality and extent of the threat requires an understanding of some basic aspects of human genetics. (19) By informing courts and legislatures of the essentials of the science and technology that must be understood to formulate, interpret, and evaluate law that involves science and technology, amicus briefs from respected researchers can serve an important function. Scholars who undertake this task, however, bear a heavy burden. Unless they forthrightly abandon their scholarly role in favor of advocacy, they should offer a clear, accurate, and balanced presentation. If they put themselves forward as experts rather than partisans, they should, to paraphrase the opinion of the Supreme Court in Kumho Tire Co., Ltd., v. Carmichael, (20) adhere to the same standards of accuracy that they would follow in their academic research. (21)
To fulfill this aspiration, a scholarly brief that offers to assist the court in understanding the technology involved in DNA databanking would employ scientific terminology correctly to explain the theory that supports DNA identification. (22) It would describe the type of DNA polymorphisms used in forensic work, the types used to construct DNA databases, and assess current information on the validity and reliability of the identification process. It would distinguish between loci that are related to diseases that are potentially stigmatizing and loci that are without social significance, and it would indicate that the scientific community believes that the loci currently used in constructing databases contain no genes and are of limited medical relevance.
On the surface, the IIT Brief seems to accept this challenge. It makes unusual claims of expertise and objectivity, (23) and it prefaces the legal argument with a description of "the evolution of DNA technology." (24) Unfortunately, the description of the science and technology seems more tenebrific than edifying. Because not all DNA profiles pose the same threat to personal privacy, it may be worthwhile to clarify the nature of the genetic information stored in DNA databases.
A. Is There "a DNA Profile for Identification and Genetic Health Information"?
The IIT Brief speaks of "a DNA profile for identification and genetic health information" (25) as if one "profile" fits all uses. According to the IIT researchers, "differing gene combinations are called alleles," and "[t]he sequence of alleles is what is used to create a DNA profile for identification and genetic health information." (26) In this way, the brief blurs together all DNA polymorphisms as due to variations in "a particular gene" or "gene combinations." (27) Of course, in classical genetics an "allele" is merely the form of a particular gene that codes for a protein and thus controls an observable trait. (28) But only some 3% of the human genome consists of genes; 97% is noncoding. Confusing the alleles of genes with the "alleles" of noncoding loci (29) and intimating that a single "DNA profile" is used both for "identification and genetic health information" makes forensic DNA profiling seem to reveal far more "health information" than it actually does.
B. Is RFLP Testing Less Invasive of Privacy than Forensic STR Testing?
After positing "a DNA profile for identification and genetic health information," the IIT scholars attribute "the potential for error where some individuals may be wrongly implicated for a crime and others may be wrongly exculpated" to "the rapid introduction of ever-changing types of DNA forensic techniques." (30) This view of "the evolution of DNA technology" seems perverse. If anything, the introduction of new techniques has led to more reliable and less subjective determinations. The causes of false matches and false exclusions probably have more to do with sample handling than with the loci employed or the methods used to characterize them. (31)
The expository problems multiply with the description of the difference between RFLP testing and PCR-based testing. The former technology typically employs gel electrophoresis to separate large restriction fragments by length. Restriction fragments are snippets of DNA produced by subjecting long DNA molecules to a bacterial enzyme (called a restriction enzyme). The enzyme cuts the molecules at sites where a particular, short sequence of bases is present. The lengths of the fragments thus reflect the distances between two restriction sites. A small fraction of the many fragments are detected--those that contain a particular sequence of bases that bind to those in a "probe"--a short, radioactively tagged strand of DNA with the sequence of interest. (32) Depending on the enzyme and probe combination, the procedure can detect alleles in a variety of genetic systems, (33) and the field of DNA diagnostics in medicine began with RFLP analysis. (34)
PCR-based testing uses the "polymerase chain reaction" to make millions of copies of a short segment of DNA, and a variety of analytical methods then can be used to discern different features of the collection of identical fragments. PCR-based tests have come to dominate forensic work because they are faster and less expensive than RFLP testing, (35) and they can be performed on much smaller samples of DNA. (36) Moreover, with the genetic systems used in PCR-based tests, determining whether two samples "match" often is easier than it is with RFLPs.
Yet, the brief makes it sound as if the transition to PCR-based systems is insidious. It claims that "RFLP analysis . . . did not provide a means of assessing an individual's current or future health profile," (37) whereas the newer "forensic technologies . . . will allow the generation and disclosure of sensitive information beyond forensic identifying information." (38) The reality is otherwise. Both RFLP and PCR-based testing can produce considerable health-related information, depending on the loci that are chosen for analysis. The particular loci that are rapidly becoming the polymorphisms of choice for DNA databases for identification are a set of 13 STR loci (39) on autosomal chromosomes. (40) The brief criticizes the use of STR loci because "[t]he STR testing method can be performed on both the X and Y chromosomes, both sources providing a wealth of genetic information on the subjects of the tests and on their relatives." (41) This statement is troublesome for two reasons. Not only can any type of genetic analysis (including RFLP testing) be used to provide information about health-related loci on the X and Y chromosomes, (42) but also tests on other chromosomes can yield information on relatives. (43) The brief adds that "STR testing can gather information regarding [the] genetic basis of metabolism and hormone receptor functions." (44) Again, however, even more biochemical information might be acquired with other loci, including RFLPs. (45)
Finally, the brief states that mitochondrial DNA sequencing is a "critical development" (46) because "mitochondrial DNA can be used to predict diseases in the individual and inter-generationally." (47) Yet, the incremental threat to privacy is trivial. Mitochondria are small structures found outside the cell nucleus. The mitochondrial genome is an utterly negligible fraction of the human genome (48) with a limited function. (49) No mitochondrial loci are included in convicted-offender databases. Indeed, mitochondrial DNA sequencing rarely is used even in case investigations, and the part that is used for identification (the D-loop) is not of medical interest.
The relevant scientific fact, then, is that whatever technology may be used to detect particular DNA polymorphisms, some polymorphisms will be medically significant, but many will not be. The significant issue for privacy protection is which loci are included in the profile. Of course, as more and more loci associated with diseases are discovered and characterized, the amount of medically relevant information that could be derived from a DNA sample will grow. In this sense, the brief correctly recognizes that "developments in DNA technology" pose more of a risk to privacy than any particular set of RFLP tests of noncoding VNTR loci. (50)
The Massachusetts Supreme Judicial Court did not adopt the IIT position that STR loci are especially dangerous. To the contrary, the court remarked that "[t]he most notable feature of STR testing is that STRs found on a DNA strand are likely not to be associated with any known genetic coding function." (51) But this could be as much a misconception as the opposite view. (52) That an allele is noncoding does not mean that it is not associated with a coding allele. Nevertheless, the particular 13 loci that are used in Massachusetts and other states that desire compatibility with the national database maintained by the FBI are generally believed not to be predictive of any diseases or stigmatizing conditions. (53)
The IIT Brief not only portrays STR loci and mitochondrial sequencing as particularly disturbing developments, but it also suggests that STR typing is prone to "twenty problems" in forensic applications that are exacerbated by "primitive" and "appalling laboratory conditions across the country." (54) The result, it maintains, is that "Massachusetts' utilization of a databank will inevitably lead to an unacceptable rate of errors . . . ." (55) The evidence cited in support of this pessimistic assessment is varied--and problematic. It includes (1) a remark from the eccentric Kary Mullis about the inability of "[p]eople hired to follow written instructions on the boxes of DNA investigation kits" to "know the chaff from the wheat"; (56) (2) observations by Eric Lander to the effect that "DNA artifacts [sic] collected at crime scenes might be degraded, highly contaminated, or even 'mixtures of samples from different individuals'"; (57) (3) the interpretation from Jay Koehler of RFLP testing in a simulated case involving six samples (one of which was a mixture) submitted to 45 laboratories in 1993; (58) and (4) a newspaper story on a serologist's perjury and falsification of laboratory records. (59)
In sum, indicating the dangers in a technology is especially important for an amicus whose very purpose is "to assess the appropriate uses of technologies," (60) but the task demands careful analysis of the particular application of the technology to determine the implications of those dangers for "public debate and policy analysis." (61) Despite the problems in the IIT group's description of the technology, the brief raises an important point--our increasing knowledge of human genetics gives the state the power to test for characteristics that most people would rather keep out of the hands of the government. The next section therefore examines the genetic privacy arguments in the IIT and the Boston College briefs. It suggests that a number of these arguments, which have undeniable force in the context of clinical genetics, are not well tailored to the context of law enforcement DNA databanking.
