Blood On Tap, Part 2: An Ethical Dilemma in Emergency Research

"Ethics is not the process of giving 'correct' answers. Neither is it allowing everyone's opinion to stand unchallenged. It is the process by which we test either some, most, or all of what we do against the reasons that we do it."
-John Freeman, MD

By the time you read this, paramedics of the Denver Health Paramedic Division will have randomized the first patients into a nationwide prehospital study of the oxygen-carrying blood substitute Polyheme. The study is groundbreaking, historical and utterly impossible without a vital piece of legislation known as 21 CFR 50.24, an exception to the requirement of informed consent in medical research. The Polyheme trial will be another chapter in the widening debate over the ethics of risk and informed consent in emergency medical research. This is an up-close look at the history and roots of the controversy as some of the experts face off over an issue with far-reaching implications in the future of emergency medicine.

Bad Blood

Possibly the most difficult question to ask of history is, "What if?" It's difficult to say if the seemingly minor 1996 amendment to the code of federal regulations known as 21 CFR 50.24 would have made much of a stir if it hadn't gotten off to such a bad start. If the first few research studies conducted under the new waiver of informed consent had gone off without a hitch, the alteration may have gone unnoticed by anyone outside of the medical research community. But it didn't work that way.

March 31, 1996. The scientific community was surprised to learn that Baxter Healthcare Corporation was halting its phase III trauma study of the oxygen-carrying blood substitute Hemassist after enrolling just 112 patients. The product, which had already showed promise in multiple phase II studies, had not demonstrated the lifesaving potential that researchers had predicted. Instead, the patients who received Hemassist had demonstrated a mortality rate that not only exceeded Baxter's predictions, but far exceeded the mortality rates of patients in the control group who had received whole blood.

In all, 24 of the 50 patients who received Hemassist died. These numbers would have been merely an unfortunate footnote in emergency research if not for one critical distinction: None of the patients had consented to be involved in the study. Baxter's Hemassist study was the first research ever conducted under the new federal regulation allowing some emergency research to be exempted from the requirement of informed consent.

Hemassist's failure launched a swarm of negative press and brought CFR 50.24 into the spotlight. The media wanted to know, "How could this happen?" An Associated Press headline proclaimed "Blood Trials Done Without Consent." Websites devoted to warning the public of becoming guinea pigs popped up, and reporters unfairly compared the research to barbaric studies conducted by the Third Reich during World War II.

The shockwave of bad press sent Baxter Healthcare's stock plummeting and eventually led Baxter to withdraw Hemassist from all clinical trials, costing the company an estimated $500 million.

All predictions had been that the trial would be a perfect debut for the new exception to informed consent rule. A product that had proved itself in multiple phase II studies would be given to critically ill trauma patients who could not possibly provide informed consent. It was the type of study for which 50.24 was designed. What could go wrong?

The Rule (In a Nutshell)

When I say "the rule," I am specifically referring to the Code of Federal Regulations Title 21, Volume One, Part 50.24, which provides the first exception to a long-standing rule requiring informed consent of all human participants in medical research. The authors of 50.24 (as it is known in research circles) didn't offer the exception lightly.