EDITORIAL Coercion in Research: Are Prisoners the Only Vulnerable Population?

Barbara E. McDermott, PhD

J Am Acad Psychiatry Law 41:8–13, 2013

Research plays an essential role in advancing medical (USPHS) commissioned Dr. Joseph Goldberger to and behavioral sciences and in improving our ability study the problem in an effort to find a cure.1 The to understand and treat illness. However, unlike clin- widely held belief at the time was that pellagra was ical care, which is intended for the sole benefit of the caused by some type of micro-organism. Dr. Gold- patient, research has the broader goal of advancing berger began observational studies and formulated knowledge and does not necessarily provide the sub- the hypothesis that diet was primarily responsible for ject with an ensured clinical benefit, raising obvious the disease. To test this theory, he recruited 11 pris- ethics-related concerns and making voluntary con- oners from Rankin State Prison in Mississippi to un- sent to participate essential. For certain participant dergo dietary modifications in an effort to induce populations, consent may be problematic. Such in- pellagra in these otherwise healthy volunteers. In ex- dividuals include those who may be unable to make change for their participation, the prisoners were informed, voluntary decisions about participation in promised a full pardon. After five months, during research, either because of impairments in cognition which time several prisoners began to develop symp- or reasoning or because they live in an environment toms of the disease, the study concluded, with 6 that is potentially coercive. of the 11 evidencing symptoms of pellagra. During those months, inmates who developed the disease History of Research on Prisoners suffered so greatly that some were willing to forgo the Prisoners are considered a group of individuals pardon to discontinue the study. for whom coercion is a major concern, primarily be- A much more extreme example highlighting the cause of the supposition that prisons are inherently subtlety of the coercive process was the malaria study coercive institutions. Previous research conducted at Stateville Penitentiary in Illinois.2 The prison was on prisoners makes the rationale for this evident. built as a panopticon, meaning that the housing units Historically, prisoners have been considered an ideal were arranged in a circle surrounding a central tower. population on which to conduct research because A guard was placed in the tower but was not visible to they are readily accessible and in a controlled envi- the inmates. Thus, the guard could be watching the ronment. As an example, in the early 1900s, pellagra inmates at any time without their knowledge, impos- had reached epidemic levels in the rural southern ing a subtle form of authority. During World War II, United States. The U.S. Public Health Service malaria was devastating U.S. soldiers fighting in the Pacific Ocean theatre. At that time, quinine was the Dr. McDermott is Professor of Clinical Psychiatry, Division of Psy- chiatry and the Law, Department of Psychiatry, UC Davis School of only effective treatment, although supplies were lim- Medicine, Sacramento, CA. Address correspondence to: Barbara E. ited. Between 1942 and 1945, the United States re- McDermott, PhD, UC Davis School of Medicine, Department of Psychiatry, Division of Psychiatry and the Law, 2230 Stockton Bou- portedly lost eight million man-days to the disease. levard, Sacramento, CA 95817. E-mail: [email protected]. In 1944, the malaria studies began at Stateville Peni- Disclosures of financial or other potential conflicts of interest: None. tentiary. Although inmates were not promised early

