The papers included in this volume were edited to conform to minimal stylistic consistency. The content and accuracy of the papers are the responsibility of the authors, not the National Bioethics Advisory Commission.
Author(s): Jack Schwartz
Federal regulations establish basic protections for human subjects that cannot be diminished by state law.
That is, a researcher or institution subject to a federal requirement must comply with it. This obligation exists whether or not a comparable requirement is imposed by state law. Moreover, a researcher may not engage in an activity prohibited by federal law even if state law were to allow it. As the Supreme Court recently reiterated, "even if Congress has not occupied the field, state law is naturally preempted to the extent of any conflict with a federal statute."1 The same principle, grounded in the Supremacy Clause of the United States Constitution, leads to preemption of state laws that conflict with federal regulations.
Yet, state law is hardly irrelevant to the research enterprise. First, to the extent that research is not subject to federal law, pertinent state law (if any) becomes the only legally applicable regulatory regime. Second, federal law, when it does apply to research, expressly preserves any additional state protections. The Common Rule contains the following non-preemption language: "This policy does not affect any State or local laws or regulations which may otherwise be applicable and which provide additional protections for human subjects."2 In addition to this general provision, the Common Rule more specifically recognizes additional state requirements for informed consent: "The informed consent requirements in this policy are not intended to preempt any applicable Federal, State, or local laws which require additional information to be disclosed in order for informed consent to be legally effective."3 Identical language appears in the regulations of the Food and Drug Administration.4 This policy decision, to preserve a role for states in the regulation of the research enterprise, is unsurprising, for the protection of human subjects may be seen as an application of a state's core function, protecting its citizens against harm. As the Supreme Court observed many years ago, "The police power of a state...springs from the obligation of the state to protect its citizens and provide for the safety and good order of society....It is the governmental power of self-protection and permits reasonable regulation of rights and property in particulars essential to the preservation of the community from injury."5 Yet, state regulation of research has not escaped criticism. Some contend that, especially as research increasingly involves multi-site collaborations, a regulatory system inefficiently increases compliance costs if it imposes requirements or restrictions that differ from one site to the next. These critics insist that federal regulation is sufficient and state regulation unnecessary or even harmful. For example, lobbyists for the Pharmaceutical Research and Manufacturers of America, commenting on a legislative proposal in Maryland to regulate research involving decisionally incapacitated subjects, argued that "creation of additional state laws that add to the federal common rule may inhibit the conduct of research on a multi-state level" and, they asserted, would "delay approval of lifesaving and life-altering drugs."6 Others, while accepting that states can play an appropriate role in the overall regulatory system, have condemned specific state activities as short-sighted and inimical to research. For example, critics of the intermediate appellate court decision in T.D. v. New York State Office of Mental Health, discussed later in this paper, suggested that its "repercussions may prevent numerous studies from going forward and ultimately put a halt to research that may have provided participants with access to effective treatment and significantly improved the well-being of other individuals suffering from mental illness and other cognitive impairments" (Oldham, Haimowitz, and Delano 1998, 154).
The Commission's consideration of the appropriate role of the states in an overall regulatory system should be informed by a survey of current state regulatory efforts. This paper presents such a survey. It discusses state statutes and regulations concerning research activities or the information used in research, the limited case law involving research, and a few particular issues for Commission attention.
Back to Top
One of the oft-noted voids in the current federal regulatory system is that some privately funded research is unregulated. That is, research is subject to federal regulation only if it is conducted or funded by agencies that have subscribed to the Common Rule; is the basis on which the Food and Drug Administration (FDA) will be asked to approve a drug or device; or is conducted at an institution that, as an aspect of its assurance to the Office for Protection from Research Risks (now the Office for Human Research Protections), has agreed to conduct all research at the institution in accordance with the Common Rule. Otherwise, research is free of federal regulation. Consequently, as the Commission has pointed out, "the absence of Federal jurisdiction over much privately funded research means that the U.S. government cannot know how many Americans currently are the subjects in experiments, cannot influence how they are recruited, cannot ensure that research subjects know and understand the risks they are undertaking, and cannot ascertain whether they have been harmed."7 If Congress sought to assert federal jurisdiction over all privately funded research, its ability to do so would depend on the scope of its Commerce Clause power in this context.8 "Every law enacted by Congress must be based on one or more of its powers enumerated in the Constitution."9 By contrast, a state need invoke no particular grant of constitutional authority to regulate privately funded research or any other activity within its borders; state legislatures have plenary power, subject only to constitutional limitations. As the Supreme Court wrote more than century ago, "The legislative power of a State extends to everything within the sphere of such power, except as it is restricted by the Federal Constitution or that of the State."10 A state legislature that deemed it appropriate to regulate research beyond the scope of existing federal regulations is unquestionably free to do so as an aspect of its sovereign power.
