By Robert Davis
USA TODAY
Copyright 2006 Gannett Company, Inc.
As Hilary Williams hung from her seat belt in the overturned wreckage of her truck — legs broken, colon ruptured and lung bruised — blood oozed from torn arteries and veins.
Hilary, 27, and her sister, Holly, 25, who was injured less severely in the crash, waited half an hour for medical help to arrive. By the time an air ambulance reached them at the crash site, about 45 miles south of Memphis, it was midday on March 15. Hilary’s face was pale, her lips were blue, her faint pulse was fast, and her blood pressure was 55/0. She was in shock.
Because she was close to death, she instantly became eligible for enrollment in a controversial clinical trial. But because she was in no shape to consider the risks, the flight nurses did not have to obtain her consent before giving her an experimental blood substitute.
Within moments, fluid the color of merlot was dripping into her veins. Hilary, a singer/songwriter and the daughter of country singer Hank Williams Jr., became one of thousands of people across the nation who, while fighting for their lives over the past 10 years, unknowingly became test subjects in medical trials.
With waived-consent studies becoming more prevalent, critics question whether the public understands how they work and whether test subjects get adequate protection.
Although the study that included Hilary Williams has brought criticism, she believes the infusion of blood substitute helped save her. “I thought I was going blind,” she says as she recovers in Nashville. “Everything was dark. My eyes were open but I couldn’t see anything. I felt at peace. I kept wanting to fall asleep, but a truck driver who stopped kept telling me to wake up and to hang on.”
But another trial, which is reported in today’s Journal of the American Medical Association (JAMA), was halted because a device called the AutoPulse, which was used to revive cardiac-arrest victims, failed to save more lives than when rescuers performed cardiopulmonary resuscitation.
According to the study, as many as 10 people in one city may not have been revived because of their participation. The JAMA study does not identify the city. But doctors have identified the site as Seattle — which has one of the highest resuscitation rates in the nation — in discussions at medical meetings since the AutoPulse trial was halted in March 2005.
Patients in these types of studies — often financed by manufacturers of the tested product — are treated under a broad federal rule that allows researchers to test emergency treatments on patients with specific, life-threatening medical conditions without their explicit consent as long as they remain under close watch of independent reviewers.
Federal officials and medical researchers say there is no substitute for this testing of emergency treatments for the leading causes of death: trauma and cardiac arrest, which claim hundreds of thousands of lives in the USA each year. To the researchers, studies with negative findings are as important as successful trials in their search for the best treatments.
Studies have included large, multi-city, randomized trials, which scientists consider the gold standard for medical research. One study that ended in October 2003 compared ways to revive patients in cardiac arrest and taught doctors and city leaders how to deploy automated external defibrillators.
Federal officials say such research is key because emergency treatments are largely untested. “Many of the treatments that are currently used in life-threatening circumstances have not been formally studied,” says Sara Goldkind, a Food and Drug Administration (FDA) bioethicist. “They have not come to be standard-of-care based upon rigorous scientific data.”
Doctors agree. “The only way to make any sort of significant advance in this final frontier is to perform these studies with waiver of informed consent,” says Martin Croce, a Memphis trauma surgeon. “It’s clear that multiple cascades of bad physiologic things happen shortly after injury. Once that snowball begins to roll down the hill, it’s very difficult to stop. To stop it, you need to intervene early. That is at the scene.”
Critics say the trials are conducted on a slippery ethical slope.
“It has to be done carefully,” says Nancy King, a University of North Carolina ethicist who says corporate sponsors of trials don’t always want to make key study information public. “It is so terribly easy to cut corners and say this doesn’t matter as much and we will just put some notices in the newspaper. It makes me concerned.”
Two trials in particular — the ongoing blood-substitute trial in which Williams became involved and the ill-fated AutoPulse study — illustrate their ability to answer scientific questions and the ethical controversy that surrounds them.
The PolyHeme trial
About five minutes after flight nurse Cindy Bailey began infusing the blood substitute PolyHeme into Williams’ veins, they were lifting off for the 20-minute helicopter flight to University of Tennessee Health Science Center in Memphis, where Croce waited with a trauma team.
Ambulances can’t carry blood because it must be kept cool and typed to match each patient. Researchers have sought for decades a substitute that can be carried by medics in the field.
PolyHeme, which was born from research after the Vietnam War, is made from chemically modified human hemoglobin by Northfield Laboratories of Evanston, Ill.
The study in which Williams was enrolled is the most widespread use yet of PolyHeme. The trial has raised concern among some ethicists and alarm in Congress, where Sen. Charles Grassley, R-Iowa, chairman of the Finance Committee, is conducting an investigation.
Grassley is concerned that people who live in the 19 states where PolyHeme is being tested have had inadequate notice about the trial. The FDA requires that community input be sought in the regions around test sites. “It is outrageous that, for all intents and purposes, the FDA allowed a clinical trial to proceed, which makes every citizen in the United States a potential ‘guinea pig,’ without providing a practical, informative warning to the public,” Grassley wrote in a letter to the FDA in February.
“This study raises significant ethical issues,” says King, who co-wrote an article in the current edition of The American Journal of Bioethics that faults the PolyHeme trial for withholding available treatments, namely blood transfusions, from patients.
In the trial, PolyHeme is appropriately given to patients without their consent in ambulances, she says, where blood is not available. But some patients, including Williams, also get the blood substitute for up to 12 hours in the hospital, where blood is available.