8 The Journal of the American Academy of Psychiatry and the Law McDermott release initially, the governor of Illinois eventually tested a polio vaccine on prisoners at the Federal guaranteed that all inmate participants would have Reformatory in Ohio because he believed that the their parole decisions reviewed at the end of their vaccine was not yet ready for the general public; Dr. sentences. It ultimately was determined that inmates Jonas Salk’s use of Michigan inmates in his work participating in the malaria studies were released an with the influenza vaccine; and Sloane-Kettering’s average of two years earlier than were nonpartici- collaboration with Ohio State University and the use pants. At the initial announcement of the effort, 200 of prisoners in Ohio for cancer research that involved volunteers were solicited; 487 men offered their ser- injecting live cancer cells into supposed volunteers. vices. Although the prisoners signed consents for Although the Green Committee encouraged the dis- these studies, no risks were mentioned in the consent closure of risks to each prisoner, there were no pro- forms. According to reports from the infamous in- cedures in place to ensure that such a disclosure oc- mate Nathan Leopold,3 each volunteer was fully ap- curred. Researchers were left with the responsibility prised verbally of the risks; one risk was death. Yet, of assuring that participants knew the risks associated inmates volunteered in increasing numbers. As a re- with the studies and were presumed to police them- sult of these studies, brought to light during the selves in this regard. Nuremberg trials as evidence that the United States In 1962, the thalidomide scandal changed Food had also performed atrocious medical studies on pris- and Drug Administration (FDA) regulations regard- oners, the Governor of Illinois (Dwight Herbert ing drug testing. Thalidomide was routinely admin- Green) convened a committee (dubbed the Green istered to pregnant women in Europe for the control Committee) to determine the conditions under of nausea. Although the FDA withheld approval of which prisoners could serve as subjects in medical thalidomide for use in the United States, thousands experiments and whether a reduction in sentence of U.S. women, many of child-bearing age (includ- could be offered as a reward for such service.4 Pub- ing some who were pregnant), participated in studies lished, in part, in the Journal of the American Medical conducted by pharmaceutical companies on its safety Association in 19485 were three rules that the com- and efficacy. When the link between thalidomide mittee proposed for conducting medical research on and birth defects became clear, the FDA modified its prisoners: consent must be obtained in the absence of procedures for drug testing and ultimately required coercion and with knowledge of the potential risks; that such studies be conducted in three phases. Phase the studies must be based on the results of animal 1 studies required the use of healthy subjects to eval- experimentation and on knowledge of the natural uate the safety of experimental medication (Phases 2 history of the disease, and must be expected to yield and 3 were for establishing efficacy). Prisoners were results not obtainable in any other fashion; and the ideal subjects for Phase 1 studies. They were willing study must be conducted by scientifically qualified to accept more risk for less money and, according to personnel and avoid unnecessary physical or emo- the assistant medical director of one pharmaceutical tional suffering, and there should be no reason to company, were guaranteed to show up. Medical re- believe that death or a disabling injury would occur. search in prisons continued to grow. One particu- The ultimate conclusion of the committee was that larly egregious example was a study by Dr. Austin medical research could be conducted ethically with Stough, an Oklahoma physician who used prisoners prisoners if those criteria were met and, to avoid un- in Oklahoma, Arkansas, and Alabama for drug and due influence or coercion, the reduction of sentence blood plasma studies in mostly unregulated condi- was not excessive. tions. One report indicated that in a prison in Ala- With the endorsement of the American Medical bama, 28 percent of the prisoners participating in Association, medical research on prisoners flour- Dr. Stough’s studies developed viral hepatitis com- ished. In the last year of World War II, the National pared with 1 percent of the prisoners who did not Institutes of Health received approximately participate. One inmate was quoted as saying, $700,000 in federal funding; by 1955, this amount “They’re dropping like flies out here.”7 The studies increased to $36 million and reached $1.5 billion conducted in Holmesburg Prison in by 1970.6 Much of this funding was allocated to further illustrated the abuse of prisoners for medical medical research that was conducted on prisoners. research. Prisoners were routinely used for testing Noteworthy examples include Dr. Albert Sabin, who of cosmetics, details of which were documented by

