Maryland State Proposed Statute For Research Using "Decisionally Impaired" Individuals: Beneficence or Abuse?

Dianne N. Irving
Professor of the History of Philosophy, and Medical Ethics
Departments of Philosophy
The Dominican House of Studies, and
The Catholic University of America
Washington, D.C. 20017
Copyright May 28, 1997
Reproduced with Permission

[Outline of speech delivered to Conference of The Law and Health Care Program of the University of Maryland School of Law, "Conducting Medical Research on the Decisionally Impaired", University of Maryland School of Law - Baltimore, MD; May 28, 1997]

I. Conclusion, Up-front

Before I run out of time, let me state explicitly my conclusions concerning the May draft for proposed legislation for the use of the "decisionally impaired" in medical research. I find this proposal grounded on an array of problematic and still very controversial ethical (and other) assumptions - and thus ill-conceived to begin with. As old Aristotle once said, "a small error in the beginning leads to a multitude of errors in the end". Legislating for such research on the grounds presented in this proposal would clearly and translate into great abuses to a multitude of cognitively impaired patients in the end - and therefore in essence most of the kinds of research promoted in this proposal would be unethical. No human being has a duty to be altruistic, or to advance scientific knowledge.

In short, because cognitively impaired patients or individuals are human persons, with the same inherent rights as all human beings; because no human beings should be used as unethical means to any goals - no matter how lofty; because of the obvious vulnerability of cognitively impaired persons to exploitation and their compromised ability to give truly meaningful informed consent, and our inability to know certainly their desires to participate in medical research; because of the past and on-going abuses involving research with the mentally ill; and because the present system of protections for the mentally ill in research are insufficient and inadequate and thus practically incapable of protecting them, only research with minimal risks, explicitly delineated, and for the direct benefit of a particular patient could be ethically consented to, and then only if the research holds out at least as much direct benefit as available standard medical therapies, and only as consented to by a legitimate legally authorized representative. In any other research, participation should be authorized by a court of law. This does not mean that research in mental diseases or any other mental disorders must come to a grinding halt - only that the process of "scientific progress" must be ethical as well as the intended goal, even if it takes a little longer to get there.

The use of an advanced directive for research using cognitively impaired patients is also rejected, including the "substituted judgments" of any agent or other persons noted in a directive specifically for research, as well as that directive's equation with an advanced directive for standard medical care. Simply because medical benefits might possibly be obtained while participating in a therapeutic research protocol does not mean that that research is anything other than research, or that it somehow equals or is equivalent to standard medical practice or "health care".

II. Controversial Assumptions

Frankly, there seems to be a rather naive understanding of ethics presented here, and an assumption that bioethics Ð including its principles of autonomy, beneficence and justice (or "principlism") - is "ethics-pure-and-simple". But there are dozens of kinds of ethics besides the kind of ethics ingrained in this proposal. It might be objected that bioethics "principlism" is perfect for any public policy, since it was formulated specifically as a sort of "neutral ethics" which would be perfect for our pluralistic society. However, there is no such thing as a "neutral" ethics - even for public policy - ask any academically credentialed ethicist. All ethical theories are normative - that is, they take a stand on what is right or wrong. And despite the mantra in bioethics, even so-called "meta-ethical" theories, which are referred to as non-normative ethical theories, are clearly prejudiced and contain multiple philosophical, logical and other distinct presuppositions - as any historian of philosophy would acknowledge. Thus the search for a "neutral" ethics is simply wishful thinking. Bioethics principlism, utilitarianism, deontology, consensus ethics and communitarianism (which could easily be another form of either utilitarianism or consensus ethics - or both) are all normative ethical theories - and all fraught with problems which have constantly been difficult to defend in philosophical academic settings. Indeed, why should bioethics principlism, utilitarianism, consensus ethics, or communitarianism be "imposed" on the rest of the citizens of the State of Maryland or any other form of our government, given that none of these ÒethicsÓ are actually "neutral"?

