Commentary on Grady et al.: Using poor,uninsured minorities to test the safety of experimental drugs

In 2004, a pharmaceutical researcher in Alabama offered a paycheck to 21 clients of a Mobile homeless center in exchange for testing an experimental smallpox vaccine. ¹ These people could expect little medical benefit from the study, of course. Smallpox was eradicated in 1980, and in the absence of a bioterrorist attack on Mobile, it was unlikely that clients of a homeless center there would need protection from it. But medical benefit was not really the point. The main purpose of these studies was to see if the experimental vaccine carried unpleasant or dangerous side effects.

In some ways, studies like this are not unusual. Pharmaceutical companies routinely pay desperately poor subjects to test the safety of experimental products. In this case, however, 2 of the 21 subjects had to be hospitalized, one of them for acute myocarditis—an inflammation of the heart muscle that can cause sudden death. A third subject contracted pericarditis. It did not help that the research team lost this subject’s medical files for two years, one of many blunders that attracted federal scrutiny. When the Food and Drug Administration (FDA) investigated in 2007, it found a host of risky research practices ranging from careless record-keeping to dubious recruiting procedures, as well as little meaningful oversight by the Institutional Review Board. ² The FDA eventually disqualified the researcher in 2008, effectively barring him from doing any more studies. But he continues to practice medicine in Alabama, marketing sexual enhancement procedures as “Dr. Orgasm.” ³

By this point, nobody should really be shocked by episodes like this. It has been over 20 years since Bob Helms began publishing Guinea Pig Zero, the dark, caustic job-zine that chronicled the world of poor people who test the safety of experimental drugs for money. ⁴ In 1996, The Wall Street Journal reported that Eli Lilly was recruiting homeless alcoholics for safety studies at its Indianapolis trial site. ⁵ In 2005, Bloomberg Markets found that a contract research organization called SFBC International was paying undocumented immigrants to test new drugs in a converted motel that was later demolished for fire and safety violations. Prior to demolition, it was the largest trial site in North America. ⁶ In 2010, Roberto Abadie published The Professional Guinea Pig, a deeply researched ethnography of semi-professional research subjects living on the margins in Philadelphia. ⁷ In 2014, I wrote about the recruitment of homeless people with schizophrenia to test the safety of experimental antipsychotic drugs, including a mentally ill veteran named Walter Jorden who died in a Phase I antipsychotic trial in New Jersey. ⁸ After decades of medical experimentation on the poor, the relevant question under debate is not “Does it happen?” but rather “Are poor subjects being exploited?”

If the companion studies of Phase I trial subjects in this issue of Clinical Trials do not directly address this question, neither do they provide much reassurance.^9,10 The subjects in these studies are largely poor, uninsured and marginalized. Only 12.5% of the subjects surveyed at Pfizer’s trial site in New Haven had full-time employment. Although it is unclear just how many subjects had incomes lower than the federal poverty level (currently set at US$12,060 for individuals), over 47% of subjects had household incomes below US$25,000 a year. In one of the studies, over 38% of the subjects had no health insurance, not even Medicaid. Equally disturbing is the fact that over half of the subjects testing the safety of new drugs for Pfizer in New Haven were Black.

Forty-five years after the end of the Tuskegee syphilis study, those figures should give us pause. In 1979, seven years after Peter Buxtun blew the whistle on the Tuskegee study, the Belmont Report warned against the recruitment of vulnerable groups such as “racial minorities” and “the economically disadvantaged” for medical research. ¹¹ The problem is not just that vulnerable groups are less capable of protecting themselves from potential harm—or, as the Belmont Report put it, “easy to manipulate as a result of their illness or socioeconomic condition.” It is not even that most research sponsors in the United States, in stark contrast to the rest of the developed world, fail to guarantee that they will pay the medical bills of subjects who are injured in their trials. ¹² The problem is also that it is unfair to ask a vulnerable group to bear the potential risks of research if they are unlikely to benefit from the results. If there are many Americans less likely to have access to the benefits of drug research than poor, uninsured Black men in New Haven, it is hard to imagine who they might be.

The authors of the Belmont Report were especially worried that researchers would be tempted to recruit subjects merely because of “their ready availability in settings where research is conducted.” At the time, that sentence would have called to mind settings such as the Willowbrook State School, where researchers intentionally infected mentally disabled children with hepatitis A, or the Iowa State Penitentiary, where, in the early 1970s, researchers fed prisoners a vitamin C-free diet through gastrostomy tubes for three months to see if they would get scurvy. ¹³ (They did.) Now that research sponsors conduct Phase I trials on the poor, there is no need to depend on institutional settings such as these. Contract researchers can ensure ready access to willing subjects by setting up in bleak urban neighborhoods next to the pawn shops, the plasma centers and the payday loan companies.

