HAROLD EDGAR
Twenty years ago, when the Washington Star told the public that the United States Public Health Service had, since 1932, maintained a study of untreated syphilis in the Negro male that was still going on, my reaction was, How could people have done this? I later worked on the participants’ lawsuit, and I learned of the study’s many complexities. In the end, though, the best explanation of “how” it could have happened is the obvious one: the researchers did not see the participants as part of “their” community or, indeed, as people whose lives could or would be much affected by what the researchers did.
Looking back on those events after two decades, there are a number of observations I’d like to make.
Harold Edgar is professor of law and director of the Julius Silver Program in Law, Science and Technology at the Columbia University School of Law.
Originally published in the Hastings Center Report 22 (1992): 32–35. Reprinted by permission of the Hastings Center and Harold Edgar.
First, I should like to describe some aspects of the legal landscape of the Tuskegee litigation. The injustices done to the participants did not fit easily into the framework of an adjudicable lawsuit. The issues that faced Fred Gray, an Alabama civil rights lawyer, and me could have filled a book. Among them: the United States cannot be sued in tort without its consent. However, in 1946 by act of Congress the government became responsible in damages, like any employer, if an employee exercising a nondiscretionary authority causes an injury cognizable under state law. Did the act permit recovery for harm done between 1932 to 1945 to people (many of whom were dead by 1946), on the theory that there had been an ongoing project whose success depended upon keeping the facts secret from the participants? Second, could unwitting (and unwilling) service in an experiment give rise to a contract claim against the government, or even a claim that government had taken property and owed compensation? Third, what relevancy, if any, did Nuremberg have? There, the United States had conducted in its own name criminal proceedings against German doctors on the theory that their criminal homicides and assaults could not be justified by their desire to gain useful information by experiments. Fundamental principles of medical ethics recognized everywhere, the prosecutors claimed, permitted experiments only with informed consent. Had our government’s conduct, therefore, violated the fundamental principles of liberty and justice embodied in the American conception of due process? If so, can the United States be held responsible for its employees’ “constitutional torts” under the 1946 act on a theory that “state” law includes federal law?
How could we prove damages, particularly when people had died from causes not obviously related to syphilis? Could we use in evidence against the government the chief finding of the study—that syphilis seems to cause pervasive medical problems previously not thought to have been caused by syphilis? If so, could we shift the burden of proof to the government and require it to disprove the relationship between syphilis and any individual’s medical condition? If not, could we use aggregate data about the men as a group (so much extra heart disease, for example), and then value the excess and award it proportionally to individuals?
The complexity of sustaining such a suit through a verdict and appeals was matched by its futility, if the ambition was to put money into the hands of those who had personally suffered injustice. Even if ultimately successful, litigation would take years, and fewer people would actually have the advantage of monetary payment. And yet was it fair to settle the action on a class basis without regard to individualized damages? We believed there were major barriers to any recovery, and it was futile to attempt to ascertain the medical damages each person sustained. The victims’ class representatives themselves made the decision to settle, and I had no doubt then nor do I now that this was a wise thing to do. As further proof that law marches to the slowest of beats, it is just this past year that the last payments to a few persons pursuant to the settlement agreements have been made.
As a result of the settlement, however, the Tuskegee litigation made no new law. The extremely interesting issue of how government misbehavior in conducting medical experiments should be analyzed for purposes of constitutional law went unanswered and remains unanswered to this day. The problem has been raised in litigations involving CIA experiments, and military experiments exposing servicemen to radiation, among others, but the Supreme Court passed over its chance to comment on the problem. In United States v. Stanley, the Supreme Court held that the courts were closed to a serviceman injured by surreptitious administration of LSD.