Additionally, the Working Group seems oblivious to the fact that bioethics as a field has come under serious fire for several years now. In fact, the bioethics principles of autonomy, beneficence and justice - which explicitly grounded the October draft - have been declared "dead" even by those who crafted them 20 years ago in the Belmont Report (e.g., Al Jonsen essentially admitted in an article that the National Commission just "made them up"), and by those involved in the leadership of bioethics "think-tanks" (e.g., Daniel Callahan at the Hastings Center - who published his uncomplimentary judgment of "principlism" several years ago). The bioethics literature is replete with books and articles, nationally and internationally, which at least make it very clear that there is certainly no consensus on the theoretical viability of the very grounding of bioethics itself. This in turn throws a shadow over the very viability of all of the documents which have been explicitly and enthusiastically grounded on bioethics principles for the past 20 years - including the OPRR regulations for the use of human subjects in research themselves. In essence, at the moment, bioethics "has no clothes".

Another more subtle assumption constantly used throughout these and similar regulations and proposals is the now-popular use of the "risk/benefit" ratio. It would seem that this is simply a ÒreasonableÓ and ÒmathematicalÓ way to determine what action would be ethically acceptable to take in situations such as we are discussing here today. But in reality the use of a risk/benefit ratio is rooted in a utilitarian theory of ethics - where the good consequences would out-weigh the bad consequences. This has exponentially led to its blind acceptance throughout public policy making in research. But, as noted, that is only one of many kinds of normative ethics - one which only considers the consequences, or in this context, benefits or risks of participating in research. Individual patients are lost in the equation; only the ÒmajorityÓ are covered.

But it is not enough for the end result to be good or beneficial. Utilitarianism fails to grasp that the means used to reach that end must also be good. In short, it is not sufficient that the goal of a research protocol be beneficial. The human subjects, who are the means by which the goal of the research is reached, must be treated ethically as well. To intentionally involve human subjects as means in research in which harm could result is unethical unless they have given their informed consent. Otherwise they are simply being used as biological material for someone else's ends or goals. Additionally, there are even some research protocols which are inherently wrong, even though a human subject has given his/her informed consent. Yet the common use of the risk/benefit ratio has lead, interestingly enough, to a kind of blind ethical relativism, i.e., that there is essentially no kind of research which is not ethically acceptable to perform, as long as a patient consents to it, and the benefits outweigh the risks.

Finally, the list of controversial assumptions which I will explicitly address with reference to the proposed legislation includes - but does not by any means exhaust - the following: "substituted judgments", the definitions of the risk categories, the definition of the benefit categories, several research categories on which the proposal is silent, research that is specifically rejected or accepted, and legal authorization and liability for harm caused the patients during their participation in research.

III. Possible Sources of Abuse to Research Participants

A. October 6, 1995, draft

Before turning to the present May draft, I want to turn to the previous October one, simply as a quasi-informative example of how research with human subjects has usually been discussed and debated. It is well to remember throughout these comments that we are talking about "decisionally impaired" persons, whose "competency" could be determined by the same researchers who are to perform the experiment. These patients would then be considered to be capable of giving "informed consent" in their Advanced Directives, to choose a Research Agent or a Health Care Agent who would use their "substituted judgments" as to what research the patient would have wanted to participate in, if he/she were still "competent".

1. "Substituted Judgment":

Note the liberal use of "substituted judgments" throughout the proposed bill. Even as controversial as they are, the use of "substituted judgments" was accepted by some bioethicists early on only if a competent patient had not specified in his/her Advanced Medical Directive exactly what kind of standard medical care he/she would want if they were later to become incompetent. There have been and are several legal challenges for even this use of "substituted judgments". In consent for research - either therapeutic or experimental - involving an incompetent or "decisionally impaired" patient -- the use of a "substituted judgment" is highly controversial, and one on which there is no consensus - at least within bioethics or the law.