II. Genetic Privacy Arguments
The IIT brief presents a series of arguments to show that the Massachusetts law invades genetic privacy--not just of the convicted offenders whose DNA is sampled, stored, and analyzed, but also of "'innocent' third parties." (62) Section A analyzes one privacy argument pertaining to the offenders who must provide DNA samples. Section B turns to an argument about the privacy rights of relatives of the offenders. These discussions show how privacy concerns that are important in genetic counseling are only marginally implicated in the law enforcement context.
A. Searching Disease-Linked Loci to Track Suspects by Pharmacy Records
Reflecting the view that statutes must require genetic information to be kept confidential, the IIT Brief complains that the statute does not exclude the possibility that
[l]aw enforcement officials may use genetic information about an offender, such as genetically based diabetes or schizophrenia, to track a suspect. They may pour [sic] through pharmacy records, for instance, to monitor the suspect's movements. That information, in turn, eventually may be retrieved by third parties such as insurance companies. Indeed, the offender may not even know his or her own genetic traits prior to the investigation. (63)
This passage evinces a fixation on arguments that may (or may not have) have merit in some contexts, (64) but suffer in the translation to law enforcement databases. Thus, one plausible aspect of "genetic privacy" is that an individual has the right not to know that he has particular alleles. For example, genetic counselors should not be able to force individuals to undergo testing so that they can discover whether they carry the allele for Huntington's disease. Some individuals with a family history that puts them at risk may prefer a life with ambiguity to one in which they know that they are destined to an early death. But the offender's interest in maintaining ignorance of his genetic traits cannot be the basis for Institute's privacy argument. If the offender is buying insulin, he knows that he has diabetes. If he does not have diabetes but could be predicted to develop the condition, the chance that he would learn of this possibility through the hypothetical genetic detective work is attenuated. Will the police really want to screen for disease-linked genes in stored DNA samples from convicted offenders who become suspects in the hope that they can find a pharmacist who sold a prescription drug to the offender in order to ascertain where the suspect was when the sale was made? Surely there are easier ways to discover an offender's whereabouts than to "track" him by his purchases of insulin.
However, the brief may be alluding to a different concern--that the offender will have difficulty securing health insurance because the police investigation will reveal to an insurance company that he has an as-yet unmanifested disease. This, too, seems farfetched. First, as just noted, this type of genetic detective work is itself unlikely to have any appeal to investigators. Second, like the police, insurers may not care about the disease-linked genes. To date, there are no competent indications that health insurers are especially interested in such information, and there are good reasons to believe that they are not chomping at the bit to institute genetic testing. (65) Third, even if insurers were motivated and allowed to use DNA test results, how would they learn about them? The Massachusetts statute does not authorize their release to insurers and imposes criminal penalties for both the unauthorized disclosure and wilful receipt of these records. (66)
B. Privacy of Relatives
The IIT Brief's argument about the privacy of relatives provides another example of the dubious extension of a bona fide issue of confidentiality in genetic counseling. When genetic testing at disease-related loci indicates a substantial chance that a relative is at risk for a serious disease for which an early intervention would be advisable, difficult problems of confidentiality as well as possible liability for failing to warn the relatives arise. (67) On the one hand, the relatives have not consented to the testing and might not wish to be informed; on the other, they might well want to know something of the risk and the options that they face.
The situation is much clearer in the criminal justice context. The IIT group speculates that
prosecutors may gather information about the offender's close relatives in certain circumstances. For example, law enforcement officials may have found a DNA artifact [sic] at the scene of a crime but have been unable to compare this artifact [sic] to that of their prime suspect because the suspect has fled and his or her whereabouts are unknown. In this kind of situation, law enforcement officials may argue that it is necessary to compare the DNA of this prime suspect's close biological relatives to that of the DNA artifact [sic] from the crime scene to determine whether to continue or cease pursuit of the suspect. (68)
Yet, the brief offers no thoughts as to why the relatives would have a right to be free from "this level of genomic intrusion," (69) and the existence of any such right is implausible. To establish this result, we can begin by simplifying and clarifying the case that the IIT scholars have contrived. The "prime suspect" P is a fugitive. His identical twin Q is a convicted offender whose DNA already has been genotyped at the 13 STR loci that are of no medical significance. The police want to know whether the trace evidence DNA might have come from P. What right does P have to prevent the police from looking back at Q's types? The answer must be "None at all." The police may query the entire database to see if any record--including Q's-- matches the trace evidence profile. Once they determine that Q's matches, they violate no right of P's by deciding to continue their investigation of P. P has the misfortune of being the twin brother of a man whose encounters with the criminal justice system have generated evidence against P. He does not have the right to suppress that evidence or to bar the police from obtaining it.
This conclusion seems perfectly clear when the evidence is not genetic. Suppose that Q had been picked up by the police for questioning about a reported rape, and he had revealed that P was the rapist. As long as Q's detention and interrogation are legitimate, his statements are admissible against P. The result does not change if the lead from Q to P results from a genetic characteristic rather than a narrative. Suppose that the rape victim studying mugshots states that the rapist is the man whose face is shown in a photograph of Q, but Q was in prison at the time of the rape. The police know that P is Q's twin brother, so they focus their investigation on P. Could P object to his becoming a target of the investigation because he did nothing that caused Q's picture to be in the police files?
As with the narrative and the mugshot, if the DNA sample is obtained properly, a relative has no plausible right to prohibit its use as an investigative lead. The concern for the privacy of relatives in genetic counseling must not be ignored, but it applies very differently in the law enforcement context.
III. The Human Subjects Arguments
Both the IIT and the Boston College briefs make much of the fact that the statute does not explicitly prohibit the profiles and samples from being used for various kinds of research. According to the IIT scholars, any "research" use of the DNA samples is proscribed by "basic ethical principles . . . such as" the Nuremberg Code and federal regulations, which "mandate that research not be undertaken without the subject's advance, voluntary, informed consent." (70) The Boston College students and faculty take this argument a step further, insisting that the perceived transgression of the Nuremberg Code violates customary international law, the International Covenant on Civil and Political Rights, (71) and the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment. (72) However, as explained below, the meaning and application of the "basic ethical principles" to the law enforcement databases and databanks is far more complex than the briefs would lead one to imagine.
