As the Nazi doctors knew well, there’s a seductive inverse relationship in human experiments: the less you protect the particular humans you study, the more you can learn about humans in general. Systematic near-freezing of Jews in concentration camps taught the German Luftwaffe a lot about staving off hypothermia, which was crucial for their pilots when they had to ditch in the North Sea. But the Nazis weren’t the only ones to understand this perverse dichotomy. Fifty years ago the U.S. government allowed black men with syphilis in Alabama to go untreated as part of the notorious Tuskegee experiments.
Even today, there are troublesome cases that keep regulators, doctors and ethicists up all night. A two-year old girl died recently when doctors forwent a known surgical cure for her bladder cancer and gave her an experimental drug instead. We still have lapses. And we still have the pull of experiment’s inverse relationship. In all but the rarest of cases, we don’t have mad scientists bent on discovery at the expense of human welfare anymore. So when little girls die in research, or when mentally ill patients are subjected to experimental harm, or when pregnant women don’t know they are in studies, it is a systemic problem: one that points out inherent flaws both in the way the government is organized to protect citizens, and in the very nature of having some human beings poke, prod, cut and drug the rest of us.
In 1995, President Clinton wrote an executive order creating a National Bioethics Advisory Commission (NBAC), composed of doctors, lawyers, and ethicists from around the country. In renewing its mandate in 1997, he said that, “Science must respect the dignity of every American. We must never allow our citizens to be unwitting guinea pigs in scientific experiments.” NBAC resolved that the president’s goal would be met if every subject in a medical trial enjoyed two protections: informed consent and independent review of research. Informed consent means that patients know everything they need to know about what is going to happen to them, and that they willingly agree to submit to the research. Independent review means that researchers open their studies to outside observers (usually a committee of doctors and community members) so that a disinterested party can assess their ethics.
But commission members say the president’s goal has not been met; Americans are not as safe as they could be. Federal protections do not cover all research. And the research the government does oversee is on shaky ground. Informing patients to get consent is difficult because average patients and average doctors speak very different languages. And independent review is compromised by mounting pressures for institutions to make money through medical discovery: A culture of “approve it now, ask questions later” pervades the research community. It may not be long before Americans lack up-front guarantees that what doctors do to them is needed, ethical, and benign. Earlier this year, the commission put it this way in a memo to the president, “The U.S. Government cannot know how many Americans currently are subjects in experiments, cannot influence how they have been recruited, cannot ensure that research subjects know and understand the risks they are undertaking, and cannot ascertain whether they have been harmed.”
Why? First there is the problem of jurisdiction. As the law stands, the federal government oversees research it funds (the lion’s share is an annual $12 billion in grants from the National Institutes of Health), and regulates research that will lead to products sold in interstate commerce (FDA trials). Some states, such as California, have laws about the ethical treatment of subjects; many do not. But that leaves a lot of research out of government’s reach. Small institutions, such as fertility clinics, privately funded laboratories, small universities, and weight centers are essentially free to do whatever they want to patients who walk in and sign forms. Despite President Clinton’s wish that no American be a guinea pig, those who want to lose weight, who need help conceiving, or who are simply pulled in by offers of money to join small studies are unprotected—and unprotectable—at least until the next wave of Federalism sweeps the country. Though subjects in these studies theoretically have redress through state tort laws, proving medical abuse is difficult. What seems to frustrate NBAC members most on this count is that they cannot discover how many Americans are subjects, abused or not, in such studies. Says one Bioethics Commissioner, Alta Charo, a lawyer at University of Wisconsin, Madison, “We are advocating for a solution in terms of a problem, but we don’t have any idea whether the problem is big or small.”
NBAC is not alone in this concern. Last year, the Inspector General of the Department of Health and Human Services (which includes NIH) issued a report saying that “[t]he long-established system for protecting human research subjects has vulnerabilities that threaten its effectiveness,” and that given the new realities of the medical marketplace—genomics, stem cells, cloning, for-profit medicine, and the ever increasing drive to publish results—these vulnerabilities are a “serious national issue.”
Both Duke University and a Veterans Affairs facility in Los Angeles had their research shut down this spring by NIH’s oversight arm, the Office for Protection from Research Risks (OPRR). Neither institution was accused of actually harming patients. OPRR took the drastic measure of halting all research because each had violated a technical regulation for independent review (Duke didn’t remove review board members that OPRR wanted removed; the V.A. approved studies without a review board quorum present.) In other words, the vulnerabilities of the system are apparent enough to the regulators that small slips require serious response. It is possible that the line between healthy research and freezing humans on blocks of ice is just that thin.