King says some patients who die from their injuries therefore will perish while “being denied an available treatment.” Though she favors the concept of waived-consent trials, she says PolyHeme could be tested on the battlefield or during elective surgeries in which “you could get prior consent.”
Steven Gould, chief executive of Northfield Laboratories, stands behind the PolyHeme trial. He says it is too complex to be performed on a battlefield and elective surgery is not the crisis situation best suited for the study of PolyHeme. Of King’s comments, he says, “We have different points of view, but that is what this is all about.”
Amid the controversy, some cities stopped enrolling patients while the data-monitoring board reviewed preliminary findings.
The data-monitoring board includes clinicians who specialize in the area under study and biostatisticians and bioethicists who review study data and “adverse event” reports. This committee tells the study sponsor when a trial needs to be modified or stopped. The members of this group have no financial interest in the outcome.
All of the sites resumed the trial after the board concluded that patients were safe and that it was safe for the trial to finish.
Results of the study are not expected for months, but Bailey and her partner Cindy Parker believe that they have already seen lives saved by the blood substitute.
“It’s unbelievable,” Parker says. “If Hilary had not received this, I don’t think she would be here today.” Says Bailey: “I think she would have arrested (died) before we got to the hospital.”
EMS officials prefer to act on big trial results because such studies determine scientifically — using a large number of patients who are randomly assigned to be tested — whether the treatment works.
“In the old days, manufacturers would come up with a whiz-bang product, and we’d buy it and try it,” says Michael Poniatowski, operations director for EVAC Ambulance in Daytona Beach, Fla.
“If it didn’t work, we chucked it,” he says. “If it worked, we kept it. Today, we need to see the data to back those statements up.”
The AutoPulse trial
One trial that was conducted amid swirling accolades and anecdotes of lives being saved was halted early after a sizable number of test subjects died.
The results of the trial, published today in the Journal of the American Medical Association, illustrate the key role that independent trial monitors play in intervening when test subjects need protection.
The device known as the AutoPulse — a portable, battery-operated device that wraps around an adult’s chest and squeezes the ribcage rhythmically to move blood through the body — began generating awe-inspiring stories of cardiac arrest survival in 2004. EMS agencies began scooping them up.
“Our medics loved the device and they reported what they felt anecdotally were amazing saves,” says Marshall Isaacs, director of emergency medical services for the San Francisco Fire Department. “They said patients pinked up.”
Emergency medical crews in Richmond, Va., who also report their findings in today’s JAMA, had more promising results when comparing the use of AutoPulse with previous rescue attempts.
But in the five-city trial that was supposed to enroll 1,837 victims of cardiac arrest, researchers tested the device head-to-head against manual chest compressions during CPR by randomly placing patients into one of two arms of the study. Half of the patients got compressions from the device, and the other half got manual compressions.
Because the federal rule for waiving consent requires clinical investigators to consult with the community and instructs the institutional review board to consider public comments before approving the study, Seattle officials met with community groups.
The survey of 400 Seattle adults found that 75.3% said it was OK to carry out the study when consent was not possible; 18.6% objected.
With a majority of those consulted supportive of the trial, researchers began enrolling patients in Seattle on Sept. 9, 2004. Calgary, Alberta; Vancouver, British Columbia; Columbus, Ohio; and Pittsburgh followed that year.
The data-monitoring board halted the trial on March 31, 2005,finding that the device appeared to be harmful to the patients.
About 6% of patients treated with the AutoPulse survived, while roughly 10% of patients who were not treated with the device and got traditional, manual CPR survived.
The FDA says the data-monitoring board provides a safety net.
“Given that people are put into these trials without consent, you owe them a continuous watch on how they are doing,” says Robert Temple, director of the FDA’s Office of Medical Policy.
And results must be reported, Temple says. “These are studies using a vulnerable population,” he says. “You owe the world an explanation of what you were doing and what the results were.”
Within the results published in today’s JAMA are the numbers that troubled the researchers and stopped the study. Of the 65 people who were treated in Seattle with the AutoPulse, seven were discharged from the hospital alive, compared with 17 of the 66 who received manual CPR.
“We’ve thought about that patient population a lot,” says Michael Sayre, associate professor in the Department of Emergency Medicine at Ohio State University in Columbus. He hopes that people across the nation learn from the deaths. “If we had not done this study, then maybe this technology would be adopted on all patients and that might have hurt more people than it helped,” he says.
The future
Medical researchers and FDA officials say that they will be looking in coming months into how these trials of emergency devices and treatments have been conducted over the past decade to see if improvements are needed.
“Is this a good reg? Is it perfect? Can it be better? Can it be worse? Those are important policy questions that we feel we can and should address,” Gould says. “To do research without prospective informed consent defies what we would all say is good ethical practices, so you have this conundrum.”
That quandary continues.
The National Institutes of Health has set aside $5 million to study both cardiac arrest and trauma over the next five years in trials using the waived-consent rule.
Recovering in her mother’s home in Nashville, Williams says her music career is on pause. She was scheduled to be photographed for a pop record album before the accident, but she is writing songs and longs to return to the stage.
And she says she is happy that science could learn from her care.
“I think it’s a great thing,” Williams says. “God and PolyHeme basically saved my life.”
Time and data analysis will tell what role PolyHeme played.