Question: How can a Research Agent (RA) or a Health Care Agent (HCA) be reasonably sure that even though a patient gave "consent" for a particular protocol in the past that the patient still wishes to participate in any research again, or in this particular protocol again? Couldn't patients change their minds about participating between then and now? Couldn't they change their minds about the amount of benefit, or about the degree of risk, to which they might want to be exposed between then and now? It is surely possible that a research protocol in which they had previously participated was an unpleasant experience to them - benefits not withstanding. Once a patient is "stuck" in an AdDir for research, is there no way out?

2. Risk Categories:

- There are three sub-categories of risks or harms: minimal risk, minor increase over minimal risk, and more than minor risk over minimal risk - the latter two categories usually having been considered high risk research (the term which I will continue to use here).

- Sometimes the October draft expresses these standard categories; sometimes the last two categories are simply conflated to "greater than minimal risk". However, it should be remembered that "greater than minimal risk" would include high risk research.

Question: If risk categories are inconsistent, vague or unclear, then how can a patient - or any agent, surrogate or monitor - try to "balance" these vague "risks" with any benefits in order to reasonably determine if the possible benefits outweigh the risks involved in participating in the research? Such a utilitarian balance of risks and benefits is incapable of being calculated.

Question: If it is not reasonably possible to determine a sound risk/benefit ratio, than how can any agent, surrogate or monitor reasonably give a meaningful and valid "substituted judgment" that the patient would have wanted to participate in the research? How is any Òinformed consent" even possible?

- The language referring to "research" in this October draft does state explicitly and consistently: "participate in this particular protocol". It does not refer to "research-in-general", but specifies the kind of research in which the patient wanted to participate.

- This draft also uses clear language in which all research risks are risks to which only individual patients could be exposed if they were to participate. Risks to "society-at-large" or to "classes of future patients with the same or similar disease or illness" are not part of this October draft.

3. Benefit Categories:

- Likewise, any possible benefits to participating in the research would be benefits directly to that particular patient who was being enrolled in a specified research protocol. There is no mention in this October draft of benefits to "society-at-large", or to other patients in the future who belong to the same "class" of patients or who might benefit from the research, even though this particular patient would not benefit.

4. "Silent" Categories:

- The October draft is "silent" on - i.e., does not even mention, discuss or articulate - direct benefit/minimal risk research [as consented to by a RA, a HCA, or a Surrogate (S)].

Question: If a draft is "silent" on any issue, does that mean that "anything goes", that it can't be done, or what? Couldn't "silence" in effect be a "loop-hole" by which researchers could perform any research which is not explicitly mentioned, regardless of risks or to whom the benefits would flow, and by which IRB's could get away with approving such research? This could clearly be a source of harm to participating patients.

5. Specifically Rejected Research:

- The October draft indicates unequivocally that no research is ethically acceptable in seven different categories:

a) A surrogate or monitor, with an Advanced Directive (AdDir) for research, cannot consent to a particular protocol if the research has no direct benefit to the patient (in both high risk categories).

The rationale used to come to a consensus was: "The Working Group concluded that reliance on the surrogate's 'substituted judgment' would be insufficient given the absence of potential for direct medical benefit and the degree of risk." (p. 13)

b) A monitor, with no AdDir for either research or for standard medical care, cannot consent for a particular protocol if the research has no direct benefit for the patient -in all three risk categories - even for minimal risk research.

Thus if research could cause "greater than minimal risk" with no direct benefit to the patient, the rationale by which the Working Group came to a consensus was: " on an advance directive alone ... would not be justified ethically. ... (H)ere no one has intimate knowledge of the individual's values and attitudes. The bare document alone is insufficient to expose the individual to research involving this degree of risk without the prospect of direct medical benefit." (p. 14)

c) If there is no advance directive - either for research or for standard medical care - a monitor may not consent even for a minor increase over minimal risk. As the Working Group concluded: "In the absence of any written indication of the individual's wishes regarding research and in the absence of anyone with intimate knowledge of the individual, the Working Group believes that research lacking a reasonable prospect of direct medical benefit should be ruled out, regardless of the degree of risk.." (p. 15)

d) High risk research with no direct benefit is not authorized because, in the words of the Working Group: "Respect for persons absolutely precludes research participation under these circumstances." (p. 15)

6. Acceptable Categories of Research:

Research could be authorized in all other categories by an RA or a HCA, even high risk research, with or without direct benefit to the patient - and even by a Surrogate in direct benefit/high risk research, with no AdDir (research or standard medical). In fact, a "Monitor", with an AdDir for research, could authorize direct benefit research in both higher risk categories.