Here’s how the current system works: the government, through OPRR, gets assurances from institutions that they will oversee the ethical treatment of subjects in the government’s stead. The institution fulfills that assurance to OPRR by setting up institutional review boards (IRBs) to review research proposals before they begin. IRBs are composed mostly of doctors from the institution, but each must also have a non-affiliated community member. These committees examine the research proposals to make sure they are ethically designed. Once an IRB approves a study, the researcher is on his own to sign up subjects, pending informed consent, and conduct the study for up to a year, when the research must be re-approved by the IRB. During that year, the doctor is on his own. The government leaves it to the institution, the institution leaves it to the IRB, the IRB adjusts what it wants and then leaves it to the investigator. And the investigator leaves it to the subject, in so far as he can experiment only on willing and informed volunteers.
The vulnerabilities in our system of protecting subjects starts with the difficulty of ensuring that volunteers are willing and informed. How does an investigator tell an average person what her study entails? How do you sign up people for a gene therapy trial when most people don’t know what genes are? Investigators try to pitch information on a sixth to eighth grade level. But, says Alta Charo of NBAC, “Americans are very poor at interpreting statistical information. They have binary attitudes. They have trouble with [assessing a] 3/10ths chance.” According to the Inspector General’s report, “Many individuals have signed up to participate in research trials with little understanding of the distinction between research and standard therapy.” This is called the therapeutic misconception—my doctor is asking me to do something; it must be in order to treat me. The case of the man who was sure he was in therapy for his heart condition but was actually in an experiment—and died not knowing better—is a classic example.
Regulators fear this is a common problem because investigators often use their own patients in studies. And the confusion of roles may not be just the patient’s. An investigator’s focus in a trial is science, which requires filling the research protocol set ahead of time and not making individual adjustments for any one subject. If the investigator is also a subject’s clinician, then, in a sense, that patient no longer has a doctor—someone who is looking out expressly for his interests and modulating care to suit his needs. The priorities are different enough, says James Childress, another NBAC commissioner, that the Hippocratic oath itself is in conflict for the clinician-investigator. Guiding the science that will help future patients and attending the needs of current subject-patients are different goals. “Those goals can come into conflict at times,” says Childress. The inverse relationship between knowledge and care rears its head.
However, the number of actual abuses is not as high as it could be, given the power doctors have over generally willing subjects. NBAC commissioners, the Inspector General’s report, and OPRR officials—despite their worries about systemic weaknesses—all credit the culture of the American medical profession with an outstanding sense of ethics. This century, we have added 26 years to our average life expectancy in this country—thanks in large part to medical interventions tested on human subjects and nutritional knowledge assayed in clinical trials. And the number of abuses is “vanishingly small” in the words of one regulator. The egalitarian nature of science assists in this good record. There are no lone wolves in medicine (weight-loss clinics aside). Just as science dictates that all findings be reviewed by the doubting eyes of peers, fifty years of medical tradition dictate that all proposed research undergo ethical screening by an institutional review board.
These review boards, made up mostly of doctors, do three things to protect human subjects. They assure that the proposed research is going to lead to worthwhile results—science must learn something by doing these things to humans. They shave risk off the research protocol so that human subjects are as unexposed to risk as possible, given what the investigator is trying to find out (heart surgery trials usually involve high risk; smoking cessation trials should not). And IRBs worry about informed consent. Review boards must have a non-doctor community representative on them. Their primary purpose is to help word the consent information to prevent therapeutic misconception and to ensure that the information includes all a subject needs to know to make an informed decision. As Alta Charo put it, transient diarrhea in a drug trial may not be medically significant to an investigator. But it is highly significant to potential subjects deciding whether to sign up.
A more serious example of the need for IRB review comes from Jonathan Moreno, a professor of ethics at the University of Virginia, who has just written a history of human subjects research called Undue Risk. A certain set of children have a condition that stunts their growth and impairs their cognitive ability. A respectable investigator submitted to his IRB a protocol to test whether a certain treatment would help these children grow. He was sure the regime would do nothing for their IQs. There was also a minimal cancer risk associated with the drugs. The researcher’s informed consent packet left out the parts about cancer and not helping the IQ. He was not being devious, says Moreno; he was just so focused on the science of the study—would the treatment help growth?—that he neglected information crucial to the parents who might allow their children to participate. The study was not approved by the IRB until he reworked the information packet.
