Ethical Issues in Biomedical Research: Diaz v. Hillsborough County Hospital Authority

Vol. 30 No. 2


Stephen F. Hanlon manages Holland & Knight's Community Services Team, which provides legal representation to people, groups, and causes that otherwise could not afford it. His major civil rights work has included challenges to high-stakes testing; challenges to indigent defense systems; housing, employment and AIDS discrimination; death penalty litigation; voting rights; and unconsented medical experimentation. Robyn S. Shapiro is a partner in Michael Best & Friedrich LLP. She has represented clients with respect to bioethics issues, medical staff matters, health information privacy, informed consent, regulatory and licensing matters, and employment and other business issues. She is the director of the Bioethics Center at the Medical College of Wisconsin.

In Diaz v. Hillsborough County Hospital Authority, 2000 U.S. Dist. LEXIS 14061, Aug. 7, 2000, a group of approximately 5,000 pregnant women brought a class action claiming that while receiving prenatal care from the defendant hospital, they were subjected to research tests without being informed that the tests were for research and without being asked to consent. Settlement of the case included not only substantial payments to the individual women but also the hospital's agreement to change its research procedures. The following discussion relates the experiences and observations of Stephen F. Hanlon, attorney for the plaintiffs in the case.

Facts of the Case

Flora Diaz was a sixteen-year-old girl in her first pregnancy when she appeared at Tampa General Hospital's high-risk clinic. She was in preterm labor, and she knew that something was wrong with her pregnancy. Upon admission to the hospital, she was given either Demerol or morphine to arrest the preterm labor. While in a drowsy state resulting from administration of these drugs, she was presented with a three-page, single-spaced, English-only, second-year-college-readability-level document and told to sign the document or "your baby will die." (It should be noted that under federal law, informed consent documents must clearly disclose the risks, benefits, and alternatives to proposed medical experiments and be written in language understandable to the proposed subjects.) Investigation revealed that in addition to Flora Diaz, 383 mostly indigent women had been "enrolled" in this clinical trial, from which the physician/researchers at the hospital sought to learn more effective ways of treating respiratory distress syndrome resulting from fetal lung immaturity. One nurse who had worked at the hospital for thirty years described the "enrollment" as follows:

I observed many patients who were young, afraid, and confused. I saw many who cried and trembled from their anxiety about the situation they were in. These patients were in pain and fear . . . there were numerous times I observed a patient who was so frightened and disoriented by her fear that she didn't hear or understand a word being said to her . . . I frequently witnessed just such a patient being enrolled in a study.

The standard of care for treating fetal lung immaturity at the hospital at the time Flora Diaz presented was corticosteroid therapy. However, researchers at the University of South Florida Medical School who staffed the high-risk clinic at Tampa General Hospital knew work on rats and rabbits had been done that indicated a combined drug regimen, using thyroid hormones and corticosteroids, might increase fetal lung maturation. They set out to construct a protocol to test the relative efficacy of corticosteroids versus such a combination.

The protocol involved a double-blind, randomized clinical trial in which an equal number of several hundred women would receive either corticosteroids (drug X) or corticosteroids combined with thyroid hormones (drug X + drug Y); neither researchers nor subjects would know which drug was being used. Whether the woman received drug X or drug X + drug Y would be determined randomly. An amniocentesis would be performed initially to determine the L/S ratio, an indicator of fetal lung maturity, and another amniocentesis would be performed every seven days thereafter that the woman remained undelivered. Data to see whether drug X or drug X + drug Y performed better would be assembled and analyzed, and an article would be written for publication in a peer-reviewed medical journal.

The Issue

The most significant issue in the protocol involving Flora Diaz was informed consent: did Flora Diaz and the other women who were part of this research understand that they had a choice not to be enrolled in this experiment, not to be randomized, and simply to take drug X (i.e., corticosteroids alone), which was the standard of care at the institution at the time?

The informed consent document, three pages single-spaced, English-only, at second-year-college-readability-level that Flora Diaz was shown and told to sign or her baby would die did not adequately disclose this choice. The danger in this case and others is that physician/researchers who need significant numbers of subjects to produce reliable data and to publish articles-which results in prestige for themselves and their institutions and, perhaps, in funding for continued research-may be reluctant to disclose the choice to the proposed subjects, fearing the patients will simply opt for the standard of care.

The Law

Federal regulations applicable to informed consent in biomedical research require that the enrollment of subjects take place in a noncoercive atmosphere. Presentation of a complex informed consent document under the circumstances present in the Diaz case to a class of low-income, high-risk pregnant women is in and of itself coercive. Add the administration of morphine and Demerol prior to the "enrollment" choice made by the 384 women in this study in a remarkably short period of time-and the delegation of the duty to obtain informed consent to a large number of physicians, residents, fellows, and nurses-and the result is a situation involving coercion and confusion that is exacerbated by institutional pressures to ignore or override the fundamental concern for truly informed consent in biomedical research.

The Theory of the Case: Liability and Damages

From the outset I was convinced that the case involved violation of a person's basic human dignity. The U.S. Supreme Court's decision in Cruzan, 497 U.S. 261 (June 25, 1990) (just after we filed Diaz), lent support to my theory. In this case, the Court held that citizens in the care of a state hospital had a constitutionally protected liberty interest in refusing unwanted medical treatment and preserving bodily integrity. The Court held: "No right is held more sacred or is more carefully guarded by the common law, than the right of every individual to the possession and control of his own person, free from all restraint or interference of another."

Justice Brennan, concurring and dissenting, anticipated my case and posited a horrific scenario that could result if the state were allowed to put the interest of third parties above the constitutional right to bodily integrity:
Indeed, why could the State not perform medical experiments on her body, experiments that might save countless lives, and would cause her no greater burden than she already bears by being fed through a tube? This would be too brave a new world for me and, I submit, for our Constitution.