Hence, the concerns expressed by the Commission about the consequences of limited federal jurisdiction might be ameliorated if privately funded research were subject to state laws that provided for the twin protections of informed consent and independent review. Only two states, however, New York and Virginia, have applied these protections to biomedical research generally.11 A third state, California, has applied a vigorous informed consent requirement comprehensively.
The New York law, enacted in 1975, applies to "human research," defined as "any medical experiments, research, or scientific or psychological investigation, which utilizes human subjects and which involves physical or psychological intervention by the researcher upon the body of the subject and which is not required for the purposes of obtaining information for the diagnosis, prevention, or treatment of disease or the assessment of medical condition for the direct benefit of the subject."12 The law excludes studies limited to tissue or fluid specimens taken in the course of standard medical practice, epidemiological studies, and "human research which is subject to, and which is in compliance with, policies and regulations promulgated by any agency of the federal government for the protection of human subjects."13 By means of this last exclusion, the New York law avoids any additional regulatory burden on research that is already subject to the Common Rule or FDA regulations.
The New York law extends to privately funded research the core requirements of the federal scheme: informed consent and institutional oversight. "No human research may be conducted in this state," the law directs, "in the absence of the voluntary informed consent subscribed to in writing by the human subject" or, in the case of a minor or other individual "legally unable to render consent," by someone who is "legally empowered to act on behalf of the human subject."14 "Voluntary informed consent" is consent that is "knowing," given by someone "so situated as to be able to exercise free power of choice without undue inducement or any element of force, fraud, deceit, duress or other form of constraint or coercion."15 The law's itemization of the "basic elements of information necessary to such consent" parallels those in the Common Rule's informed consent provision:16 an explanation of the research procedures; a description of risks and benefits, if any; a disclosure of alternatives to research participation; an offer to provide additional information; and notice of the subject's right to discontinue participation at any time.17 The law requires each research institution, public or private, to establish a "human research review committee," the membership requirements of which are similar to those in the Common Rule's specification for IRBs.18 The committee is to secure from its parent institution a "statement of principle and policy in regard to the rights and welfare of human subjects," which is subject to approval by the New York commissioner of health.19 The committee is then to review each proposed instance of human subject research within the institution for "its necessity," adequate protection of the subjects' rights and welfare, a favorable ratio of benefits to risks, "adequate and appropriate" informed consent methods, and sufficiently qualified researchers.20 Although no annual or other set requirement for continuing review is specified, the committee is to "periodically examine each existing...research project with regard to the proper application of the [previously] approved [institutional] principles...." and is to report any violation to the commissioner of health.21 In addition to this oversight role, indistinguishable in its main elements from that of an IRB operating under the Common Rule, a committee is also called on to play a more direct role when vulnerable subjects are to be involved. Beyond the informed consent of the subject or legally authorized representative, "the consent of the committee and the commissioner [of health] shall be required with relation to the conduct of human research involving minors, incompetent persons, mentally disabled persons, and prisoners."22 The Virginia law, enacted in 1979, applies to "human research," defined as "any systematic investigation utilizing human subjects which may expose such human subjects to physical or psychological injury as a consequence of participation as subjects and which departs from the application of established and accepted therapeutic methods appropriate to meet the subjects' needs."23 The law exempts some of the same activities as are exempt from the Common Rule: educational testing, survey and interview procedures, and observation of public behavior, if the resulting data are not linked to personal identifiers or if, in any event, disclosure would not be damaging to the subjects;24 the surveying or interviewing of public officials or candidates for public office;25 and research involving existing, unlinked data or specimens.26 In addition, the law exempts the disease prevention and control activities of the Virginia Department of Health.27 Finally, research is exempt if it is "subject to policies and regulations for the protection of human subjects promulgated by any agency of the federal government...."28 For human research that is not exempt, the law requires both informed consent and approval by a "human research review committee."29 If the subject is not competent, informed consent may be given by a legally authorized representative.30 Proxy consent may not be given, however, for participation in "nontherapeutic research unless it is determined by the human research committee that such nontherapeutic research will present no more than a minor increase over minimal risk"31a description of risk drawn from the federal regulation on pediatric research32 but criticized by the Commission (National Bioethics Advisory Commission 1998, 4043). The elements of informed consent are an abbreviated version of the Common Rule requirements.33 Likewise, the law's standard for protocol approval by a human research review committee is an abbreviated version of the Common Rule requirements.34 The law omits the Common Rule's reference to an IRB's authority to impose additional safeguards when the research subjects are drawn from vulnerable populations.35 The law's continuing review provision simply states the committee's obligation to "require periodic reports from each existing human research project to ensure that the project is being carried out in conformity with the proposal as approved";36 unlike the Common Rule, there is no mention of a maximum one-year interval between reviews or of third-party observation of the consent process or the research itself.37 Virginia seeks to promote compliance with its research law through licensing laws for physicians and other health care professionals; failure to comply "shall constitute unprofessional conduct," a ground for discipline against licensees.38
California enacted its Protection of Human Subjects in Medical Experimentation Act in 1978. The law's scope is narrower than New York's or Virginia's, for it contains no provisions for IRB review. Instead, it gives strong legal force to the informed consent requirement.