Volume 41, Number 1, 2013 9 Coercion in Research

Allen Hornblum in his book, Acres of Skin: Human oping guidelines for the ethical conduct of research; Experiments at Holmesburg Prison.8 The title was de- the Belmont Report,10 named for the conference rived from a comment by Dr. , a center in which it was penned, was drafted in re- professor of dermatology at the University of Penn- sponse to this charge. The report described three ba- sylvania Medical School. When he first entered the sic principles of ethical research: respect for persons: prison, he reportedly recalled in a newspaper inter- persons should be treated as autonomous, and if au- view: “All I saw before me were acres of skin. It was tonomy is diminished, additional protections should like a farmer seeing a fertile field for the first time” be in place; beneficence: human subjects should not (Ref. 8, p 37). be harmed and the research should maximize bene- Although not conducted on prisoners, the Tus- fits and minimize risk; and justice: benefits and risks kegee syphilis study set in motion a chain of events of research should be distributed fairly. Many of the that effectively halted medical research in correc- recommendations of the Belmont Report were in- tional institutions.9 In the early 1900s, syphilis was corporated into the Department of Health and Hu- seen as a major health problem, with as many as man Services (DHHS) Title 45 Code of Federal Reg- 35 percent of people in their reproductive years ex- ulations (CFR) Part 46, known as the Common hibiting symptoms of the disease. The Tuskegee Rule.10 One major aspect of the Common Rule was study, initiated by the USPHS in 1932, was designed the requirement that research be reviewed by a com- to provide evidence of the significant effects of un- mittee on ethics: an institutional review board (IRB). treated syphilis, presumably in an effort to encourage These bodies were intended to provide oversight of the development of new treatments for the disease. research (researchers?) to ensure that the three basic The study involved 600 African-American share- principles of the Belmont Report are not violated. croppers from Alabama, 399 infected with syphilis In 1976, the Commission effectively banned med- before the study began and 201 noninfected. Partic- ical research in prisoners. Two concerns were ipants in the study were provided medical care, deemed most relevant for prisoners: whether prison- meals, and burial insurance; all participants were led ers bore the fair share of burdens and benefits (jus- to believe that they were receiving treatment; and tice) and whether they could truly give voluntary none provided informed consent. In 1947, penicillin consent (respect for persons). The Commission com- was established as an effective treatment for syphilis, mented: but rather than close the study and treat all the sub- . . . [T]he strong evidence of poor conditions generally pre- jects, investigators withheld penicillin from study vailing in prisons and the paucity of evidence of any neces- participants. The participants were never told that sity to conduct research in prisons have been significant they had syphilis, they were told that they had bad considerations of the Commission. An equally important blood, and the researchers prevented the men from consideration has been the closed nature of prisons, with the resultant potential for abuse of authority [Ref. 12, accessing the rapid-treatment centers established by p 12]. the USPHS for those infected with syphilis. The study ended in 1972 when it was made public in an They issued multiple recommendations, many of article appearing in . By the time which were adopted in the Common Rule. However, the study ended, 128 men had died, 40 wives had the Commission went beyond the recommendations been infected, and 19 children had been born with to propose a requisite standard of living for inmates congenital syphilis. A class-action lawsuit resulted in in prisons conducting research, making research out a $10 million settlement by the U.S. Government of the question for most, if not all, prisons. Such and sweeping changes in the conduct of research. On standards included limiting prison populations, pro- May 16, 1997, President Bill Clinton issued a public viding single-occupancy cells, segregation of offend- apology to all participants in the Tuskegee study as ers by age and offense as well as potential for violence, well as to their wives and children. and “effective procedures assuring that parole boards As a result of the Tuskegee study, in 1974 Con- cannot take into account prisoners’ participation in gress passed the National Research Act, which cre- research and that prisoners are clearly informed that ated the National Commission for the Protection there is absolutely no relationship between research of Human Subjects of Biomedical and Behavioral participation and determinations by their parole Research. The Commission was charged with devel- boards” (Ref. 12, p 19).