7. Legal Authorization and Liability:

Note that surrogates are not now required to be authorized by a court of law. In the Attorney General's July 26, 1995 letter [page 2] distributed with the October draft, it would seem that court involvement is considered "unnecessary", and that this opinion should be translated into law: "If ultimately we think that health care agents or surrogates ought to have the authority to consent to non-therapeutic research under limited circumstances, and if researchers ought to enjoy immunity for acting upon such a consent, we should recommend a change in the law". (Jack Swartz, July 26, 1995 letter, p. 2 - to the Secretary of Health and Mental Hygiene). However, there is still court authorization required in this draft if a monitor, with no advanced directives at all, wishes to enroll a patient in any research which would have no direct benefit to that patient.

Question: If there is now very little involvement of the courts, and if patients are seriously harmed while participating in the research, is Justice truly served if injured patients have no recourse to the courts, while researchers and consenting agents or surrogates continue to "enjoy immunity for acting upon such ['substituted'] consents"?

B. The May 5 draft, as proposed

1. "Substituted Judgment":

- Note again the liberal use of the very controversial issue of "substituted judgment" throughout the May draft - and here specifically within the context of research.

2. Risk Categories:

- Now, risks refer not just to an individual patient. If the risks are involved in "direct benefit" research, it can now also be understood as risks and benefits for the "class" of persons of which the particular patient is a member with the same or similar disease or illness. That is, in this case the direct benefit need not be to the individual patient who is taking part in the research, but can also be only for the direct benefit to future patients in the same "class".

- There are virtually no risk categories articulated if the research might possibly be directly beneficial to a particular patient or future patients in the same "class".

- Only research with no direct benefit is broken down into the usual risk categories of minimal risk and high risk (or, "greater than minimal risk") research.

Question: If risk categories are not specified in direct benefit research, then how can a patient - or any agent, surrogate or monitor - try to "balance" these "silent" "risks" with benefits in order to reasonably determine if the possible benefits outweigh the "silent" risks involved in participating in the research? Again, risks and benefits are incapable of being calculated.

Question: If it is not reasonably possible to determine a sound risk/benefit ratio, then again, how can any patient - or any agent, surrogate or monitor - reasonably give a meaningful and valid "substituted judgment" Ð or even Òinformed consent"?

Question: Isn't it a bit overly unrealistic to expect a patient to altruistically accept risks for "classes of future patients" who they do not even know - even high risk research in which they receive no benefit? Are these patients being pressured to participate in the research by disingenuously appealing to their altruistic sense of "duty to future patients"?

- The language referring to "research" in this draft no longer specifies "participating in this particular protocol" - but rather uses the general language of "participating in research".

Question: Should any patient even be allowed to sign an AdDir for research, when that person has no knowledge of what kind of research protocol might be involved? Is that really informed consent? And can a patient legally sign off for anything - even for such total unknowns?

Question: If an AdDir for research does not specify a particular research protocol, how can a RA, HCA or S give a "substituted judgment" for a patient who didn't even know what specific kind of research protocol he/she wanted to participate in Ð or even Òinformed consent"?

3. Benefit Categories:

- Benefit categories are clearly broken down into research that could possibly be for:

Question: If the usual benefit categories are all merged together, with no distinction according to degree of benefit involved - including research involving high risk - then --- you guessed it --- how can a patient - or anyone - try to "balance" these merged categories of benefits with risks in order to reasonably determine if the possible benefits outweigh the risks involved in participating in research? Once again, such a utilitarian balance of risks and benefits is incapable of being calculated.

