By Arthur Kellerman
The Atlanta Journal Constitution

ATLANTA — Typically, health stories focus on medical "breakthroughs." But occasionally, a treatment turns out not to be as effective as hoped.

These so-called "negative" studies are often regarded as failures. They shouldn't be. Identifying ineffective or inefficient treatments is extremely important because it allows us to focus our health care spending where it can do the most good.

The April 1 online edition of the New England Journal of Medicine contains an excellent and courageous example of a negative study. For many years, doctors and the public have been drawn to a high-tech device called an automated external defibrillator, or AED.

AEDs are designed to treat cardiac arrest. Each year, approximately 160,000 Americans collapse outside the hospital from cardiac arrest. More than 90 percent die. For more than three decades, the American Heart Association has fought cardiac arrest by promoting a concept known as "The Chain of Survival."

It has four links: early access to emergency services (via 911), early CPR (ideally from a family member or bystander), early defibrillation (to restart the heart) and early advanced care by skilled EMS personnel.

AEDs strengthen the third link by allowing a first-responding firefighter and even a citizen to promptly deliver countershocks to a victim of cardiac arrest. After studies showed that putting AEDs on firetrucks and in public settings like airports saved lives, it was only a matter of time before AED advocates proposed that they should also be placed in homes. One company even secured FDA approval to sell an AED "over the counter" without a doctor's prescription.

The argument for home use is this: Cardiac arrest is an unexpected event, 80 percent of cases occur at home, most require treatment with a defibrillator and the sooner the defibrillator is used, the more the odds the victim will survive.

Given these facts, it seemed logical that placing AEDs in homes would save lives.

There was only one problem. There was no scientific proof that this costly strategy works better than teaching family members to call 911 and start CPR.

To answer the question, the National Institutes of Health sponsored a huge seven-nation study. More than 7,000 patients agreed to participate. Half were given an AED, and their spouse or companion was trained to use it. The other half relied on CPR and a call to 911. The median follow-up interval was three years.

The results were eye-opening. The two groups had identical rates of death from all causes. Those who kept an AED in their home were no more likely to survive cardiac arrest than those who did not.

Is this study a failure? Absolutely not.

With this knowledge, we can let go of the false hope that a massive program to place AEDs in homes will reduce deaths from cardiac arrest. Had this study not been done, millions of people might be convinced to shell out $1,500 to $2,000 apiece to buy an AED. Over time, public pressure could grow to make AEDs coverable under private and public health insurance.

All that spending would have been for naught. We need more comparative effectiveness studies like this one to make health care more effective and affordable.

With this knowledge, families can focus on interventions that work. For example, CPR doubles or triples a victim's chances of surviving cardiac arrest. Once you learn CPR, you can take it anywhere — to the pool, to work and even to grandmother's house.

With this knowledge, communities can improve the quality of their emergency medical (ambulance) systems. Today, rates of successful cardiac resuscitation vary widely from one city to the next. If the poorest-performing cities improved to the level of the best, we could save thousands of lives.

But the most remarkable finding of all is the researchers had fewer events to study than expected. Why? Because prevention works! Nationwide, the rate of cardiac arrest is falling year by year, because fewer people smoke and more monitor their diet and exercise. Also, doctors have improved treatment of high blood pressure, diabetes and coronary artery disease — conditions that increase the risk of cardiac arrest.

Every ER doc knows that the best way to survive a cardiac arrest is not to have one in the first place. In this group, survival is 100 percent. Each year tens of thousands of lives are saved through prevention. And that is worth celebrating.

Dr. Arthur Kellermann is professor of emergency medicine and associate dean of health policy at Emory University School of Medicine.

Copyright 2008 The Atlanta Journal-Constitution