The Tuskegee study was, however, surely a major force in the development of American bioethics. Among other things, it was a direct cause of the National Research Act in 1974, which required the establishment of institutional review boards at institutions receiving federal grants. These IRBS have been transformed from peer group meetings in which controversial research might be discussed into legally mandated, professionally diverse, and bureaucratically independent centers that have the power to shape and regulate research within an institution. Increasingly, conventional product regulation, such as the FDA’s regulation for commencing testing of insignificant risk devices, builds on the presence and protective capacities of IRBS. And yet who gets on IRBS, the extent to which they are subject to institutional controls, how successfully they carry out their mission, and at what costs, remains, I submit, largely a mystery. Nonetheless, the core IRB concept—an institutionally based, multidisciplinary ethics committee—has been adopted as the paradigmatic American response to bioethics dilemmas.
Apart from its violation of human rights, the Tuskegee study was a serious incident of what is now called “science misconduct.” The story is told brilliantly by Jim Jones in his book Bad Blood. Jones found and organized the physicians’ and the study’s early papers in the National Archives. These materials showed not only that the study was ethically wrong measured by the ethical standards of the 1970s, but that it had been, from the start, a program built upon deception. In 1932 the researchers told participants they were getting free medical treatment. Without that promise, the researchers could not get participants’ cooperation in giving blood and undergoing spinal taps. The subsequent program built on that lie, although it was a complicated lie.
The central fact about the forty-year study was that its scientific rationale made no sense. The fact that the government was running a long-term study when no person in charge had any idea of what the original protocols were, or where they were, is itself shocking. On reading the protocols, moreover, it is apparent that the forty-year study was an afterthought.
The initial plan was simply to go to Alabama for a one-shot visit to take a snapshot of secondary syphilis in a population. The snapshot would show, for example, whether people who had had syphilis for ten years were much worse off than those who had had syphilis for five years. (Some readers may not realize that the government did not infect the participants; they had the disease already. The men ranged in age from their twenties to their sixties when the study started.) The initial plan rested on a highly dubious premise, namely, that the participants were capable of giving adequate information about how long they had been infected. When they got the data, however, the physicians could not read their snapshot. Although they had generated a lot of data, no one knew whether the X-ray pictures that seemed to reveal massive amounts of heart damage were either accurate or, if they were, indicated syphilis rather than other causes. Only after doubts came up about the interpretability of data was a decision made to go back to Alabama to create a control group of age-matched nonsyphilitics, and “bring the material to autopsy.” The goal was to learn by autopsy whether the heart problems were indeed caused by syphilis.
The forty-year study then proceeded upon a false claim that the men had not been treated. They were treated in 1932, although completely inadequately, by then-current standards. They were treated with as much mercury and arsenicals as the government doctors doing the first work in 1932 could get—the only limit on treatment was the availability of drugs. The participants’ cooperation was obtained by providing such treatments. The researchers then administered a spinal tap with no anesthetic, called it treatment, and left town. How did it happen that the long-term study explained itself as a study of “untreated” syphilis?
Despite their having administered drugs, and thus destroyed the sample, the researchers pressed on. Moreover, they did what people still do today: they assumed their error had no biasing impact. Interesting also, for it too has its modern analogy, was their reasoning, which was shaped by their capacity to measure. Since modest amounts of arsenicals and mercury did not cause seroconversion of the blood (the Wassermann tests still were positive), they assumed that treatment had no possible effect upon syphilis. Whether it did or not is impossible to tell, because among other things, in later years, they had no way to control for who was getting what antibiotics in private medical visits. The aggregate data suggest, however, that the younger men did not do as badly when measured against the control group as did the older ones. Sampling biases, however, could readily account for the difference, inasmuch as one would predict that sicker people, among those who were in their forties and fifties in 1932, would find more attractive the promise of free treatment and thus show up in greater numbers.
This was bad science. It was also unethical medicine. It was during the first fifteen years of follow-up that the most outrageous incidents occurred, including switching newly infected control subjects into the participant group while making no effort to protect them or their families, and keeping participants from becoming aware of their disease or of treatment available for it when new public health efforts were launched again after the Depression.