The law applies to "medical experiments": invasive procedures or the use of a drug, device, or radiation "in a manner not reasonably related to maintaining or improving the health of the subject or otherwise directly benefiting the subject."39 The term also includes the withholding of medical treatment "for any purpose other than maintenance or improvement of the health of the subject."40 Research that would otherwise be subject to the law is exempted, however, if it is conducted within an institution that holds an assurance of compliance with the Common Rule.41 Also exempt, by virtue of a 1997 amendment, is emergency research that meets the main requirements of the FDA's exception from informed consent requirements.42 The California law requires informed consent from the subject or one of several specified legally authorized representatives.43 If a representative provides consent, the research must be "related to maintaining or improving the health of the human subject or related to obtaining information about a pathological condition of the human subject."44 Conducting a medical experiment without first obtaining informed consent subjects the principal investigator to civil damages, if the failure is negligent, or to criminal prosecution, if the failure is willful.45 The elements of informed consent largely parallel those of the Common Rule. The California law also requires explicit identification of the research sponsor or funding source, or the name of the manufacturer if a drug or device is involved,46 and contact information for "an impartial third party, not associated with the experiment, to whom the subject may address complaints about the experiment."47
California,48 Louisiana,49 Michigan,50 and Rhode Island51 have enacted prohibitions on efforts to clone a human being.52 Contrary to the Commission's recommendation (National Bioethics Advisory Commission 1997), none of these enactments contains a sunset clause. Laws of this kind are analyzed in detail in a previous commissioned paper (Andrews 1997).
Since 1978, New York law has required those who engage in "recombinant DNA activity" to be certified by the commissioner of health. The commissioner in turn, is required to adopt regulations that are the "substantial equivalent of [the containment' and experimental guidelines' sections, including any revisions] of the recombinant DNA research guidelines of the National Institutes of Health...."53 Rather more directly, Oregon requires anyone "carrying out recombinant DNA research [to] comply with the recombinant research guidelines adopted by the National Institutes of Health and any subsequent modifications thereof."54
The Department of Health and Human Services (DHHS) regulations that apply to research involving fetuses, pregnant women, and in vitro fertilization, 45 CFR Part 46, Subpart B, contain language that, although close to a parody of legalistic style, does unmistakably affirm the primacy of state law: "Nothing in this subpart shall be construed as indicating that compliance with the procedures set forth herein will in any way render inapplicable pertinent State or local laws bearing upon activities covered by this subpart."55 Consequently, although Subpart B allows research of this kind to go forward with certain additional protections, states are free, as many have, to enact highly restrictive or prohibitory legislation. For example, as reported by Professor Andrews in an earlier commissioned paper, nine states have banned research involving in vitro embryos (Andrews 2000, A-4).