A. The Statutory Authorization for Research
The briefs presume that the Massachusetts law countenances the use of DNA samples for medical research. But why should we assume this? The statute distinguishes between "DNA records" and "DNA samples." A sample is "biological evidence of any nature that is utilized to conduct DNA analysis." (73) A record is "DNA information that is derived from a DNA sample and DNA analysis . . . ." (74) The statute specifies the permissible uses for the records, (75) and it penalizes all unauthorized uses. (76) In particular, the statute gives the director of the state police laboratory discretion to
make DNA records available . . . for the limited purpose of (1) advancing DNA analysis methods and supporting statistical interpretation of DNA analysis, including development of population databases; provided, however, that personal identifying information shall be removed from DNA records . . . and (4) advancing other humanitarian purposes. (77)
Oddly, the law contains no corresponding list of the authorized uses for samples and imposes no penalty for their misuse. (78) Should this gap be interpreted to mean that state employees are free to use and disseminate the samples for any purpose, including medical research? Or might it mean exactly the opposite--that the legislature assumed that the samples would not be used for any purpose except to produce the records and therefore found it unnecessary to develop a list of allowable uses for the samples? The Boston College Brief ignores the issue, asserting without analysis or explanation, that "[t]he Act envisions that the DNA will be disseminated to a number of different entities" for the same uses that could be made of the records. (79) Similarly, the IIT Brief seems to assume that whatever can be done with records can be done with samples. (80)
The Supreme Judicial Court adopted the opposite interpretation. It construed the authorization to disclose the records for certain purposes combined with lack of any such authorization for disclosure of the sample as a prohibition on any release of the samples. (81) In addition, the court noted that under the regulations promulgated by the state crime laboratory, even the STR profiles could not be released for "humanitarian purposes." (82) The court concluded:
The plaintiffs assert that [a section] of the Act, which compels disclosure of the records to comply with Federal statutory or grant obligations, and allows disclosure of records for various scientific or humanitarian purposes, may somehow lead to leakage of complete genetic profiles. Their speculation that data may be used wrongfully is contrary to the language of the Act and concerns provisions that are not before us. (83)
Although the court's disposition may have sufficed to sustain the Massachusetts law against the attack that it was unconstitutional on its face, efforts to advance human knowledge with the records in offender databases and samples and offender DNA banks may yet materialize. Should these be seen as Frankensteinian experiments that use offenders "as genetic guinea pigs in the interests of furthering 'science'" (84) and rejected as "a particular category of cruel, inhuman or degrading treatment or punishment" (85)? It is time to consider these claims of the amici, even if they were inapplicable to the statute under attack. We begin with a brief review of medical experiments that clearly deserve these epithets.
B. The Nuremberg Precedent
The Nuremberg Code is a list of ten "basic principles" of ethical "medical experiments on human beings" promulgated by the United States military tribunal that found 15 of 23 Nazi doctors guilty of participating in "plans and enterprises involving medical experiments without the subjects' consent . . . in the course of which experiments the defendants committed murders, brutalities, cruelties, tortures, atrocities, and other inhuman acts" (86) on known, identified, and unwilling victims. Among other things, they injected concentration camp inmates with yellow fever, smallpox, typhus, cholera, diptheria, and spotted fever to test vaccines; they immersed inmates in freezing water until they died to test methods for treating hypothermia; they transplanted sections of bones from one inmate to another; and they fed inmates poisons or shot them with poisoned bullets.
The Code obviously goes beyond these atrocities, (87) and its first principle is that "[t]he voluntary consent of the human subject is absolutely essential." But the Code speaks in terms that presume that a specific human being is the subject of an investigation into the treatment of a disease or comparable condition, (88) and in applying this Code, one should not lose sight of the evils it was intended to combat. Whether the same consent is essential for gathering other sorts of data on people or for using preexisting information is not immediately obvious. Consider the following research:
(1) The Internal Revenue Service selects taxpayers to audit as part of a study to find information on individual income tax returns that could be used to make statistically valid predictions of who might be filing returns that falsely report income, exemptions, or deductions. The Service does not ask the taxpayers to consent to the research.
(2) Researchers use a law school's records of its students' undergraduate grade point averages to see whether these grades are correlated to performance in law school. The researchers do not ask the students for their consent.
(3) Scientists use fingerprint records taken when people are arrested for crimes to determine how variable fingerprints actually are and to develop and validate a computerized system of fingerprint identification. They do not contact the arrested individuals to secure their consent.
(4) Geneticists use tissue samples from pathology repositories to investigate whether a gene is associated with a disease. The sources of the samples consented to the removal of the samples and to their use in medical education and research generally, but they did not consent to this specific use.
(5) Physicians seeking data on the effects of psychedelic drugs administer LSD to soldiers without informing them of the experiments. (89)
Perhaps the Nuremberg Code proscribes all this research. Certainly, it applies to the fifth case. But its application to the other cases is indirect or unclear. The Code was offered to delineate "reasonably well-defined bounds" on "medical experiments on human beings" implicit in "the ethics of the medical profession." (90) Although the Code elaborates on the meaning of "voluntary consent," it contains no definition of "human subject" or "medical experiment." As one moves outside the historical context of nontherapeutic, harmful interventions on unwilling and uninformed subjects, the moral or legal status of research cannot be resolved by dogmatic assertions about the inexorable need for informed consent, (91) invocations of the horrors of the concentration camps, (92) or reflections on the study in the United States of the progression of syphilis in African-American men. (93) The events at the camps in Auschwitz, Buchenwald, Dauchau, Natzweiler, Ravensbrueck, and Sachsenhausen--and the legal response to them--are of vital importance. This history compels us to ask not only "How could that have happened?," but also "How much harm can be inflicted on human subjects of research for the sake of medical progress and national survival?" (94) Yet, the Nuremberg Code--important as it is--does not reveal the moral or legal status of research that poses no significant risk of any harm to any human being. If research with DNA records or samples from convicted offenders is of this type, then voluntariness may not be the talisman that is required to ward off evil. We turn, therefore, to the types of research that might be conducted with these materials.
C. Biometric DNA Research
1. Research on Anonymized Records
The Massachusetts statute provides that
The director may, in his discretion, make DNA records available to authorized persons or organizations, upon written or electronic request, for the limited purpose of . . . advancing DNA analysis methods and supporting statistical interpretation of DNA analysis, including development of population databases; provided, however, that personal identifying information shall be removed from DNA records disclosed for such purposes . . . . (95)
With no personal identifiers, the records with no names attached merely would form a list of at least 13 STR numbers per person. (96) The only research that might be done with this data set is statistical. For example, one could compute the relative frequency for each allele and use these frequencies to estimate the probability of a random match between a crime scene sample and an unrelated person. (97) One could perform tests of Hardy-Weinberg equilibrium at each locus. (98) For a set of records, one could compare each profile to every other profile to help test the theoretical claim that every STR profile is likely to be unique. (99)
This research is not unlike the dermatoglyphic research in case (3) of the previous section. If anything, it is even less problematic, for the researchers do not know the identities of sources of the data. Yet, the Boston College students and faculty seem to maintain that such research violates the demand for informed consent codified in the opinion of the Nuremberg tribunal. "The requirement of consent in the Nuremberg Code," they insist, "is absolute and is never tied in the text to a determination that the human subject will be harmed by the experiment." (100)
However, as we have just seen, whether the statistical use of records should be considered a "medical experiment" on "human subjects" within the scope of the Nuremberg Code is debatable, (101) and the unbounded, absolutist view of the brief that it is immoral to compute statistics from data on human subjects who do not freely assent to the particular use of the data seems difficult to defend. Consider a data set consisting of records of blood or breath alcohol concentrations taken from drivers suspected of driving while intoxicated. Most of these data, let us assume, have not been obtained with the voluntary consent of the drivers (especially if they were, in fact, intoxicated at the time), but rather under threat of having one's driving license suspended as provided for in the jurisdiction's "implied consent law." Would it be unethical for the sheriff's office to make these records available, in anonymized form, to a researcher to analyze the distribution of alcohol levels in the sample? This would seem to be as much a "medical experiment" as the statistical research allowed with anonymized DNA records, yet the suggestion that it is unethical for the reasons that the concentration camp experiments were immoral seems not only implausible, but morally blind.