Moreno and other ethicists, as well as NBAC, and the Inspector General worry about what would happen if IRBs stopped being thoughtful. What if they became rubber-stamp approval boards, rather than deliberative bodies with the power to alter protocols? It is a change that regulators do not want to see. However well-meaning individual investigators are, the ethical sense of the medical community does not reside in each doctor individually any more than does scientific truth. One investigator’s results are not scientific truth until others assess them and agree. The same is true for ethical treatment of human subjects. An investigator’s judgment of proper treatment is always jeopardized by the inverse relationship rule. The temptation, however unconscious, to increase risk to gain return is ever-present. That’s why outside eyes and independent adjustment are needed.
Nevertheless, IRBs are being squeezed for time and resources. The old, hallowed halls of academic research are rapidly changing as managed care and pharmaceutical companies bring marketplace pressures to bear on institutions. Administrators want more research done faster, for the prestige—and bottom line—of the institution.
A public affairs official for the V.A. contends that institutions all over the country cut corners. “Many of us feel under-appreciated,” says Dr. David Gaba, the chair of one of Stanford University’s two medical IRBs. “The work is unrelenting. It doesn’t wax and wane.” In fact, it just waxes. His IRB oversaw 1,350 protocols last year, meeting once a month for four hours. Members do spend time at home reading and reviewing, but the statutory work of meeting to amend and vote on research proposals is limited to just over two minutes per protocol on average. And the number of protocols increases 7 to 10 percent annually. Stanford—an institution that has not yet run into trouble with OPRR—is adding a third medical IRB to ease the burden on members, a move the Inspector General of Health and Human Services applauds. But many institutions are not adding IRBs, despite the mounting workload and pressures to approve protocols.
In addition to being “overworked and overwhelmed” as one ethicist put it, IRB members must endure what amounts to begrudging tolerance among researchers. “I’ve been doing research with human subjects for many years,” says Dr. Marguerite Hays, a V.A. administrator and researcher. “I find it very frustrating to have to wait for a committee to tell me I can make a small change in protocol.” That wait averages 37 days in academic settings—too long for the new market pace. Most doctors consider themselves ethically equal to IRB members and scientifically superior to them in their own sphere of study. For the members of the review boards, it is hard to ignore that attitude. “Every time you find something wrong, you are risking the irritation of your own colleagues at your institution—and you are giving yourself more work,” says Alta Charo of NBAC.
Almost no one in the medical community argues that review boards are unnecessary. (For one thing, IRBs cover doctors’ backs. Investigators who stick to IRB-approved protocols are exculpable, should ethical questions arise.) But no one is moving to make sitting on review boards any easier. In the words of one reviewer, “The need to cost out all your services is becoming more critical to the institution. Service on an IRB is not billable time.” That is true for the institution and it is true for the IRB members, who are not generally paid for IRB work. David Gaba thinks IRB service “should be viewed as a positive thing in your [academic] promotion. But getting a couple of studies published is viewed as 100 times more important to your promotion.”
The situation is only going to worsen as the pace of research increases, as the need for money and fame that come from research grows, and as the content of research becomes more ethically challenging. Genetics and stem cell research and privacy issues in studies such as AIDS trials are giving IRBs complex issues that need more time, not less.
So far, OPRR has been firm about shutting down research institutions that cut back on their IRBs. (OPRR’s method is money—it holds the purse strings from the government and can freeze NIH-funded research immediately). But OPRR officer Tom Puglisi claims “OPRR is small for the task it has to do. We’re overseeing 5,000 institutions. My staff consists of 15 people. It’s [a system] of monitored self-regulation.” OPRR makes very few site visits. Its ability to ensure the protection of human subjects in research relies on assurances from institutions that they will do the right thing. And the IRB is the crux of that assurance.
Because NIH is in the business of giving out money for clinical trials, OPRR is naturally reluctant to shut down research. The government wants doctors everywhere to experiment on human beings. That’s how medicine is advanced, how discoveries are made, how you and I live longer. But against a backdrop of Nazi experiments and the Tuskegee syphilis trials, medicine must be careful. Despite her frustration with IRBs, Dr. Hays cautions doctors away from ethical hubris. She notes that certain EU countries have banned diagnostic tests involving radioactive tracers (common here, and harmless). “I think that this is a reaction that legislators have to what they see as abuses. We have to be very careful not only not to allow abuses, but not to allow the appearance of abuses. You may find that you are not allowed to do the research at all.”
Research on fellow humans is a privilege, not a right. Both the flow of government cash and the willing trust of potential subjects could dry up if medical research institutions aren’t careful. Without the safeguards of informed consent and robust independent review, research trips itself, says NBAC’s James Childress. Abuse invites reactionary legislation—and, with it, the loss of the research privilege and its contribution to the greater good. “It is protection that provides the basis of trust that makes it possible for subjects to participate,” he says. “What is right is what also turns out to be in the long run most effective.”