Amid the wide variety of state laws in this area are those that seemingly attempt to codify aspects of the federal regulations. For example, Subpart B limits DHHS-supported research involving pregnant women to studies that are intended "to meet the health needs of the mother and the fetus will be placed at risk only to the minimum extent necessary to meet such needs" or, for other research, "the risk to the fetus is minimal."56 New Mexico has incorporated substantially the same language in a statute applicable to all clinical research, whatever its funding source.57
The DHHS regulations that apply to research involving prisoners, 45 CFR Part 46, Subpart C, also contain language explicitly preserving more restrictive state law: "Nothing in this subpart shall be construed as indicating that compliance with the procedures set forth herein will authorize research involving prisoners as subjects, to the extent such research is limited or barred by applicable State or local law."58 Some states do indeed bar research involving prisoners if the research has no prospect of direct medical benefit. For example, regulations in at least four states prohibit "the use of an inmate for medical, pharmaceutical, or cosmetic experiments," although the regulations allow, to quote the Massachusetts provision, for participation in research that is "medically appropriate for a specific inmate."59 Similarly, Oregon law states flatly that, "There shall be no medical, psychiatric, or psychological experimentation or research with inmates in Department of Corrections institutions of the State of Oregon."60 The pertinent definitions limit the ban, however, to what the law terms "nontherapeutic" procedures.61 Laws in some states, like Arizona, generally allow the conduct of whatever research is approved by the prison director and chief of inmate health services.62 Georgia regulations vest broad approval discretion in the commissioner of corrections but add the requirement of "periodic reports [to the commissioner] during the course of the project."63 Other states distinguish in more detail permissible from impermissible research. California law, for example, prohibits all "biomedical research," except for participation in investigational new drug research deemed to be in the prisoner's best medical interest.64 "Behavioral research," on the other hand, is permissible if it concerns certain penological matters and presents "minimal or no risk and no more than mere inconvenience to the subjects of the research."65 This language is drawn from Subpart C;66 other reasons for prisoner research that are at least theoretically permissible under the federal regulation are omitted from the California law and, accordingly, are prohibited.67 Virginia law draws lines based on benefit and risk: "Nontherapeutic research using institutionalized participants shall be prohibited unless it is determined by [a] research review committee that such nontherapeutic research will not present greater than minimal risk."68
Unlike the other subparts of 45 CFR Part 46, the DHHS regulations concerning children as research subjects, Subpart D, do not expressly preserve state law on the subject, apart from deferring to state law for determining "the legal age for consent to treatments or procedures involved in the research...."69 Nevertheless, Subpart D does not purport to preempt any state laws and, since the regulations are framed in terms of research that DHHS "will conduct or fund,"70 they have no direct effect on state regulatory efforts.
No state has enacted a local analogue to Subpart D. Rather, states have focused on the need to enact protections when children are receiving social services from the state or are in state facilities. The Oregon law on the care and treatment of indigent children, for example, states flatly that, "No child shall be used for the purpose of experimentation."71 Massachusetts law, by contrast, allows unrestricted research participation with parental consent or, for children in state custody after a termination of parental rights, with court approval.72 Somewhere in the middle are laws like the Illinois regulations governing research involving children served by the Department of Children and Family Services; under these regulations, research risk may not exceed minimal;73 "the purely experimental use of drugs in research" is prohibited;74 and the selection of research subjects "will not be based solely on administrative convenience, availability of a population living in conditions of social or economic deprivation, or convenient access to the population."75 Most state laws do not directly address the question whether, under some circumstances, minors may themselves consent to research participation. An exception is Virginia's general regulatory statute, which provides that informed consent must be obtained from both "a minor otherwise capable of rendering informed consent" and the minor's legally authorized representative.76 Most states have statutes or common law decisions identifying the criteria for a minor to become "emancipated" and so fully able to make his or her own health care decisions (under Maryland law, for example, if a minor is married or the parent of a child).77 Other state laws identify particular conditions for which minors are granted the right to consent to treatment typically, sexually transmitted diseases and substance abuse.78 Laws of this kind might be construed as authority for a minor to consent to expected-benefit research within the scope of the statutory grant of decision-making authority, although the application of these laws to research remains largely unsettled (Hershey and Miller 1976, 118). Oklahoma law is unusually explicit in this regard: Under the circumstances specified in the law, a minor may consent to "health services," which do not include "research or experimentation with minors except where used in an attempt to preserve the life of that minor, or research as approved by an appropriate review board involved in the management of reportable diseases."79
Many states have acted on the view that residents in psychiatric facilities are in need of special protection against the risk of research abuse. As the Secretary of the then-Department of Health, Education and Welfare observed in an early draft of regulations to protect "the institutionalized mentally infirm in research," these individuals "might lack the...capacity to comprehend relevant information, and to make informed judgments concerning their participation" in research.80 In addition, "they experience a diminished sense of personal integrity as a result of confinement in an institution."81 As the federal government struggled, unsuccessfully, to find an appropriate policy response to these considerations (Hoffmann, Schwartz, and DeRenzo 2000), states adopted a variety of provisions. One approach is prohibitory: Missouri law declares that, "No biomedical or pharmacological research shall be conducted in any mental health facility...or in any public or private residential facilities...unless such research is intended to alleviate or prevent the disabling conditions or is reasonably expected to be of direct therapeutic benefit to the participants."82 A more permissive approach, typified by a Montana law, codifies the patient's "right not to be subjected to experimental