2. Biometric Research on Anonymized Tissue Samples
The research just described is confined to the "records" of essentially meaningless, but distinctive numbers obtained from STR or other loci that are recorded for use in database searches. As the Landry court read the Massachusetts law, disclosure or use of the blood samples would not be allowed. But the samples could be useful for developing and validating new loci or analytical methods. Although a state might choose to conduct such research with other DNA sources, the use of offender data or samples is directly connected to operating and administering an offender DNA databanking system, and it surely is conceivable that legislators might contemplate the use of a convenient source of DNA for this research.
It is hard to see how forensic identification research of this type invades any right of the sources of the samples. The only possible risk of harm might be that a researcher would analyze disease- or behavior-related loci to which a valid expectation of privacy might attach. This risk might be minimized by requiring that the research be limited to other loci. If it were thought that researchers cannot be trusted to follow the approved research protocol, the samples could be anonymized before being delivered to the researchers.
However, the IIT Brief argues that "DNA samples can never sufficiently be made anonymous. The very premise of DNA forensic technology is that a blood sample can always reveal the identity of the person from whom it is withdrawn." (102) Perhaps there are some situations in which DNA samples cannot be made anonymous, (103) but in the context of forensic identification databases, the claim is fatuous. The premise of DNA databases is that a database is required to link unidentified samples to known individuals. No one can determine the source of an unknown sample without comparing it to a sample whose origin is known, and that is why the government establishes offender databases. If the researchers are not given access to the database, but only to anonymized samples, they cannot link the samples to their sources. (104)
Perhaps recognizing the weakness of its claim that anonymization is impossible, the IIT group also insists that research using anonymized samples would be unethical without the explicit consent of the sources of the samples. (105) To support this view, the brief relies on the judgment at Nuremberg (106) and provisions of the Code of Federal Regulations governing federally funded research. (107) But the claim that the Nuremberg Code prohibits research on legitimately collected, anonymized tissue samples is perilous for the same reason that the claim that it prohibits statistical research with biometric data derived from those samples is implausible: the research involves no intervention or experimentation on any identifiable human being.
Neither do the federal regulations (108) generally known as "the Common Rule" (109) clearly establish that the research would be unethical. Even if we indulge the assumption that the regulations codify a well formulated moral position, their application to anonymized tissue samples is immensely more controversial that the brief deigns to admit. The IIT scholars simply assert that "in most cases," "federal regulations governing the use of stored tissue samples . . . require the informed consent of the tissue sample donor before release of the sample to researchers." (110) The implication, of course, is that the Rule would require such consent from offenders before their samples could be used to improve the system of DNA databanking for forensic identification. But this conclusion is not obvious from the text of the Rule. The provision of the Common Rule pertaining to archival tissue samples states that "[r]esearch involving the collection or study of existing data, documents, records, pathological specimens, or diagnostic specimens" does not require consent or approval from an institutional review board "if these sources are publicly available or if the information is recorded by the investigator in such a manner that subjects cannot be identified, directly or through identifiers linked to the subjects." (111) This text is a powerful illustration of the evil of passive voice. Is the point of the exemption to enable the investigator to use samples legitimately collected in the past without attempting to contact the sources of the samples as long the investigator cannot use the available records to identify of the sources? Is the purpose to enable the samples to be used only if no one can trace the samples to their sources using the information recorded by the investigator? Or is it to preclude research with the samples only when no one in the universe can link the samples to their sources using the information recorded by the sample repository as well as the researcher? Only the rather extravagant, third interpretation of the exemption is consistent with the portrayal of the Rule in the IIT Brief.
Although I suspect that the first interpretation is the most plausible--and the only one in keeping with the practices of tissue repositories that were generally regarded as ethical when the regulations were drafted (112)--I shall not press that point. All I mean to demonstrate here is that the correct construction of the Common Rule is not apparent. As a result, a purportedly scholarly amicus brief that presents but one choice as if it were the obvious and uncontested meaning of the Common Rule is misleading. The historical fact is that although "it is widely accepted that informed consent must be obtained for the many projects that involve the direct prospective involvement of individual subjects," "[t]he role of informed consent has been much less clear for research that does not require such personal involvement but rather can be performed using tissue samples." (113) Indeed, it has been said that "[w]hether previously collected human biological materials may be used for newly conceived research has been the most controversial issue in research ethics for the past four years." (114) The brief blithely ignores this controversy, giving only a few of the arguments for the particular interpretation of the Common Rule offered in 1995 by a group of individuals said to be an "interdisciplinary panel of physicians, lawyers, and ethicists." (115)
D. The Real Issues
The difficulties in applying the Nuremberg Code and the Common Rule to traditional genetic research with tissue samples do not mean that all, or even any, research with banked offender DNA samples should be permitted. It merely means that the issue cannot be resolved by incantation of the necessity for informed consent. Even if the Nuremberg judgment and the federal regulations clearly proscribed new research with old samples obtained by clinicians or researchers, much more analysis would be required to decide whether the same limitations should apply to offender databanks. The obvious, but as yet unexplored fact, is that DNA samples in convicted-offender databanks are different from tissue samples stored in hospitals or research laboratories. A patient has a right, grounded in personal autonomy, to decline to donate tissues to medical research. (116) But where a patient can refuse to have the tissues removed at all, a convicted offender has no such choice--the law demands that he supply a sample of DNA. The samples can be acquired legitimately for a public purpose without regard to the desires or wishes of the "donors," because there is no moral or legal obligation to honor their wishes over the benefits that DNA databanking offers society. Unless a further research use of the samples poses some real risk of harm to the individuals who are compelled to donate their DNA, the usual foundation for the supposed need for consent is weakened. For example, one justification sometimes offered for a requirement of explicit consent for a research use of a voluntarily obtained tissue sample is that "[p]ublic commitment to obtaining consent for research promotes the willingness of people to seek medical care because patients can be reassured that they do not give up their right to decide whether to participate in research when they enter the health care system." (117) This rationale has no bearing on samples acquired outside of the health care system. Other justifications for a strict consent rule, such as "[d]emonstrating respect for persons," (118) carry weight in the law enforcement context, but ascertaining their boundaries and implications requires careful analysis rather than simple slogans.
If consent is not a categorical imperative for all research on DNA from offender samples-- just as it is not a sine qua non for obtaining the samples--then assessing the propriety of particular research projects becomes a more complex task. For instance, it has been suggested that "[c]ollections of DNA samples from criminals . . . are likely to be perceived as particularly rich research resources by those interested in studying genetic factors involved in anti-social or aggressive behavior" and that such studies "ultimately mischaracterize--and stigmatize--groups of people," thereby "legitimizing draconian 'medical' responses . . . like eugenic sterilization." (119) This is an argument about a harm that could flow from certain non-biometric research with offender samples. If it were sound, it would have implications both for requiring consent on the part of the individuals whose samples would be used for this particular research and who might be so harmed, and for the permissibility of such research regardless of consent. But it is a complex (and perhaps strained) argument that is much richer than the superficial claim that informed consent is necessary to use any archived samples. And, it raises much deeper questions: Should we regard scientific research as ethically suspect because valid findings might pose social dangers? Should government prohibit such research even if it is ethically acceptable out of fear that it will be misinterpreted or misused?
It may be that statutes in some jurisdictions are too generous in the types of research that they would allow with law enforcement DNA samples. This article does not answer that question. My claim about research is more limited. I have tried to show that much more argument and analysis than has been brought to bear to date would be required to establish that the research provisions of these laws countenance unethical biological research, and I have contended that the amici in Landry v. Attorney General present one side of a complex and fractious debate about archived tissue samples in biomedical research without revealing the existence of other views and without considering the differences between law enforcement databases and databanks and repositories of pathology samples.
