Two years ago, researchers for Deerfield-based Baxter International Inc. began the first advanced trial of an oxygen-carrying blood substitute with great hopes that the product might one day revolutionize emergency medicine by providing a ready treatment for people of all blood types.
That didn’t happen. The trial was halted because out of 52 emergency trauma patients around the country who received the blood substitute called HemAssist, 24 died–two more than would have been predicted given the severity of their injuries and far more than the death rate in a group that received only routine treatment.
Though the trial’s disappointing results were widely reported last summer, the assessments did not examine its most controversial feature: Consent was never obtained from patients in the study. But recently, some medical experts have called for federal regulators to revisit the policy that permits some experiments on humans to be conducted without the consent of patients or their families.
The absence of consent was allowed under a historic 1996 change in U.S. Food and Drug Administration regulations. Baxter’s trial was the first study to take advantage of the new rules, which were designed to facilitate research in emergency medicine that could not happen if doctors had to take the time to get consent.
But in addition to encouraging new discoveries, the regulatory change broke with a 50-year-old doctrine requiring informed consent in virtually all experiments on humans. The failure of the Baxter trial has led some ethicists to question the wisdom behind the rule change.
“People get involved in something to their detriment without any knowledge of it,” said George Annas, a professor of health law at the Boston University School of Public Health. “We use people. What’s the justification for that?”
First articulated at the Nuremberg trials as a way to prevent hideous experiments such as those performed by Nazi doctors during World War II, informed consent became the guiding principle of research on humans in this country. Violations of the rule normally have met with wide outrage, from the clandestine federal syphilis experiments on African-American Tuskegee airmen to recent allegations that researchers at the National Institutes of Mental Health gave subjects the “date rape drug” ketamine without fully disclosing some psychosis-inducing side effects.
In contrast to the secrecy of those studies, however, the HemAssist trial began under significant public scrutiny as the first under the new FDA regulations. There also was little reason to think the product would have an adverse effect, since initial trials in Europe did not indicate any problem with unexpected deaths.
Although no Chicago-area hospitals participated in the trial, one of its co-leaders was Dr. Max Koenigsberg, director of emergency medical services at Illinois Masonic Medical Center and an overseer of the Chicago EMS system. Koenigsberg said even though the trial’s outcome was negative, the procedures used were sound.
“Nuremberg was never intended to withhold needed therapies from patients,” Koenigsberg said. “The problem with Nuremberg and Tuskegee was that the medical community and the public were never told.”
The new regulations require a level of community notification unheard of for most scientific studies, including extensive community meetings, press releases and post-study follow-up.
Yet it’s unlikely that any of the patients who wound up being transfused with the blood substitute in emergency rooms had been reached by the public notification. Even supporters of the FDA rules say such notification cannot replace direct consent from patients or their relatives.
“Public notification means nothing,” said Dr. Arthur Caplan, director of the Center for Bioethics at the University of Pennsylvania. “I know people are enamored of it, but it means nothing.”
Caplan was lead author of a December article in the Journal of the American Medical Association that called for modifications to the FDA informed-consent regulations. His suggestions include requiring researchers do more to prove that a study can only be performed on patients who are so incapacitated by trauma that they cannot give consent themselves.
Waiving informed consent of patients has always been allowed for studies with very minimal risk, in which researchers use only data that would be collected in the normal course of clinical care. In addition, federal regulators made rare exceptions to the informed-consent requirement when experiments involved patients in cardiac arrest or similar conditions when death appears likely.
The 1996 rule allows institutions to waive informed consent for studies of emergency treatment in which available treatments are unproven or inadequate.
Human blood, for example, currently cannot be given while a patient is in the ambulance speeding from the scene, and there can be a delay at the hospital if blood is in short supply. Blood substitutes such as HemAssist would not require refrigeration or cross-matching for blood type, and so would be immediately available.
The regulations state that informed consent can be waived only if the patient is unconscious or incapacitated, and the family cannot be reached. Risks from the experiment must also be reasonable compared to standard treatment.
Animal studies and preliminary trials on humans suggested HemAssist was relatively safe, although some research suggested it caused an increase in blood pressure. The U.S. Army considered using similar blood products during the Persian Gulf war in 1991 but decided against it because of the blood pressure concern.
Trauma patients in the Baxter study received two to four units of HemAssist within an hour of arriving at the hospital, in addition to regular human blood and a saline solution. Randomly assigned patients in a control group got only human blood and saline. Researchers hoped HemAssist, like human blood, would help supply oxygen to the body’s organs in the critical time immediately following severe injury from a car accident or shooting.
That justification met the federal requirement that the product could be shown to fill a need left by human blood, which often is not immediately available. But Boston University’s Annas believes such theories are not enough for researchers to bypass informed consent.
“It’s presented as if it’s better, and we don’t know that,” Annas said. “That’s the definition of research–if we knew it worked, we’d be doing it.”
Analysis of the Baxter data has given no conclusive explanation of why the observed death rate for HemAssist was 46.2 percent, compared with the predicted rate of 42.6 percent. In another mysterious finding, mortality for the control group was just 17.4 percent, even less than the rate of 35.5 percent that would have been expected given the severity of that group’s injuries. Given such irregularities, Koenigsberg said, it is impossible to show HemAssist caused the excess deaths.
In fact, many experts believe the FDA rules–and even Baxter’s study–will be beneficial to patients in the long run.
Dr. Norman Fost, director of the medical ethics program at the University of Wisconsin, said some doctors had resorted to unofficial, unsupervised emergency medicine research before the FDA rules were changed.
“They would do informal studies, with the disadvantage that the data was never collected in a systematic way so you could tell whether the benefits were worth the risks,” he said. “Most standard treatment in emergency room settings has never been shown to be safe and effective.”
Fost, who lobbied for the FDA’s relaxation of informed-consent requirements, said the Baxter study could have the unintended positive effect of deterring unmonitored research.
“This study may have prevented many deaths,” Fost said. “Otherwise, the drug may have . . . been used in an uncontrolled, unstudied way.”
The goal, Fost says, is to do what the patient would want if he or she could give consent.
A post-study survey at Lehigh Valley Hospital in Allentown, Pa., revealed that most families were satisfied with the treatment even after the trial proved a failure, according to Dr. Mark Cipolle, director of surgical research at the hospital.
“I was a little nervous about calling the families of patients who died,” Cipolle said, “but it wasn’t bad at all.”
No lawsuits have arisen from the blood trial, a Baxter spokeswoman said.
Still, Baxter’s main blood substitute competitor, Evanston-based Northfield Laboratories Inc., said it has no plans to seek waivers for informed consent for studies of its product.
“I honestly am not convinced it’s a method we need to use,” said Northfield Chairman and CEO Richard DeWoskin. “I think experimental research should be consensual.”
Northfield has tried to limit preliminary trials to patients who are less severely injured, or whose families can be reached in time to give their consent.
“We miss some patients, but we still get enough to do the study,” DeWoskin said.
In fact, no company besides Baxter has conducted a no-consent experiment under the new rule, and only a few have begun the confidential preliminary process of preparing for one, FDA officials said.
That may indicate that the rules as currently formulated are still too burdensome for drug companies, Fost said.
Yet Bonnie Lee, a senior policy analyst for the FDA, said even the 1996 change was intended to be strict. Still, Lee said, the fact that few companies have plans to use the new rule may be significant.
“If it’s not going to accomplish what we wanted it to accomplish, maybe it needs to be changed (again),” Lee said. “In some cases that could mean tightening the regulations or opening up other areas. We would never consider opening it so broadly as to allow just anything to happen.”