research without the express and informed consent of the patient, if the patient is able to give consent," and of the patient's guardian or other appointed proxy.83 In addition, the law calls for notice of the proposed research involvement to the patient's next of kin and the attorney who represented the patient.84 "The proposed research must have been reviewed and approved by the mental disabilities board of visitors before consent may be sought. Prior to approval, the board shall determine that the research complies with the principles of the American Association on Mental Deficiency and with the principles for research involving human subjects required by the United States Department of Health and Human Services for projects supported by that agency."85 Michigan's law similarly requires the responsible agency to establish a review process that complies with federal law.86 Some state laws, like South Dakota's, merely require a review process without standards: "No person may be the subject of any experimental research or hazardous procedure unless the research or procedure is approved and conducted in the manner prescribed by the secretary of human services."87 Delaware's law on pharmaceutical research involving patients in its facilities is particularly interesting as a model of a detailed protective regime. Among its requirements are that research be limited to patients who can give informed consent;88 that capacity to consent be determined by, and the consent process itself be observed by, a "health care professional who will receive no financial benefit from the research";89 and that a patient "participating in double blind research shall be advised both verbally and in writing that the patient may receive a placebo for the duration of the research instead of medication. The term placebo shall be fully defined both verbally and in writing."90 The informed consent requirement may be waived only if "no accepted pharmaceutical or other therapy exists for the type of illness affecting the patient or the patient has not responded to accepted pharmaceutical or other therapies,"91 and only then with court approval.92 Involuntarily admitted patients are ineligible.93 Patients who are eligible are to be diagnosed prior to study entry and are to be monitored by "psychiatrists who will receive no financial benefit form the research."94 Finally, pharmaceutical research may be approved only by a two-third affirmative vote of the IRB.95
State laws range from a prohibition in Texas96 to general requirements for informed consent by the resident or legally authorized representatives (District of Columbia97 and Montana98) to a detailed regulatory regime that replicates key elements of the Common Rule (California99).
Many states have codified a nursing home residents' "bill of rights." These statutes or regulations invariably include a "right to refuse to participate in experimental research"100 or a requirement that informed consent be obtained prior to research participation.101 Rhode Island and Wisconsin have extended similar protection to home health care patients;102 Maryland, to patients in ambulatory care facilities;103 and New York, to hospice patients.104
States provide tangible economic support to the research enterprise in a variety of direct and indirect ways: for example, grant programs to encourage private research on particular topics of concern,105 tax abatements and other economic development incentives for research enterprises or facilities,106 and mandated insurance benefit laws that encompass clinical trials or other research.107 These nonregulatory aspects of state law are beyond the scope of this paper.
About a dozen states have enacted "controlled substances therapeutic research acts," which create an exception to narcotics control laws for marijuana use in certain cancer- or glaucoma-related research.108 Given the narrow scope of these laws, they warrant no further discussion.
States also facilitate research through their regulation of dead bodies. State anatomical gift acts uniformly allow cadaver or organ donation for research purposes, although variations from state to state may affect access to tissue under some circumstances. These laws, as they affect human stem cell research, were discussed in Chapter 3 of the Commission's report (National Bioethics Advisory Commission 1999, 32) and in two commissioned papers (Andrews 2000; Kinner 2000).
Of greater interest as the Commission considers the regulatory system as a whole, and the focus of this portion of the paper, is the interaction of state privacy laws and research. In what respects do state laws affect the acquisition of information for research purposes or research information itself?
hold vast amounts of demographic and health data of potential value to researchers. Every state is a repository of birth, death, and similar demographic information. For public health reasons, every state mandates reporting of some communicable diseases. Largely to aid research, most states have created cancer or other disease-specific registries. For health planning purposes, some states require health care providers or third-party payers to report details of patient encounters. Other collections of state records (for example, files on state employees) might yield important information for researchers.
Whenever a state assembles a pool of personally identifiable and highly sensitive information, the state recognizes its obligation to protect the privacy of the individuals. Although the protective measures vary widely, and some are open to criticism as inadequate (Pritts, Goldman, Hudson et al. 1999), virtually every law that mandates or authorizes the collection of the information also requires that personally identifiable information be kept confidential.
Yet, confidentiality requirements could vitiate the very benefit to research that is a primary or secondary goal of these data collection efforts. To avoid this effect, the laws that mandate confidentiality invariably contain an authorization for access by researchers if the researchers bind themselves not to redisclose personally identifiable data. Vital records laws typically have a broadly phrased research exception to their confidentiality requirements.109 Disease registry and mandatory reporting laws allow access for research purposes.110 To cite but one example, the Illinois law that mandates the reporting of Reye's Syndrome cases allows individually identifiable data to be made available for "health-related research" to a researcher who agrees not to redisclose the data.111 Finally, public records laws, which always exempt individual health and other sensitive information from mandatory disclosure, usually allow researchers access to this information.112 For example, all government records in Georgia that by law are "confidential, classified, or restricted may be used for research purposes by private researchers" if a researcher is deemed qualified, the research topic "is designed to produce a study that would be of potential benefit to the state or its citizens," and the researcher agrees to protect the confidentiality of the records.113 Apart from its confidentiality requirement, this broad authorization for research access, like comparable laws elsewhere, does not impose any requirements related to the protection of human subjects.