* * *
The amicus briefs discussed here are no less adversarial than the parties' briefs. To the extent that they are simply the work of another group of lawyers, this poses no major problem. An appellate court can consider the competing arguments, make its own review of the cases and secondary literature, and seek the best possible interpretation of the law. The lesson for the courts is to do precisely this. Had the Massachusetts court relied solely on the representations of the Institute for Science, Law, and Technology, or the Owen M. Kupperschmid Holocaust Human Rights Project, its understanding of the technology of DNA identification and the pertinent ethical principles would have been incomplete if not distorted.
As interpreted by the Massachusetts Supreme Judicial Court, the convicted-offender DNA databanking law is defensible on constitutional and ethical grounds. Whether additional research uses of data or samples should be allowed without the consent of the offenders who were forced to provide the samples is a question of public policy that cannot be resolved by absolute and sweeping assertions that information on people can never be used without their consent. The developing norms of research on human subjects should be examined in the context of law enforcement databanks, and the briefs are valuable reminders of that fact.
* D.H. Kaye is Regents' Professor and Fellow, Center for the Study of Law, Science, and Technology, Arizona State University. He is grateful to Michelle Hibbert, Emily Hughes, Josephine Ross, and the Council for Responsible Genetics for providing copies of briefs or other materials relating to Landry v. Attorney General, 709 N.E.2d 1085 (Mass. 1999), and to George Sensabaugh, Elliot Goldstein, and Gary Marchant for discussions of some of the biological issues considered here.
1. 709 N.E.2d 1085 (Mass. 1999).
2. Middlesex County Superior Court Judge Isaac Borenstein also held the statute to be inconsistent with article 14 of the Massachusetts Declaration of Rights. Landry v. Harshbarger, C.A. No. 98-462 (Super. Ct. Aug. 12, 1998).
3. Id. at 12.
4. For a more detailed account of the litigation at the superior court level, see Eric T. Berkman, State Cannot Blood-test to Establish DNA Registry, Mass. Law. Wkly., Aug. 24, 1998, at 1, col. 4.
5. Dr. Carl Selavka, director of the state laboratory, commented that "[n]ow a single judge has delayed an opinion of the elected people of the state that this is an important crime-fighting tool." Mass. to Ask Court to Lift Ban on Testing DNA of Prisoners, Patriot Ledger (Quincy, Mass.), Sept. 19, 1998, at 12, available in WESTLAW, 1998 WL 8101516.
6. Dr. Paul Billings, a long-time advocate of restraint in using genetic information, said the ruling showed a balancing trend. "There are beginning to be decisions that show there have to be rules--that all this ethics stuff isn't blowing smoke." Sally Lehrman, Prisoners' DNA Database Ruled Unlawful, 394 Nature 818 (1998).
7. See infra notes 8-10. Oddly, the Supreme Judicial Court opinion in Landry does not include the IIT brief in its list of amicus briefs.
8. The Harvard group argued that the statute violates the Fourth Amendment by authorizing the government to use physical force to remove blood. Brief for Amici Curiae, Criminal Justice Institute at Harvard Law School, and Massachusetts Correctional Legal Services, Landry v. Attorney General, 709 N.E.2d 1085 (Mass. 1999). That argument is not examined here.
9. Brief for Amicus Curiae, Owen M. Kupperschmid Holocaust Human Rights Project of Boston College Law School and Criminal Justice Clinic of Boston College Law School, Landry v. Attorney General, 709 N.E.2d 1085 (Mass. 1999) [hereinafter BC Brief].
10. Brief for Amicus Curiae, Institute for Science, Law and Technology, Illinois Institute of Technology, Landry v. Attorney General, 709 N.E.2d 1085 (Mass. 1999) [hereinafter IIT Brief].
11. See Landry v. Attorney Gen., 709 N.E.2d 1085, 1092 n.11 (Mass. 1999).
12. Id. at 1091.
13. Id. (quoting Rise v. State, 59 F.3d 1556, 1560 (9th Cir.1995)) (alteration in original).
14. IIT Brief, supra note 10, at 27.
15. BC Brief, supra note 9, at 5.
16. See infra Part III.
17. In addition to the laws of the 50 states summarized in M. Dawn Herkenham, U.S. Dep't of Justice, State DNA Database Statutes: Summary of Provisions (1999), the DNA Identification Act of 1994, 42 U.S.C. § 14133 (1995), governs the national database created from state records maintained by the FBI.
18. IIT Brief, supra note 10, at 2-6.
19. Cf. D.H. Kaye, Bible Reading: DNA Evidence in Arizona, 28 Ariz. St. L.J. 1035 (1996) (identifying inconsistencies and fallacies, resulting from an incomplete understanding of genetics, in opinions on the admissibility of DNA evidence).
20. 526 U.S. 137 (1999).
21. Id. at 152 (explaining that the requirements for admitting expert testimony at trial "is to make certain that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.").
22. For such expositions, see Committee on DNA Technology in Forensic Science, Committee on DNA Technology in Forensic Science, National Research Council, DNA Technology in Forensic Science (1992) [hereinafter NRC I]; Committee on DNA Forensic Science: An Update, National Research Council, The Evaluation of Forensic DNA Evidence [hereinafter NRC II] (1996); David H. Kaye & George F. Sensabaugh, Reference Guide on DNA Evidence, in Reference Manual on Scientific Evidence (Federal Judicial Center ed., 2d ed., forthcoming 2000).
23. The brief begins by advertising itself as the product of an institute of "nationally and internationally known experts" who publish their findings "in peer-reviewed scientific and legal journals." IIT Brief, supra note 10, at 1. The signatories to the brief are Professors Lori B. Andrews and Harold J. Krent, and their student, Michelle Hibbert, of the Chicago-Kent College of Law. These authors emphasize the institute's funding from the National Institutes of Health and the National Science Foundation for projects "analyzing the social, economic, psychological and cultural consequences of DNA testing" and "analyz[ing] disputes over body tissue, including in the forensic DNA context." Id. at 1-2.
24. Id. at 2-6.
25. Id. at 2.
26. Id. The brief also seems to suggest that all genetic variation takes the form of length polymorphisms or that all "DNA analysis" is limited to probing this type of variation. Id. For more complete descriptions of the classes of polymorphisms examined in forensic and medical testing, see, e.g., NRC II, supra note 22; Kaye & Sensabaugh, supra note 22.
27. In classical genetics, an allele is a form of a single gene, not a combination of genes. In DNA typing, the term "allele" has been extended to encompass base-pair sequences that are not part of any gene. In neither situation is an allele "a differing gene combination."
28. If the human genome were a cookbook, and the recipe for ice cream were a gene, then the alleles would be the specific recipes for the flavors vanilla, strawberry, chocolate, pistachio, and so on.
29. The word "allele" originally referred only to coding DNA in that genes were posited as the units of heredity that controlled or influenced inherited traits (phenotypes) such as eye color.
The term is now applied to any sequence variation in the DNA, including the polymorphisms of noncoding loci. Because the entire DNA sequence (the genome) in a sex cell is copied from parent to child, one can think of the sequence variations themselves as inherited characteristics. But one should not lose sight of the fact that most of the human "genetic material" contains noise or information that is irrelevant to the characteristics that we normally regard as genetic traits. See Sydney Brenner, The Human Genome: The Nature of the Enterprise, in Human Genetic Information: Science, Law and Ethics 6, 8-9 (1990).
30. IIT Brief, supra note 10, at 2.
31. See, e.g., Kaye & Sensabaugh, supra note 22.
32. This description is simplified, and DNA probes with chemiluminescent markers also can be used.
33. See, e.g., David H. Kaye, DNA Evidence: Probability, Population Genetics, and the Courts, 7 Harv. J.L. & Tech. 101 (1993).
34. See, e.g., Stuart H. Orkin, Genetic Diagnosis by DNA Analysis, 317 N. Eng. J. Med. 1023 (1987) ("[T]he field began to emerge in 1976, when Kan et al. reported the use of molecular hybridization to detect the deletion of the alpha-globin genes in hydrops fetalis, reflecting the severest form of alpha-thalassemia.").