Yet, one of the worst faults of the project in later years had nothing to do with whether administering penicillin was a good idea: that point was complicated initially by the question of whether the massive doses required were safe, and by the belief that the disease was probably benign in people who had had it so long. The harm came from the failure aggressively to follow up on diagnosed medical problems having nothing to do with syphilis, while simultaneously handing out placebos that many men understood to be medical care. By the end of the study, a number of participants had private doctors, but many others did not, and the failure to follow up adversely affected controls as well as syphilitics.
For reasons I do not understand, front page news coverage results today when science fraud occurs. Why there should be surprise that misdeeds, let alone dumb mistakes, happen in a billion-dollar business I do not know. And yet if one looks to standards in the recent past, one realizes that contemporary scientists are vastly more alert to these problems than were people who were highly regarded in their own day. It is, therefore, not the case that things are always getting worse.
One irony about Tuskegee is that the study was racist to the core, in that no such program could possibly have continued so long but for the central fact that participants were African Americans. (Indeed, I believe race injected itself in the decision whether to shut the program down in the 1960s. Sadly—for the government should have legal authority to keep its promises—there was no easy way to secure payments of death benefits to the participants unless there was an ongoing study. The men rightly counted on them as burial insurance, and if they stopped, the story might become public. So I believe the research kept going in part to keep the story private.) Despite the centrality of race, however, the researchers were not personally vicious racists. They sought and secured research help in the African-American community, and they were personally committed to vastly expanded public health services for African Americans.
I disagree with one of Jones’s conclusions, however. He suggests the physicians believed that blacks and whites were different in terms of how syphilis affected them, and that the study was intended to shed light on this difference. The former proposition is true, but my recollection is that the latter had nothing to do with the enterprise. What drove the physicians was an attempt to undermine the conclusion of the best prior study of untreated syphilis, a study done in Oslo, Norway. It may well be that, as Jones recounts, the Oslo data suggested neurosyphilitic complications were less common among whites than supposed. That was not, however, the conclusion that attracted American attention. The Oslo study suggested that if the individual infected does nothing about syphilis, in two out of three cases nothing adverse happens to him. The Oslo data were subject to all sorts of sampling biases, and the American doctors didn’t believe them. If true, however, the proposition had implications for how one spent the limited funds available for public health during a depression.
On the issue of racial impact of the disease, the physicians involved unquestionably believed both before the study and as a result of the study’s data that syphilis does affect different groups differently. They believed neurosyphilis is a less common outcome among blacks than whites, and that crippling heart disease is more common among blacks than whites. Indeed (and I thought it was cause for grievance), they used that belief to justify continuing the study even though they could not monitor for neurosyphilis because the participants would never again accept a spinal tap. I do not know whether their beliefs about the relative impact of the two complications among blacks and whites are true. The physicians did not think that their belief in this regard was particularly interesting or surprising, however. Nor was this specific belief a prop for a broader, more systemic set of beliefs about blacks.
There has been a sea-change in attitude about such matters today. At present, any assertion of group difference is always a potentially explosive issue. The horrors of the Holocaust and the pervasive racism that still afflicts American life make it impossible for many people even to contemplate that different groups may be different in lots of ways. Acknowledging that fact may permit others to urge social policies that are frightening, and so the intellectual leadership suppresses the observations in order to keep control of the inappropriate inferences that others might draw. Yet what the Tuskegee doctors could contemplate—group differences—will be forced upon the medical world in the effort to understand problems ranging from why some people get lung cancer while others do not, to how come the French eat so much fat and do not get our rates of heart disease. Indeed, the fact that groups may be different (why does a vaccine work in Finland and not in Alaska?) may have major implications for how we appraise “research” as opposed to “experimental therapies,” for we draw lines between these categories more sharply than the reality makes desirable.
Finally, the Human Genome Project will bring to the forefront of human consciousness awareness of the range of variability not only among individuals, but among groups. To acknowledge those differences, while insisting on their irrelevance to respect for individual dignity and equality of right, is a challenge we shall have to face.