10 The Journal of the American Academy of Psychiatry and the Law McDermott

Modern-Day Research on Prisoners only when the benefits outweigh the risks; require The Common Rule identifies certain “vulnerable” that all research be conducted with collaboration populations, including pregnant women, human fe- from key stakeholders (prisoners and prison staff) tuses and neonates, prisoners, and children. The def- and enhance oversight of research involving prisoners. inition of prisoner includes not just individuals in- Because these recommendations were so contro- versial, with some factions expressing concern that carcerated in prisons, but any individual committed the IOM was imposing even further restrictions on to a facility in lieu of incarceration as well as indi- research15 and others believing that the revisions viduals detained in jails awaiting arraignment, trial, would open the door to further unethical re- or sentencing. Only minimal-risk research is allowed; search,16,17 the recommendations have not yet been for prisoners, minimal risk is defined as “the proba- adopted and incorporated into the Common Rule. bility and magnitude of physical or psychological harm that is normally encountered in the daily Has the Pendulum Swung Too Far? lives, or in the routine medical, dental, or psycholog- The Question of Voluntary Consent ical examination of healthy persons.” Further, only four types of research are permissible: studies of the As the arguably unethical research of the past has causes, effects, and processes of incarceration or exemplified, the Green Committee’s expectations criminal behavior; studies of prisons as institutions or that researchers police themselves regarding ob- prisoners as incarcerated persons; research on condi- taining voluntary, informed consent fell short. The tions particularly affecting prisoners as a class; and Tuskegee Study was a particularly egregious exam- research on practices that have the intent and prob- ple. Participants not only did not provide consent, they were unaware that they were being enrolled in a ability of improving the health or well-being of the research study. The biomedical and psychological subject. Both Types 3 and 4 require approval by studies conducted on prisoners, too numerous to the Secretary of Health and Human Services, a task document in this article, made it clear that voluntary so onerous that most federally approved research is consent was particularly problematic in this popula- limited to Types 1 and 2. However, only federally tion. The primary impediment to voluntary consent funded research and certain other federal agencies are in prisoners is coercion: the forces that exist inside required to comply with these regulations; biomedi- prison walls that may subtly or not so subtly encour- cal research can continue in prisons if it is not feder- age participation in research, so that participation is ally funded (e.g. funded by pharmaceutical compa- 13 not technically voluntary. In 1948 the Green Com- nies), and state laws allow such conduct. mittee proposed that granting a pardon to an inmate Because the Common Rule imposes severe re- serving a life sentence in exchange for research par- strictions on federally funded research but allows ticipation was excessive and clearly coercive. In the other biomedical research to continue unregulated case of the Stateville Penitentiary, where correctional in prisons, the DHHS requested that the Institute officers were ever-present and could be watching at of Medicine (IOM) form a committee to re-evaluate any time, the exercising of authority and potential the guidelines contained in Subpart C of the encouragement for research participation was more Common Rule, which were codified in 1981 and subtle. The IOM has suggested that protections af- revised in 1991, to determine if revisions were nec- forded prisoners should extend to parolees and essary. The committee, in a 265-page document probationers. These individuals are under the juris- 14 published in 2007, presented their recommenda- diction of the criminal justice system, with clear con- tions, which in essence consisted of five changes: sequences (including incarceration) for failure to co- expand the definition of prisoner to include individ- operate with the terms and conditions of probation uals whose liberty is restricted by the criminal justice or parole. However, they are not residing in a con- system (e.g., individuals on probation or in diver- trolled environment, locked behind closed doors and sion programs); widen the scope of the regulations to high walls, which is in part what led to the abuses in include all research conducted on prisoners regard- the past. Are probationers and parolees less likely to less of the funding source; eliminate the category- be able to provide voluntary, informed consent be- based approach to allowable research and shift to a cause of their involvement in the criminal justice risk-benefit approach, allowing research on prisoners system? Are they as susceptible to coercion?

Volume 41, Number 1, 2013 11 Coercion in Research

A recent study highlights the complicated ques- Conclusions tions regarding voluntary consent and vulnerability 18 There are no simple solutions to these questions. to coercion. A sample of IRB members was re- Clearly, research conducted on prisoners in the past cruited and provided written vignettes documenting involved coercion, both from the authority imposed hypothetical research studies. Participants in the hy- by prison officials and the fact that release decisions pothetical studies were varied along illness type indeed were based on participation. The IOM at- (psychiatric versus medical) and illness severity tempted to resolve many of the problems associated (low versus high). IRB members were asked to make with the conduct of research with prisoners in its judgments of decisional capacity, coercion, and risk. recent review of Subpart C of the Common Rule. Their judgments of vulnerability to coercion varied, The committee met with stakeholders, including depending on illness severity and type. For low- both prisoners and researchers, for a thorough review severity illnesses, vulnerability to coercion was higher of the scope of the problem and to determine if revi- in psychiatric patients; for high-severity illnesses, sions to the current standards were necessary. Prison- there were no differences in IRB members’ judgment ers are not the only research participants vulnerable of vulnerability to coercion: psychiatric patients to coercion, as the IOM acknowledged by including and medically ill participants were viewed as equally nonincarcerated individuals under control of the vulnerable. These results suggest that vulnerability to criminal justice system in its definition of prisoners. coercion is not simply based on one’s living environ- Yet, the recommendations stirred such controversy ment and may be based, at least in part, on how that five years later, none of them has been adopted. high-stakes the research is. The study by Luebbert and colleagues18 makes it Along these lines, how many research participants clear that IRBs believe prisoners are not the only actually provide informed, voluntary consent? Con- research participants vulnerable to coercion. When sent forms for clinical trials often exceed 15 to 20 life is at stake, whether within the confines of prison pages, as the researcher dutifully outlines all the pro- walls or not, it appears that many research partici- cedures and the risks and benefits of participation, pants are willing to take excessive risks. Can a termi- according to all established regulations. What level nally ill individual intelligently weigh the risks asso- of understanding do these participants truly have? ciated with dangerous research when facing death? Is Many of them have life-threatening illnesses and may the answer to create more and more restrictions to be desperate for a treatment that could effect a cure. protect classes of participants? Perhaps not. Perhaps One particularly poignant example was publicized in the answer is to explore more systematically the ques- a local newspaper. Two surgeons were performing an tions associated with consenting to participate in experimental treatment on patients with a particu- research, both in prisoners and other vulnerable larly aggressive type of cancer. The prognosis for in- populations. While capacity to consent has been re- dividuals diagnosed with this disorder is poor; it is searched extensively, especially in individuals for almost always fatal, with death occurring within whom capacity is suspected to be impaired, coercion months of diagnosis. The procedure was experimen- in research has been substantially less extensively re- tal and according to news reports (which included searched. Coercion was clearly a force in prisoners confidential letters from the university) had not been when freedom was granted in exchange for research fully tested according to FDA guidelines. Three pa- participation. Under what specific circumstances tients consented to the procedure and, according to and with what specific participants is coercion also a public documents, two of the three actually re- problem? Instead of expanding and changing defini- quested the procedure. Given that these individuals tions, it seems most appropriate to explore in a more had been in essence given a death sentence and were systematic way the role that coercion plays in bio- eager to undergo any type of experimental treatment medical research in all participants, not simply those to extend their lives, could any type of informed con- who reside in (presumably) coercive environments. sent be truly voluntary? The results of the study of the IRB members suggested that such individuals References (i.e., those with extremely high-severity illnesses) are 1. Office of NIH History: Dr. Joseph Goldberger and the War on clearly at risk of coercion and may be less able to Pellagra. Available at http://history.nih.gov/exhibits/goldberger/ provide voluntary, informed consent. index.html. Accessed December 27, 2012