35. They are especially significant to the creation of large databases of DNA from criminals because the process of PCR amplification and subsequent characterization can be automated.
36. In addition, the output of the tests often is easier to interpret than the pictures (autoradiograms) that record the results of gel electrophoresis.
37. IIT Brief, supra note 10, at 3. If, by "RFLP analysis," the brief means "RFLP analysis of the VNTR loci conventionally used in forensic work," then the statement would be true. If "RFLP analysis" means "analysis of restriction fragment length polymorphisms," as one might think, then the statement is false.
38. Id. The brief also states that "PCR does not have as high a level of individual discrimination." Id. This is misleading, since a battery of PCR-based tests could be either more or less discriminating than a set of RFLP-based tests. The specificity depends on the number of loci tested, the number of alleles per locus, and the allele frequencies. It is practical to examine more loci with PCR methods, and the thirteen loci that the FBI has selected for databanking yield an "average match probability [of] one in 180 trillion." National Commission on the Future of DNA Evidence Meeting V Proceedings, May 7, 1999 (statement of James Crow) <http://www.ojp.usdoj.gov/nij/dnamtgtrans5/trans-h.html> (visited Sept. 14, 1999).
39. "STR" stands for "short tandem repeat." In the DNA fragments that come from the STR locus D7S820, for instance, one individual might have an allele that consists of, say, six repeats of the sequence GATA, while another person might have an allele with ten repeats. See STR Fact Sheet--D7S820 <http://www.cstl.nist.gov/div831/strbase/str_d7s8.htm (visited Jan. 26, 2000). STRs are much shorter than RFLPs.
40. The human chromosomes that determine an individuals sex are denoted X (female) and Y (male). The remainder are known as autosomal chromosomes.
41. IIT Brief, supra note 10, at 4.
42. An example of possible relevance to law is the X-linked gene for a rare syndrome that involves "mild mental retardation and control of aggressive behavior." William S. Klug & Michael R. Cummings, Concepts of Genetics 653 (5th ed. 1997). "Using a variety of RFLP markers," the locus associated with "a pattern of aggressive behavior and lack of impulse control . . . was mapped to the short arm of the X-chromosome," near the locus that codes for monoamine oxidase A (MAOA). Id. "Monoamine oxidases are enzymes that degrade . . . neurotransmitters," and the allele that causes the condition contains a point mutation that results in a lack of MAOA activity. Id. Given the current understanding of the MAOA gene, the presence of the disease-causing allele could be determined with a PCR-based test (but not one that involves STRs), or affected individuals could be identified by a closely linked RFLP marker.
43. Thus, allegations of paternity (and maternity) are tested quite effectively with loci on autosomal chromosomes. See, e.g., D.H. Kaye, DNA Paternity Probabilities, 24 Fam. L.Q. 279 (1990).
44. IIT Brief, supra note 10, at 4.
45. Among the disease-related genes that were located with now classic RFLP family linkage studies are the gene for type 1 neurofibromatosis and the gene for Marfan's syndrome. See Klug & Cummings, supra note 42, at 464-66.
46. IIT Brief, supra note 10, at 4.
47. Id. at 5.
48. In contrast to the haploid nuclear genome of over three billion base pairs, the mitochondrial genome is a circular molecule 16,569 base pairs long that contains only 37 genes. Alan G. Atherly et al., The Science of Genetics 193 (1999).
49. In these organelles, certain energy-storing molecules are synthesized. Id. at 276.
50. This view may be weaker than the IIT group's claim that "the new developments in DNA technology threaten a far more serious invasion of privacy than under the DNA databank schemes previously assessed by other courts." IIT Brief, supra note 10, at 6. It is, however, more defensible. The assertion in the brief seems to overlook the fact that other courts have upheld DNA databases that use STRs, or that retain samples that could be analyzed with mitochondrial or STR testing. See, e.g., In re J.W.K., 583 N.W.2d 752 (Minn. 1998). Moreover, even if it were true that no other court had considered a "DNA databank scheme" involving modern technology, whether mitochondrial and STR testing add so many more loci of social concern is open to question.
51. Landry, 709 N.E.2d at 1089 n.4.
52. The court cited the 1996 NRC Report for this proposition. Id. The National Academy of Sciences committee actually reported that "STR loci appear to be particularly appropriate for forensic use. Like VNTRs, they can be chosen to be in noncoding regions and can therefore be expected to be selectively neutral." NRC II, supra note 22, at 117 (emphasis added).
53. Many of the CODIS STRs occur in the noncoding portions (introns) of genes. See FBI Core STR Loci (visited Jan. 26, 2000) < http://www.cstl.nist.gov/div831/strbase/fbicore.htm>. As such, some might be linked to diseases resulting from mutations in nearby coding portions (exons). However, the predictive power of the STR alleles is very slight because the same alleles are commonly found in individuals without any such diseases. Telephone interview with George Sensabaugh (Jan. 26, 2000). For example, suppose that 90% of individuals with a given disease have a particular STR allele, while 10% of the population carries the same allele. If the disease occurs in 0.5% of the population (which is high for a single gene disorder), then of every 10,000 individuals, about 50 (0.5% × 10,000) will have the disease, and about 45 of them (90% × 50) will have the allele. But because approximately 1,000 (10% × 10,000) people will have the allele, only 45/1,000 people with the allele will have the disease. Even thought the probability that someone with the disease has the allele is large (90%), the probability that someone with the allele has the disease is small (4.5%). Because the "predictive value positive" of the genetic test is low, the allele is not a useful diagnostic marker for the disease. This situation is typical of conditions that have a low prevalence. See, e.g., Joseph L. Gastwirth, The Statistical Precision of Medical Screening Procedures: Application to Polygraph and AIDS Antibodies Test Data, 2 Stat. Sci. 213 (1987); D.H. Kaye, The Validity of Tests: Caveant Omnes, 27 Jurimetrics J. 349 (1987).
54. IIT Brief, supra note 10, at 20 nn.3 & 21.
55. Id. at 26.
56. Id. at 22 n.4.
57. Id. at 23. Lander was critical of RFLP testing of VNTRs as implemented in several cases in the early years of forensic work. Yet, as early as 1994, Lander wrote that "no remaining problem . . . should prevent the full use of DNA evidence in any court." Eric S. Lander & Bruce Budowle, DNA Fingerprinting Dispute Laid to Rest, 371 Nature 735, 735 (1994). Moreover, the brief fails to reveal that the 1992 National Academy of Science panel, on which Lander served, concluded that "[i]n principle, a national DNA profile databank should be created . . . ." NRC I, supra note 22, at128. Although the committee recommended that no such database be created "yet," its hesitation had nothing to do with the supposedly high rate of errors in DNA testing. Rather, the committee explained, "we expect current [RFLP-based] methods to be replaced soon with techniques that are simpler, easier to automate, and less expensive . . . ." Id. at 129.