For some records, a research exception to state confidentiality laws derives from a federal model. The federal law that requires substance abuse patient records to be kept confidential, for example, permits certain research-related disclosures without patient consent.114 So, too, do the federal regulations governing the confidentiality of child abuse and neglect records.115 Many state laws parallel the federal exceptions.116
Much research would be foreclosed if investigators were unable to obtain baseline clinical data about the subjects. Of course, subject consent for access to medical records should be obtained if possible. State lawmakers, however, have by and large accepted that patient consent may not always be feasible and have accommodated researchers by building into medical records privacy statutes an exception for researchers' access without consent. Sometimes the exception is just that open-ended. Rhode Island law, for example, permits a health care provider to release confidential health care information without a patient's consent to qualified personnel for the purpose of conducting scientific research, provided that information about individual patients is never disclosed by the researchers.117 Other laws impose the useful check that the research be approved by an IRB. Maryland's medical records act, for example, permits disclosure without patient consent "for educational or research purposes, subject to the applicable requirements of an institutional review board" and to an agreement not to redisclose any patient identifying information.118
State laws governing HIV testing typically provide that testing is to be done only with informed consent and after appropriate counseling. Some states have enacted an exception for anonymous testing done for research purposes.119 Maryland exempts certain research projects from mandatory reporting by unique patient identifying numbers, but only if, among other requirements, the research is either related to HIV vaccine development or "is not primarily intended to provide medical treatment to participants" and has been approved by an IRB.120
The "certificate of confidentiality" procedure under federal law allows for sensitive, individually identifiable research data to be protected against compelled disclosure.121 A few states have enacted their own comparable protections. The Georgia legislature, for example, has found that "protecting the confidentiality of research data is essential to safeguarding the integrity of research in this state, guaranteeing the privacy of individuals who participate in research projects, and ensuring the continuation of research in science, medicine, and other fields that benefits [sic] the citizens of Georgia and other states."122 To that end, this law protects "confidential raw research data" against compelled disclosure.123 New Hampshire law also deems "personal medical and /or other scientific data of any kind whatsoever obtained for the purpose of medical or scientific research" by the state health commissioner or authorized researchers to be confidential, not admissible in evidence and to be used "solely for medical or scientific purposes."124
Hundreds of state laws authorize or direct state agencies to conduct epidemiologic or other public health research. Most common are the ubiquitous laws that require reporting to state health authorities of various communicable diseases. These laws then authorize the use of the information for public health purposes. The statutory language is usually capacious enough to encompass not only classic epidemiologic investigations, aimed at protecting the public health by identifying the source of a disease that has recently spread quickly among a group of people, but also epidemiologic research in the regulatory sense, aimed at the acquisition of "generalizable knowledge" about the causes and transmission of a disease within a population.125 In Illinois, for example, AIDS cases are to be reported to an AIDS Registry, and the Illinois Department of Public Health may collect "such information concerning those cases as it deems necessary or appropriate in order to conduct thorough and complete epidemiological surveys of AIDS...in Illinois, and to evaluate existing control and prevention measures."126 Depending on their purpose and design, these "surveys" and "evaluations" might or might not be "research" under the Common Rule's definition.
Other state laws, when describing the overall public health responsibilities of state officials, use language that either explicitly authorizes "research" among other permitted activities or is phrased broadly enough to encompass it. An example of the former is this Georgia authorization: "The Department of Human Resources and county boards of health are empowered to conduct studies, research, and training appropriate to the prevention of diseases and accidents, the use and control of toxic materials, and the prevention of environmental conditions which, if permitted to develop or continue, would likely endanger the health of individuals or communities."127 An example from Maryland illustrates the way in which a broadly worded provision can authorize research implicitly. The Maryland health secretary has a statutory duty to "investigate...[t]he causes of disease and, particularly, the causes of epidemics."128 To investigate the cause of an epidemic may well not be "research," because its design may not constitute "a systematic investigation...designed to develop or contribute to generalizable knowledge;"129 to investigate the causes of disease may well be. The statutory authorization covers both.
None of the three laws cited here is linked to any requirements for human subject protection. This omission is not atypical; the laws that authorize state-conducted research usually do not specify a procedure for review of the research or impose a requirement for informed consent. Presumably, state researchers seek to follow prevailing ethical norms, but state law ordinarily does not mandate any specific procedures.