Twenty years ago, a spokesman for Eli Lilly explained to the Wall Street Journal that the research subjects the company was recruiting from a homeless shelter were driven by “altruism” and wanted “to help society.” ⁵ Even at the time, that explanation sounded implausible. As Grady and her colleagues found, the vast majority of subjects in Phase I trials sign up because they need the money. Not only was money the most common motivation; it was cited at a rate five times that of the nearest competitor (57.6% vs 11.1%). When asked to compare money versus helping future patients as a motivator, 72.3% cited money while only 24.7% cited the desire to help future patients. Over 75% of subjects said they would be more willing to endure invasive procedures if they were offered more money.

Does that mean that money is more important to their decision making than the risks of the studies? No, of course not. It should come as no surprise that poor people—even desperately poor ones—are reluctant to enroll in studies they believe might injure them. Nor is it a surprise that subjects prefer low-risk studies without invasive procedures, conducted by a familiar physician. Why substantially fewer subjects were willing to enroll in first-in-human drug and vaccine trials and in studies of psychiatric drugs than in other studies was not explored. However, it is worth remembering that such studies have produced several high-profile incidents in recent years, such as the suicide of Tracy Johnson during a Phase I duloxetine study at an Eli Lilly trial site in 2004, ¹⁴ the life-threatening injuries of six paid subjects in a TGN1412 trial at Northwick Park in 2006, ¹⁵ and last year’s Phase I study of BIA 10-2474 at a BioTrial facility in France, which left one subject dead and four others with possible brain damage. ¹⁶

If there is a surprise in these companion studies, it is the conclusion that although subjects in Phase I trials are largely poor and unemployed, their poverty and unemployment “do not appear to affect either their motivations for participation or factors important to their research enrollment decisions.” Exactly how the motivations of poor, unemployed research subjects to enroll in research studies for money could be unaffected by their poverty and unemployment goes largely unexplained. Grady and her colleagues suggest that the decision to sign up for a study may be like looking for a job, where subjects evaluate risks and decide if they are worth the payment, possibly reassured “by knowing that adverse events are usually mild for healthy volunteers.”

A more nuanced explanation has been offered by medical sociologist Jill Fisher and her research team, who conducted semi-structured interviews with 178 subjects in Phase I studies at seven different United States trial sites. ¹⁷ “Some of our informants admit that they are hanging on ‘by the skin of their teeth’,” Monahan and Fisher write, “staying in motels and living from day to day, just a step away from homelessness.” ¹⁷ Nearly 62% of their subjects were either Black or Hispanic, and 17% had household incomes of less than US$10,000 a year, well below the poverty level. Almost 80% had taken part in more than one study, with 26% participating in more than eleven. Some said they had enrolled in over 200 studies.

While the choices of these subjects to enroll in studies may appear to be rational calculations, the way the subjects speak about their decisions suggests something deeper and more emotionally driven. One obvious factor is financial desperation. “I guess the desperation far outweighed the concerns,” said a Hispanic man on his tenth study. “You know, when someone’s desperate, like they are not even gonna think twice, so I guess that’s where I was at.” ¹⁸ Some subjects directly compared their fear of risky studies to a far greater fear of poverty. “I was devastated, you know your money’s cut in half; my rent was not [halved] or my gas bill,” said a White woman who had recently lost her job. “And I frankly found that more terrifying than whatever I was going to do during the trial.”

For many subjects, the decision to enroll was shaped by their experiences with prison and the police. One subject said, “(T)here’s a little risk, but it’s well worth it. Like it’s not as risky as going to sell some drugs or robbing a bank, you know.” Some explained that the risks of an experimental drug were minimal compared to the risks of their everyday lives, where arrest was a constant threat. “I’ve been in different situations, so this little drug, that’s not gonna scare me more than what’s going on out where in them places I’ve lived at,” another subject said.

In contrast to Grady and her colleagues, who believe that their findings should “alleviate some concerns about distorted judgment among healthy volunteers,” Fisher worries that many of the subjects she interviewed have become desensitized to the potential risks of studies. ¹⁹ Partly this is because of the casual, routinized fashion in which staff members present those risks, but it is also because of the particular way that trials are typically conducted. Most Phase I trial sites are standard “feed ‘em and bleed ‘em” facilities, with relatively unsophisticated medical capabilities; most of the studies conducted at those sites are homogeneous, with little variation between protocols; and most are set up for maximum efficiency, like a factory assembly line, with minute-by-minute scheduling and barcodes for the subjects. Even the language used by staff members transforms harms into data points; staff call the injuries produced “AEs,” so that even the word “adverse event” often goes unmentioned. The result is what Fisher calls “the banalization of risk.” Many subjects she interviewed appeared far less worried about the risk of being harmed by an experimental drug than the risk of failing to qualify for the next study.

If current practice is any indication, most research sponsors are comfortable with using America’s vast supply of poor, uninsured minorities to test the safety of experimental drugs. Morally justifying such a practice requires one to imagine the decisions of subjects as fully autonomous and free, rather than the product of financial desperation. It also requires imagining that the bargain being offered to subjects is fair—not just that the payment is reasonable, but that the research oversight system will protect them from harm, that the conditions of the study are not degrading, and that if subjects are injured, their medical bills will be paid and they will be compensated for their pain, suffering and the inability to work. No doubt there are some United States Phase I trial sites where most of this is true. But I am afraid there are many more where it is not.