58. Id. at 20. The brief states that "in the 223 tests, matches were identified in 18 cases where they did not exist." Id. However, Koehler and his colleagues discussed pairwise comparisons rather than "cases," and they caution that "the use of . . . mixed samples . . . introduces some interpretive ambiguity [such that] it is not clear whether an erroneous match call between sample A and the second fraction of the mixed sample (B2) should be scored as an error" when one considers that "match calls on non-matching samples and fractions of samples . . . are unlikely to falsely incriminate an innocent suspect because the suspect matches the other fraction of the mixed sample." Jonathan J. Koehler et al., The Random Match Probability in DNA Evidence: Irrelevant and Prejudicial?, 35 Jurimetrics J. 201, 208 (1995). A more complete presentation would have noted this ambiguity and perhaps presented another statistic used by Koehler et al. to avoid this problem. (They counted 4/181 false declarations of matching pairs. Id. at 209. This figure is still disturbingly high, although for reasons not mentioned in the IIT brief, it is probably not representative of database matches at the forensic STR loci.) The other data sets of pairwise comparisons analyzed by Koehler are not mentioned in the brief even though most produced error rates smaller than 18/223. Neither does the brief reveal the conclusion reached by Koehler and his colleagues:
Despite some ambiguities in the interpretation of proficiency test data, performance on the CACLD and CTS tests provides some basis for estimating a lower bound diagnosticity value. When the results from the two CACLD studies are combined, P("M"&-M|-S) = 5/6,388 (approximately 1/1,300). When the results from the three CTS tests are combined (excluding fraction comparisons and errors), P("M"&-M|-S) = 4/803 (approximately 1/200). When the results from all of the CACLD and CTS studies are pooled, P("M"&-M|-S) = 9/7,191, or approximately 1/800.
Id. at 210. Of course, the Institute is free to urge the court to infer from all the available data that DNA databanking will produce so many errors that the law cannot stand. But this brief merely offers bits and pieces of the information on possible error rates.
59. IIT Brief, supra, note 10, at 23 n.5. The brief depicts Fred Zain's outright falsification of records as an example of "innocent people [being] convicted" (id. at 23-24) because "[a]lthough the DNA per se did not falsely convict the defendant, and actually DNA later exonerated the defendant, the fact remains that an innocent man went to jail because of the technician's actions." Id. at 23 n.5. Regrettably, incidents of perjury involving police, technicians, and scientists predate forensic DNA technology. See D.H. Kaye, Science in Evidence 14-15 (1997).
60. IIT Brief, supra note 10, at 1.
61. Id. at 2.
62. Id. at 35.
63. IIT Brief, supra note 10, at 28.
64. The example assumes, unrealistically, that there are known markers for common disorders like diabetes or schizophrenia. These diseases have a genetic component, but they are not single-gene diseases, and the multiple genes and environmental factors that interact to cause such diseases are poorly understood.
65. Literature on the extent to which insurers seek genetic test data is noted in D.H. Kaye, Respecting Genetic Privacy: A Foreword to the ASU-SB Conference on Law, Science, and Technology, 40 Jurimetrics J. 1, 7 (1999).
66. Mass. Gen. Laws Ann. ch. 22E §§ 10, 12, 13 (West Supp. 1999).
67. See, e.g., Ellen Wright Clayton, What Should the Law Say about Disclosure of Genetic Information to Relatives?, 1 J. Health Care L. & Pol'y 373 (1998); Sonia M. Suter, Whose Genes Are These Anyway? Familial Conflicts over Access to Genetic Information, 91 Mich. L. Rev. 1854 (1993).
68. IIT Brief, supra note 10, at 33.
69. Id. at 34. The brief asserts that "[p]eople are being denied health and life insurance based on relatives' DNA" and that "family members" of offenders are worried that collecting DNA will make them uninsurable (presumably because the police will give the samples to insurance companies who are forbidden by law from using them). Id. at 34-35. If this is meant to establish a possible harm as the foundation for a right to confidentiality in the information in a relative's DNA, it is far too weak a foundation to bear that construction.
70. Id. at 30-31 (emphasis in original).
71. 999 U.N.T.S. 171, Dec. 19, 1999. The United States ratified the covenant in 1992. BC Brief, supra note 9, at 6.
72. 1465 U.N.T.S. 85, June 26, 1987. The U.S. became a party to this agreement in 1994. BC Brief, supra note 9, at 6.
73. Mass. Gen. Laws Ann. ch. 22E § 1 (West Supp. 1999).
75. Id. § 10.
76. See id. §§ 12 (making it a crime punishable by a fine of up $1,000 and imprisonment of up to six months for an official to disclose a record or a part of it to "any person or agency not authorized to receive such record") & 13 (making it a similar crime to obtain a record or part of it "without proper authorization").
77. Id. § 10(d).
78. The director merely is instructed to "promulgate regulations governing the . . . storage and disposal of DNA samples." Id. § 6.
79. BC Brief, supra note 9, at 4 n.1 (emphasis added).
80. See, e.g., IIT Brief, supra note 10, at 26-27 (interpreting the text of § 10(a), which makes "DNA records and analysis" available to all state and local law enforcement agencies, as permitting a prosecutor to maintain a trophy case filled with "the DNA of those he or she successfully convicted" and a parole board "to examine the DNA sample to determine whether an individual has a particular gene linked to violence").
81. 709 N.E.2d 1096 ("[T]he Act limits the purposes for which the DNA records may be distributed, and does not permit dissemination of the DNA samples themselves.").
82. Id. at 1089 n.5.
83. Id. at 1096 (notes omitted). In a footnote, the court remarked that:
When promulgating final regulations for the Act, the director may want to provide more detail as to tests that may be performed on the DNA samples that are being collected and stored. While the Act only authorizes use of those portions of DNA samples that are relevant for identification purposes, the indefinite storage of the entire DNA sample, see 515 Code Mass. Regs. § 1.05(4), creates some concern that the samples could be misused at some point in the future to search for and disclose private genetic information.
Id. at 1096 n.20.
84. IIT Brief, supra note 10, at 33.
85. BC Brief, supra note 9, at 10.
86. United States v. Brandt (The Medical Cases), 2 Trials of War Criminals Before the Nuremberg Military Tribunals Under Control Council Law No. 10 (1949).
87. See Jay Katz, Human Sacrifice and Human Experimentation: Reflections at Nuremberg, 22 Yale J. Int'l L. 401, 411 (1997) ("The tribunal articulated a vision of the limits of scientific medical research that was clear and unambiguous.").
88. The Declaration of Helsinki, which has been said to "supersede" the Nuremberg Code, Jay Katz, The Consent Principle of the Nuremberg Code: Its Significance Then and Now, in The Nazi Doctors and the Nuremberg Code: Human Rights in Human Experimentation 227, 231 (George J. Annas & Michael A. Grodin eds., 1992), is even more clearly circumscribed. Its third principle, for instance, presupposes that "[b]iomedical research involving human subjects" can affect directly the health of the subjects. For that reason, it specifies that all such research "should be conducted only . . . under the supervision of a clinically competent medical person" and that "[t]he responsibility for the human subject must always rest with a medically qualified person." World Medical Association, Declaration of Helsinki (revised by the 41st World Medical Assembly (Hong Kong, 1989)).
89. See United States v. Stanley, 483 U.S. 669 (1987); George J. Annas, The Nuremberg Code in U.S. Courts: Ethics versus Expediency, in The Nazi Doctors and the Nuremberg Code: Human Rights in Human Experimentation 201, 212-13 (George J. Annas & Michael A. Grodin eds., 1992).
90. United States v. Brandt (The Medical Cases), 2 Trials of War Criminals Before the Nuremberg Military Tribunals Under Control Council Law No. 10 (1949).
91. E.g., BC Brief, supra note 9, at 16 (asserting, without analysis, that "[a]lthough the figure [sic] of DNA evidence challenged in this case is nowhere near as horrific as the atrocities committed by Nazi doctors, it is nonetheless clear that the Nuremberg Code applies.").
92. Cf. Arthur L. Caplan, The Doctors' Trial and Analogies to the Holocaust in Contemporary Bioethical Debates, in The Nazi Doctors and the Nuremberg Code: Human Rights in Human Experimentation 258 (George J. Annas & Michael A. Grodin eds., 1992); Manfred D. Laubichler, Frankenstein in the Land of Dichter and Denker, 286 Science 1859, 1860 (1999) ("Genetics today is not the same as eugenics and racial hygiene in the 1930s, which were more concerned with technocratic solutions on the level of whole populations than with any detailed understanding of the role of genes in development and disease. . . . Framing the debate exclusively in terms of literary images . . . or the crimes and ideology of the Nazi Period is not enough.").