One notable exception is Florida's approach to research conducted under the authority of the Florida Department of Health. Human subject research is subject to review and approval by the Review Council for Biomedical and Social Research, a nine-member body appointed by senior state officials (the Governor, the President of the Senate, and the Speaker of the House) and consisting of three members "knowledgeable in biomedical research," three members "knowledgeable in behavioral research," and three members "from the client advocacy community."130 The Review Council is to be "guided by the ethical standards for human research set forth in the [Belmont Report]."131
The earliest case law on "research" practices involved physicians' ad hoc experimentation in the course of clinical care. As a New York court observed more than a century ago, when standard therapy exists, "there should be no departure from it, unless the surgeon who does it is prepared to take the risk of establishing by his success the propriety and safety of his experiment."132 These cases are based on the premise that, in the clinical setting, the patient is entitled to expect the skillful application of potentially efficacious standard therapy. "Use of an unproven method of treatment which damages the patient has generally been considered negligence, even if carried out with the highest possible degree of care, unless it is clear the patient knew that it was research or innovation" (Holder 1978, 738; Goldner 1993).
With informed consent, the use of experimental procedures is permissible.133 Conversely, "A physician who uses a patient as the subject of an experiment of any sort without a full disclosure of all of the risks involved and the nature and purpose of the investigation, especially in a situation in which consent might be obtained by duress, express or implied, may be in serious legal and disciplinary difficulties" (Holder 1978, 742).134 Professor Holder's reference to "disciplinary difficulties" is illustrated by the aftermath of the research scandal at the Jewish Chronic Disease Hospital. The physician-investigators' decision not to inform patients of the injection of live cancer cells was the equivalent of a deliberate misrepresentation of a material fact, the kind of unprofessional conduct that can result in discipline (Katz 1972, 6063).
Indeed, when research is subject to federal or state informed consent regulations, the breach of those regulations is likely to be viewed as itself a basis for liability. In Daum v. Spinecare Medical Group, Inc., a recent California case, the plaintiff alleged that his physicians had failed to obtain his informed consent for implantation of a experimental spinal fixation device. According to his version of events, he was not informed orally that "the surgery was part of an FDA-approved clinical investigation" of the device.135 The investigational status of the device, however, together with an account of risks and alternatives, was set out in a written consent document, presented for signature on the morning of surgery. In the trial court, experts testified for both the plaintiff and the defendants on the issue whether the informed consent process fell below the standard of care. After the judge instructed the jury that they should decide the case based solely on their assessment of the expert testimony, the jury found for the defendants.
The appellate court reversed, holding that the jury should have been allowed to consider all of the evidence, not just expert testimony, in deciding whether the standard of care had been breached. In particular, the jury should have considered the effect of federal and California informed consent requirements. Because the physician-investigators had agreed to comply with FDA regulations, the regulations themselves established the standard of care. The plaintiff, the appellate court held, "presented sufficient evidence that his injury resulted from the kind of occurrence the statutes and regulations...were intended to prevent: participation in a clinical trial without the subject's fully informed consent in writing, with a copy for the subject and under circumstances permitting a free and deliberate choice."136 Under this analysis, failure to adhere to regulatory standards for informed consent amounts to the tort of "research negligence," the holding in a well-known Canadian case.137 Another California decision, Moore v. Regents of the University of California,138 suggests that some courts will view the informed consent doctrine as encompassing disclosure not only of the protocol-related information specified in the Common Rule but also of material economic incentives affecting the researcher. Moore, arising in a clinical setting, held that a physician who used a patient's surgically removed spleen to establish a patented cell line should have disclosed his research and economic interests to the patient prior to removing the spleen.
One can also speculate (to be sure, without any case law on point) that an aggrieved research subject might under some circumstances bring a breach of contract action against a researcher. This kind of claim could arise if a research protocol were described one way in an informed consent document but carried out another way in practice,139 or, under a third-party beneficiary theory, if research were conducted in a manner inconsistent with a single or multiple project assurance. Whether a plaintiff would have the kind of economic damages that are cognizable in a contract action, however, is open to doubt.