12 The Journal of the American Academy of Psychiatry and the Law McDermott

2. Comfort N: The prisoner as a model organism: malaria research at www.hhs.gov/ohrp/policy/belmont.html. Accessed January 23, at Stateville Penitentiary. Stud Hist Philos Biol Biomed Sci 40: 2013 190–203, 2009 11. 45 C.F.R. § 46.101 (1991) 3. Leopold N: Life Plus 99 Years. Garden City, NY: Doubleday, 12. The National Commission for the Protection of Human Subjects 1958 of Biomedical and Behavioral Research: Report and Recom- 4. Harkness JM: Nuremberg and the issue of wartime experiments mendations: Research Involving Prisoners. DHEW Publication on US prisoners: The Green Committee. JAMA 276:1672–5, No. 76-131, 1976 1996 13. Thomas DL: Prisoner research: looking back or looking forward? 5. Green DH: Ethics governing the service of prisoners as subjects in Bioethics 24:23–6, 2010 medical experiments. JAMA 136:457–8, 1948 14. Gostin LO, Vanchieri C, Pope A (editors): Ethical Considera- tions for Research Involving Prisoners. Washington, DC: Na- 6. Rothman DJ: Strangers at the Bedside: A History of How Bio- tional Academies Press, 2007 ethics Transformed Medical Decision Making. New York: Basic 15. Test drugs on inmates? Yes. Editorial. USA Today, August 21, Books, 1991 2006 7. Starr DP: Blood: An Epic History of Medicine and Commerce. 16. Hornblum AM: When prisoners are victims. Chronicles of New York: Alfred A. Knopf, 1998 Higher Education, October 6, 2006 8. Hornblum AM: Acres of Skin: Human Experiments at Holmes- 17. Obasogie O, Reiter KA: Human subjects research with prison- burg Prison. New York: Routledge, 1998 ers: putting the ethical question in context. Bioethics 25:55–6, 9. Jones JH: Bad Blood: The Tuskegee Syphilis Experiment. 2011 New York: Free Press, 1993 18. Luebbert R, Tait RC, Chibnall JT, et al: IRB members judg- 10. Belmont Report: The Belmont Report: ethical principles and ments of decisional capacity, coercion, and risk in medical and guidelines for the protection of human subjects research. Available psychiatric studies. J Empir Res Hum Res Ethics 3:15–24, 2008

Volume 41, Number 1, 2013 13