Question: If it is not reasonably possible to determine a sound risk/benefit ratio, than how can any agent or surrogate reasonably give a meaningful and valid "substituted judgment" that the patient would have wanted to participate in the research Ð or any Òinformed consent"?

4. "Silent" Categories:

- The draft is "silent' on 5 of the 7 categories of research noted in the October draft - all 5 categories involving research with no direct benefit, and high risk research.

- In 4 of the 5 "silent" categories, the October draft had explicitly rejected consent by a surrogate or a monitor in no direct benefit research in which there was a prospect of high risk to the patient.

- Where in the October draft a HCA with an AdDir for medical care could consent to no direct benefit/high risk research, the May draft is now also silent.

Question: What does this "silence" mean? That those earlier rejected categories of research are now acceptable? Why?

5. Specifically Rejected Research:

- There are now no categories of specifically rejected research.

6. Acceptable Categories of Research:

- Given the "silence" on much of the categories in the May draft, it would appear that most if not all categories of research could be acceptable.

7. Legal Authorization and Liability:

- There is now no involvement of the courts at all in any of this research. And there is no liability to any one as well. As long as the cognitively impaired patient has given his/her "informed consent" in an advance directive - research or standard medical - a host of agents or surrogates can now act on that directive through the use of their "substituted judgments".

It doesn't take much of an imagination to realize how much harm and abuse could be caused to cognitively impaired patients by such ubiquities. However, it may be rather difficult to visualize in the present format. To see the several potentials for abuse to patients more clearly and quickly, I have re-categorized the May draft into the same format as the previous October one.

C. May Draft Re-Categorized

- If I superimpose this newly-formatted May draft onto the October one it is much clearer that the Òno direct benefit/high risk" categories which are now "silent" in the May draft actually coincide with most of the same categories of research which were soundly rejected in the October draft.

- Note also that three other categories of Òno direct benefit" research, clearly rejected in the October draft (regardless of degree of risk), are now also all "silent" - even though those categories had required authorization by a court of law. In fact, this entire bottom line of categories is not even mentioned or discussed in the present May draft. I will return to this particular point in a moment.

IV. Three Representative Cases

Let me use a few hypothetical cases which demonstrate quickly what could be legally authorized by the State of Maryland if the May draft is voted on and passed into law. It would also be a means of highlighting just a few of the ethical problems involved. According to the May draft:

Case 1:

Why is this a realistic possibility?

a) Because the patient has signed an AdDir for research-in-general.

Question: Isn't it a fact that Advance Directives even for standard medical care are both ethically and legally controversial and have led to significant harm to persons when their wishes could not be known under different circumstances, or because they had changed their minds but not their AdDir's?

b) Because the patient does not need to know which specific protocol, even if he/she had never participated in a protocol like the one being considered here before.

Question: Even with an AdDir for research, how can a patient or agent give a valid "substituted judgment" when they can't know what kind of protocol the patient would have wanted to participate in? Isn't this a substituted judgment about a substituted protocol?

c) Because the patient is being allowed to "exercise his/her autonomy".

Question: Could this be a clear abuse of the bioethics principle of autonomy - and a prime example of why the abuse of this principle was one of the major reasons why many of the most respected leaders in the field of bioethics have rejected bioethics principlism all together?

d) Because the monitor was appointed by an IRB.

Question: Isn't there a real possibility of a conflict of interests if members of the IRB or the monitor have a stake in this particular research protocol?

e) Because risks in all of the direct benefit research are now sanctioned, regardless of any degree of risk; and because the draft is silent on 5 categories of high risk research.

Question: As noted previously, wouldn't it be impossible for anyone, including the RA, to balance the risks and benefits inherent in this research, and therefore to give a valid "substituted judgment" that the patient would have wanted to participate in the research?

Case 2:

A HCA with only an AdDir for standard medical care can enroll a patient in all categories of beneficial research - even if the benefit is only to the "class" of patients and not directly beneficial to the patient. And this includes research which is in the highest risk categories. (Since now the May draft is "silent" on the no direct benefit/highest risk categories, perhaps the HCA can enroll a patient in those kinds of research as well.)