93. Compare BC Brief, supra note 9, at 19, with Arthur L. Caplan & George J. Annas, Letter, 285 Science 48, 48-49 (1999) ("Analogies must be generated with caution. Sloppy analogies to historical events such as Tuskegee abound. Caution and accuracy are crucial so as not to demean or deprecate the horrific moral abuses to which human beings were subjected in the past in the name of medical progress.").
94. Katz, supra note 87, at 404.
95. Mass. Gen. Laws Ann. ch. 22E § 10(d) (West Supp. 1999).
96. See supra note 39. Most STR loci are heterozygous, meaning that there usually are two alleles per locus. The full STR profile of an individual at the 13 CODIS loci, therefore could consist of as many as 26 numbers (13 pairs).
97. See, e.g., Ian W. Evett & Bruce S. Weir, Interpreting DNA Evidence: Statistical Genetics for Forensic Scientists (1998).
98. Id.; Daniel L. Hartl & Andrew G. Clark, Principles of Population Genetics (3d ed. 1997).
99. See L.A. Foreman & I.W. Evett, Statistical Analyses to Support Forensic Interpretation for a New Ten-locus STR Profiling System, Int'l J. Legal Medicine (submitted 1999).
100. BC Brief, supra note 9, at 14; see also id. at 4 (complaining that "the seized DNA will be taken so it can be experimented with, to thus adjust and improve the scientific process of DNA testing in general."). The brief also complains that the Massachusetts statute "violates the Nuremberg Code on its face" because it "specifically directs that force be used if the prisoners, probationers and parolees doe not acquiesce to the removal of their DNA." Id. at 14. However, the forcible-removal provision is obviously a red herring here. The act makes it a crime for "[a]ny person required to provide a DNA sample" to refuse. Mass. Gen. Laws Ann. ch. 22E § 11 (West Supp. 1999). Acquiescence under pain of fine and imprisonment for resistance is involuntary.
Whether the IIT scholars share this view of the Code is unclear. They write that "Section 10(d) is highly problematic," but they proceed to discuss disclosure of the "samples" rather than the records. IIT Brief, supra note 10, at 29-31.
101. See supra Part III.B; Katz, supra note 88, at 231 ("[T]he first principle of the Nuremberg Code, if adopted, may very well have required modification; it clearly would have required exigesis.").
102. IIT Brief, supra note 10, at 30.
103. See, e.g., Henry T. Greely, Iceland's Plan for Genomic Research, 40 Jurimetrics J. __, __ (2000).
104. The example in Eric T. Juengst, I-DNA-fication, Personal Privacy, and Social Justice, 75 Chi.-Kent L. Rev. 61 (1999), is not to the contrary. Professor Juengst worries that "sample anonymity offers no protection [against] the risk of being discovered to be a convicted criminal by any researcher who already knows your genetic identity and finds you in the database." Id. at 69. As an illustration, he posits
a clever employer, like a University Department of Forensic Science, who is concerned about promoting anyone with a criminal conviction. If the candidates DNA profiles can be voluntarily acquired ("a student project"), and an adequate research proposal concocted, the Department could scan the state database in search of its promotion candidates . . . .Id. at 69 n.24.
This example goes to the question of whether anonymizing the database records--not the samples--is sufficient protection against privacy. As to that question, the example uncovers no invasion of privacy. A conviction is a public record. Rather than the convoluted scheme to search the DNA database, the university easily could search these records. Unless the candidate has a right to conceal a criminal record from an employer, the university's discovery that the candidate has a criminal record violates no right to informational privacy.
105. IIT Brief, supra note 10, at 30-31 ("Moreover, even if the sample has been anonymized, ethical principles governing research . . . mandate that research not be undertaken without the subject's advance, voluntary, informed consent.") (emphasis in original).
106. It invokes the Code by citing a sentence in United States v. Brandt as quoted in United States v. Stanley, 483 U.S. 669, 710 (1987) (O'Connor, J., dissenting). Id. at 30-31.
107. Id. at 31.
108. These administrative regulations, have been adopted in nearly identical form by 17 federal agencies to govern research on human subjects funded by or conducted for these agencies. See 45 C.F.R. pt. 46 (1998). Despite the name, the Common Rule is not a single rule, but a set of procedures that must be followed. The Rule requires that a protocol describing the research be submitted to an Institutional Review Board (IRB), which determines whether the potential benefits of the research outweigh its risks.
109. See, e.g., Henry T. Greely, Breaking the Stalemate: A Prospective Regulatory Framework for Unforeseen Research Uses of Human Tissue Samples and Health Information, 34 Wake Forest L. Rev. 737 (1999); David Korn, Genetic Privacy, Medical Information Privacy, and the Use of Human Tissue Specimens in Research, in Genetic Testing and the Use of Information 16 (Clarisa Long ed., 1999).
110. IIT Brief, supra note 10, at 31.
111. 45 C.F.R. § 46.101(b)(4) (1998). Of course, an offender databank sample is not literally a pathological or diagnostic specimen, but I am not asking whether the Common Rule would apply to federally funded or conducted research using a state or federal offender databank as a source of tissue. Rather, I am asking whether the representations of the amici about the need for consent with anonymized tissue samples in the tissue archives maintained by pathologists are accurate and complete. With regard to the question of the applicability of the Common Rule to a forensic databank, however, it could be argued that the sample is "diagnostic" even thought it was not originally acquired to diagnose a disease. The argument would be that an offender blood sample, for instance, is functionally equivalent to a patient's blood sample and was obtained to "diagnose" a biometrically revealing trait. Furthermore, even if this argument were rejected and IRB review were required, it could be argued that an institutional review board could waive the consent requirement.
112. See generally, e.g., Korn, supra note 109.
113. Ellen Wright Clayton et al., Informed Consent for Genetic Research on Stored Tissue Samples, 274 JAMA 1786, 1786 (1995).
114. Greely, supra note 109, at 737.
115. IIT Brief, supra note 10, at 31 (citing Clayton et al., supra note 113, at 1788). The brief fails to note that the "panel" consists of only eight participants self-selected from a larger advisory group convened by the National Institutes of Health and the Centers for Disease Prevention and Control--a group that was unable to reach any consensus. Greely, supra note 109, at 742. (The "panel" of "physicians, lawyers, and ethicists" were Ellen Wright Clayton (Vanderbilt University), Karen K. Steinberg (CDC), Muin J. Khoury (CDC), Elizabeth Thomson (National Center for Human Genome Research), Lori Andrews (IIT), Mary Jo Ellis Kahn (National Breast Cancer Coalition and Virginia Breast Cancer Foundation), Loretta M. Kopelman (East Carolina University School of Medicine), and Joan O. Weiss (Alliance of Genetic Support Groups). Clayton et al., supra note 113, at 1788.) Moreover, the amicus brief fails to inform the court that the interpretation of the Common Rule offered by these eight individuals triggered "an explosion of articles and position statements affirming, denouncing, and qualifying the . . . article from, among others, the American Association of Medical Colleges, the American Society of Human Genetics, the American College of Medical Genetics, the Human Genome Organisation, and a committee of the American College of Pathologists." Greely, supra, at 744 (notes omitted).
116. See, e.g., Gerald Dworkin, The Theory and Practice of Autonomy (1988); R.J. Levine, Ethics and Regulations of Clinical Research (2d ed. 1986). But see John Harris, Ethical Genetic Research on Human Subjects, 40 Jurimetrics J. 77 (1999).
117. Clayton et al., supra note 113, at 1789.
119. Juengst, supra note 104, at 69-70.