The T.D. case,140 discussed in the Commission's Capacity report (National Bioethics Advisory Commission 1998, 7172), may prove to be the forerunner of another branch of research-related case law: claims that the civil rights of specially vulnerable subjects were violated. With respect to involuntarily hospitalized adult patients and children in state psychiatric facilities, New York's intermediate appellate court held, in a later-vacated aspect of its opinion, that the federal and state constitutions and state common law guaranteed certain protections for no-direct-benefit research, including adequate notice to potential subjects, review procedures for determinations of incapacity, and judicial approval of surrogate consent (Hoffmann and Schwartz 1998; Oldham, Haimowitz, and Delano 1998). An earlier federal case suggested that certain research practices with prisoners as subjects might violate the constitutional proscription of "cruel and unusual punishment."141
The singular title of this paper, "The Role of the States," might better have been phrased in the plural. The regulation of human subject research varies markedly from state to state, from efforts at comprehensive regulation in a few to the imposition of limited protections in others. Their regulatory styles differ as well. Some codify informed consent and independent review requirements without detailed specification of methods; others implement these protections through incorporation into state law of federal standards, sometimes augmented by useful elaboration or specification; and still others add requirements or restrictions that reflect a state's own policy choices. Indeed, in this last category are enactments that may be subject to constitutional objection for example, restrictions on fetal research that may be worded so vaguely as to offend due process standards (Andrews 2000; Gelfand and Levin 1993).142 This diversity is the product of a policy environment in which an incomplete federal regulatory scheme encourages a focus on state-level agendas, be it advocates pursuing restrictive regulation in particular areas or research enterprises seeking to block undesired regulation. Whether in state legislatures or the courts, and whether they turn out to be mild or draconian, sporadic state responses to controversial research practices can be anticipated unless, of course, the Commission were to recommend, and Congress were to pass, a broadly preemptive federal law.
Thus, a fundamental policy issue for the Commission is whether it is content with the prevailing relationship between federal and state law, in which federal standards are a floor, requiring compliance with both the federal standards and state law, if the state law is more protective of subjects; or whether the Commission believes that federal law should be both a floor and a ceiling, reflecting a conclusive judgment about the extent to which the research enterprise is to be subject to regulation, a judgment that should not be diminished by state-to-state variation. Were the Commission to adopt the latter position, it would need to consider carefully a federal constitutional issue beyond the scope of this paper: the extent to which Congress, exercising its power over interstate commerce, or, by delegation, an Executive Branch agency, may preempt state law in this area (Andrews 1997).
This paper assumes, however, that the Commission will accept as inevitable, and perhaps even desirable, a continuing role for the states in an overall regulatory scheme. If so, the Commission may wish to consider the following policy issues:
A comprehensive solution to the problem of unregulated research is for Congress to enact a law that extends the protections of informed consent and independent review to all research in this country, whatever its funding source or site. Of course, Congress has not done so, and it is difficult to predict when, if ever, it will. Therefore, the Commission may wish to consider endorsing an alternative, albeit far more cumbersome, means of extending these protections: state-by-state enactment of general regulatory laws, like New York's or Virginia's. If these laws contained appropriate exemptions, they could help fill the regulatory gap while avoiding duplication of existing federal regulation. Such an initiative might be furthered if the National Conference of Commissioners on Uniform State Laws, a well-respected source for state legislative initiatives, developed a model research regulatory act. Alternatively, states might be encouraged, when they regulate discrete areas of research or research involving vulnerable subjects, to adopt federal standards, a strategy that has the salutary effect of applying known standards, with a history of implementation, to privately funded research. Examples include the New York and Oregon laws on recombinant DNA research143 and the California law on research in facilities for those with developmental disabilities.144
Presumably, the Commission views favorably the basic policy decision, made by the states in a myriad of contexts, allowing researchers to gain access to otherwise confidential health information, whether maintained by state agencies or private health care providers. Nevertheless, overly permissive state laws not only risk compromising individual privacy but also reflect a lost opportunity to gain better protections for subjects. If more states followed the example of some and established as a prerequisite for access that the research be approved by an IRB, the policy goal of extending independent review to all research would be furthered. The Commission may wish to consider recommending this approach to the state legislatures. It is noteworthy that the National Conference of Commissioners on Uniform State Laws has included an IRB review requirement as part its Uniform Health-Care Information Act.145
Assuming that the federal policy decision about state regulation remains as permissive as it now is, the highly diverse landscape of state regulation surveyed in this paper will persist. The Commission may wish to consider encouraging DHHS to establish a clearinghouse on state regulation of research. The clearinghouse would be a source of information for state legislators considering new proposals, for federal research managers considering the implications of state regulation on their priorities, and for researchers themselves. The existence of a central source of information might itself promote greater consistency among the states.
Some federal regulatory systems incorporate shared enforcement responsibilities between federal and state (or local) agencies. Perhaps the most notable example is in the enforcement of equal employment opportunity laws, which involves "work-sharing" agreements between the Equal Employment Opportunity Commission and designated state and local agencies (Lindemann and Grossman 1996, 12211223). Other examples may be found in federal environmental laws.146 Analogies to these highly detailed regulatory schemes are imperfect at best, but they suggest the merit of exploring cooperative agreements to augment federal oversight of research. For example, it is not difficult to imagine a contract between DHHS and a state legislative audit agency under which the auditor's performance audit of a state research university or a state health department would include a special focus on compliance with standards for human subject protection. While many practical problems would have to be addressed to make such an arrangement feasible, the Commission may wish to consider recommending that approaches of this kind be explored.