by Alexander Volokh
August 2, 1995

Imagine that you've had a heart attack. Amid the wails of sirens and the flash of red lights, paramedics are rushing to help you. Skilled in the art of CPR, they remove your shirt and prepare to revive you. But much to their surprise—and yours—a sign about your neck reads, "Do not resuscitate."

The real shocker is that, unbeknownst to you, this sign was placed there by the Food and Drug Administration. In April 1993, the FDA blocked testing of the cardiopump, a new cardiopulmonary resuscitation device for heart-attack victims. As a result of FDA policy, thousands of people with heart disease are literally being protected to death.

The cardiopump, manufactured by Ambu International of Denmark, is a modest device. It weighs a mere pound and a half and looks like a modified toilet plunger, with a pliable cup that fits onto the heart-attack victim's chest and a combination hand- grip/pressure gauge instead of the wooden handle. Manual CPR exerts downward pressure on the chest, but the chest has to re- expand naturally. The cardiopump can apply pressure in both directions. Says Dr. Jeffrey Shultz of the University of Minnesota: "It turns the chest into a bellows. It allows you to pull blood back into the heart and air back into the lungs."

Just how effective is the cardiopump? In St. Paul, Minn., where tests were carried out, 54% more patients survived long enough to be admitted to intensive care if they were given cardiopump-assisted CPR, and 35% more survived to be discharged from the hospital. (Because of small sample size, the hospital discharge statistics aren't statistically significant.)

U.S. tests for the cardiopump began in 1992 and continued until April 1993, when FDA barred further use of the pumps. According to FDA spokeswoman Sharon Snider, "This product is considered to be a 'significant risk device,' and informed consent is required for any trials of significant risk devices."

The FDA's view is shared by virtually no one. None of the institutional review boards at the 12 medical institutions where the tests had been carried out called the cardiopump a "significant risk device." More important, the cardiopump is already in use around the world—Austria and France (where it is standard in physician-manned ambulances), England, Germany, Sweden, Canada, Australia, Japan and Chile, among others. The FDA, in its treatment of the pump, stands alone.

The concepts of "significant risk" and "informed consent" may well be important, but their relevance here is highly questionable. "Significant risk" is meaningless unless you compare the risk of the device with the risk of not using it. The survival statistics of manual CPR are dismal. About one million people have cardiac arrests every year in the U.S., and an estimated 700,000 are given CPR. About 680,000 of these 700,000 do not survive to leave the hospital. If the FDA had jurisdiction over manual CPR, it might well prohibit it on the same grounds.

As for informed consent, keep in mind that these are patients whose hearts have literally stopped beating. "You're talking about a life-or-death situation where our current therapy is nearly hopeless," explains Dr. Joseph P. Ornato, chairman of the American Heart Association's emergency cardiac care committee. "The reality is, there is no reasonable means to get informed consent."

The issue of informed consent for unconscious patients may be a complex question in some cases. This, however, is not one of them. The cardiopump is hardly experimental. In France, the average approval time for a medical device is a year and a half; the cardiopump was approved in March 1993 after just nine weeks. But the FDA has been of little help, aside from vaguely offering to "work with the company" in devising acceptable test procedures. It will take years just to produce studies with enough "informed consent" patients to satisfy the FDA.

The world-wide medical community is somewhat nonplussed by the U.S. government's attitude toward medical technology. The feeling is widespread that the Americans are shooting themselves in the foot by making it so difficult for new drugs and devices to reach the market. As Dr. William Grace of St. Vincent's Hospital in New York put it: "We are a Third World nation when it comes to the development of ... drugs for people with life-threatening diseases."

Keith Lurie, a professor of medicine at the University of Minnesota who helped develop the cardiopump, estimates that the device will increase one's chances of survival by about 10-50%. That kind of improvement "has not been reported since the first description of manual CPR more than three decades ago." Since 20,000 CPR patients a year survive to leave the hospital, even a 10% increase would mean 2,000 extra lives saved. In the St. Paul trial, the increase was 35%, which extrapolates nationally to 7,000 lives, or roughly 20 a day.

Yet strange as the FDA's actions might seem from a medical standpoint, they are perfectly understandable politically. If the FDA mistakenly approves a device that has adverse side effects, this would result in highly undesirable publicity. On the other hand, if the FDA delays in approving a life-saving device, the people who die are politically invisible. For a regulator, it's easy to err on the side of overcaution, even when the results are deadly.

The irony is that we applaud FDA approvals when we should really be asking why we had to wait so long. In 1987, when FDA approved streptokinase, the first drug that could be intravenously administered to reopen the blocked coronary arteries of heart-attack victims, it was estimated that the drug could have saved 22,000 lives in the two years between the filing of the new drug application and the approval. But these deaths received less media attention than the deaths of five patients last year during a test of the hepatitis B vaccine fialuridine.

Now Congress is considering charging device manufacturers "user fees" of up to $50,000 per new-product application; FDA Commissioner David Kessler calls this the solution to FDA's backlog problem. But backlog isn't the problem. No amount of staff and funding increases can erase the bureaucratic incentive to impose restrictions as illogical as informed consent from patients with no pulse.