Why is this a realistic possibility?

a) Because now "standard medical care" has been literally equated with "health care" which has been equated with "therapeutic research" - directly in the October draft and thus indirectly in the May draft. If a research protocol might provide some "medical benefit" it is thereby considered to be "standard medical care", or "health care".

Question: Is it really accurate and true to equate "standard medical care" with "therapeutic research" or "health care"? I think not. Would the American Medical Association, or any other medical associations, agree with such a drastic degree of medical research deconstruction? Additionally, this "equation" flaunts virtually all of the international research guidelines - including the Nuremburg Code, the Declarations of Helsinki, and the CIOMS/WHO international guidelines. In fact, it is difficult for me to believe that it accords with the present OPRR regulations and the Common Rule, even as lax as they are. The abuse potential in just this single slip of the language could cause incalculable harm to great numbers of cognitively impaired patients. I question whether even competent patients are legally allowed to take part in such research.

b) Note that according to the OPRR regulations, the director of NIH can waive any and all of these categories and restrictions - as well as any and all OPRR regulation requirements - if he wants to. An IRB, with the sanction of the Director of NIH, can also have any restrictions in the May draft waived - for the "advancement of scientific knowledge", "if the information can not be obtained in any other way", or perhaps even for "national security" reasons.

Case 3:

This case represents one of particular concern to me - albeit hypothetical - and possibly an inappropriate extension of the logic. But if we look at the proposed amendment to the now-existing Health Care Decisions Act (as presented in Appendix B), I suggest this case in terms of a final question:

Question: Could any persons, other than the RA, HCA, S or M, enroll cognitively impaired persons in low-to-high risk/direct or no direct benefit research - even if the benefit is only to a future "class" of cognitively impaired persons - if none of the above "list" of consenters is available?

Recall the bottom of the October chart which defined research about which the May draft is silent. And consider Appendix B in the May draft which concerns an amendment to the now existing Health Care Decisions Act:

- If the present HCD Act is amended with respect to "surrogate" consent as described in Appendix B, and

- If guardians appointed by the court, spouses, adult children, parents, adult siblings, friends, or other relatives may then be substituted as "surrogates" if no real surrogates are available, and

- If all of these "substitute surrogates" could consent just as any regular surrogate could consent for research in the May draft, and

- If, in the May draft, standard medical care is equated with therapeutic research,

- Then, even a friend of a cognitively impaired person (perhaps the lady in the drug store down the street) could enroll him/her in both categories of high risk research with no direct benefit to that patient, based on the patient's AdDir for standard medical care. Furthermore, no one would have any legal liability for any harms or abuses caused to the patient while participating in this research.

Is this the State of Maryland's interpretation of Justice - Justice due Constitutionally to some of the most vulnerable citizens in this State?

V. Conclusions

As well-intended and as hard working as I am sure the Working Group has been, I find this proposed statute ill-conceived and conducive to a great deal of abuse on many levels. And if this statute passes in the State of Maryland, it could set a legal precedent for all our other states. Perhaps it is the stated goal of this endeavor which in essence contains the seeds of my concern. In the words of the October draft, "The goal of this project is that 'for a change, law may be the handmaiden of ethics [I presume they mean bioethics] and ethics served by the law rather than vice versa'". Indications given by both the October and the May drafts imply strongly that bioethics has in fact roundly succeeded in making the law its handmaiden - and in the process it has crushed any prospect of due process, or recourse for serious harms or abuses to our most vulnerable citizens. It is, unfortunately, proof-positive that it does matter which ethics one uses for science and public policy. And it seems to me that despite several important advances, what the Working Group has actually accomplished is to substitute its gbioethics judgment" for the judgment of the rest of the citizens of this State.

[Addendum: The Maryland House passed this proposed statute; it was defeated by 1 vote in the Maryland Senate.]