The question of damages presented a much more difficult problem. Months of research demonstrated that there was no law on point. However, I was helped by an article written by Marjorie Maguire Schultz that articulated for the first time a legally and constitutionally protectable interest in medical choice for human subjects such as Flora Diaz. 95 YALE L. J. 219 (Dec. 1985). This interest, Schultz argued, is distinct from protection of bodily security (which is secured by the law against unconsented contact) and bodily well-being (which is secured by rules governing professional negligence). Neither bodily security nor bodily well-being coincides with autonomy, argued Schultz. A harm that results from a violation of subject autonomy is dignitary harm and is independently compensable, even in the absence of "actual injury."

The Lawsuit

We filed the Diaz lawsuit on January 31, 1990. It asserted a new claim for dignitary harm: the plaintiff-research subjects were harmed by conduct that overrode their autonomy, treated them as less than human, and denigrated them as human beings. We asserted they were entitled to compensation even in the absence of physical injury to themselves or their babies. If successful, this claim would defeat the standard response of the biomedical research industry to this type of claim, namely, "no harm, no foul."

Understandably, the hospital and the university medical school vigorously defended the claim. The defendants also vigorously resisted class certification and challenged my right to obtain the hospital records of the class members, which disclosed patient identifying information, without the patients' formal, written consent. I prevailed in obtaining the patient identifying records. My research disclosed that patients' state-law right of confidentiality with respect to their medical records is not an absolute right but rather one the courts balance, and that when an institution begins to use its patients' right of confidentiality as a shield for its own behavior, and when the institution rather than the patient asserts the claim, the claim is very weak. In this case, the overriding interest in access to the truth in order to assert constitutional claims ultimately prevailed.

Nonetheless, obtaining judicial approval for access to patient records of members of the class who were not named as plaintiffs presented a difficult and perplexing problem for the Court. After lengthy litigation, we were finally granted access to that information so we could move for class certification. The class certification motion was vigorously resisted, but eventually, in 1996, the widest possible class was certified by the court: all pregnant women who had been subjected to medical experiments since November 1986-a class of approximately 5,000 women. This was the first time such a class had been certified by any court in this country. In our case, it was essential to a successful claim of significant damages.

Finally, the case was set for trial, and we instituted meaningful mediation with the chief executive officers of both defendant institutions. After months of subsequent mediation sessions, this case was settled for $3.8 million. In the consent decree entered in this case, Judge Henry Lee Adams Jr. specifically cited Ms. Schultz's article in support of the theory of "dignity harm."

The Impact of the Settlement

The settlement was widely reported and is now the subject of much discussion in the biomedical research community. It represents the first substantial monetary award to a class of subjects of biomedical research who did not assert a claim of physical injury. I believe it has the potential to significantly affect how biomedical research is conducted in this country, primarily because it did in fact eliminate the traditional no-harm, no-foul response of the biomedical research industry to this kind of claim. The term "dignitary harm" is now recognized in the law. It is alive and well in research institutions and in courtrooms all over this country, which now recognize that violation of subject autonomy, standing alone, is compensable as causing dignitary harm.

Moreover, in a more general sense, the issue of oversight of the informed consent process in biomedical research is now receiving enhanced scrutiny in professional journals, among policymakers, and within research institutions. With specific respect to the defendant institutions in our case, significant institutional reform occurred after we filed both a federal lawsuit and an administrative complaint with the Office of Protection from Research Risks, to the credit of both Tampa General Hospital and the University of South Florida. (These reforms eventually obviated the need for extensive, wide-ranging injunctive relief, so litigation could focus on securing an appropriate award of damages for the class.)

The Lessons Learned

By the end of the case, all parties agreed with the proposition that informed consent is a process, not a document, and that the researcher-subject conversation is more important than the consent form in the process. However, the process starts with the important step of writing the document, which forces researchers and physicians to clarify their thinking in terms of the choices the proposed subject is going to be asked to make about participating or not participating in the specific research project.

All parties also agreed, by the end of the case, that readability alone is not sufficient for an informed consent document. Merely applying a standard readability test to the document is only one step-the document needs to be both readable and effective in communicating information in order to obtain a truly informed consent. A host of other factors-design, cultural relevance, format, length, density, and style-all enter into the question of the document's potential for effective communication.

I am convinced, after my ten years of work in this case, that if a research institution focuses its efforts on developing an informed consent document that communicates information and choices to the proposed human subject as effectively as possible, the atmosphere surrounding enrollment in biomedical research studies will tend to be noncoercive, as required under the federal regulations. I believe that the converse of that proposition is equally true: that a complex and difficult-to-understand informed consent document is conducive to a coercive atmosphere in the enrollment process in biomedical research. The document itself is coercive, intentionally or not, when it is unduly long, complex, and incomprehensible. This type of document sends a message to proposed human subjects that they have no meaningful role in the process because it is something that can be understood only by people with greater knowledge than they possess. Conversely, when the institution engages in a studied attempt to communicate effectively and does that as well as possible under the circumstances, in a written document or a videotape, the proposed human subject and all those involved are immediately put on notice that there is a meaningful role for the human in this process. That role involves adequately understanding described choices and knowingly and voluntarily making those choices.

The focus on the effectiveness of communication resulting from this lawsuit is a healthy one that ultimately will benefit both the biomedical research industry and proposed subjects of human research. To further advance the efficacy of the informed consent process, we must provide opportunity for feedback from the actual subjects of human research and recognize that their participation and input is valuable and necessary.


  • About the Magazine

  • Copyright Information