An article yesterday on a study of the drug lidocaine incorrectly stated the number of control-group heart patients -- those who had not received the drug -- who developed dangerous heart rhythms. The correct number is 12.

A shot in the arm with a commonly used heart drug, if given by ambulance workers to patients with chest pain, could save thousands of heart-attack victims who now die before reaching the hospital, according to a study published in today's New England Journal of Medicine.

The study by two Dutch researchers of more than 6,000 patients shows for the first time that injecting the drug lidocaine in a shoulder muscle as soon as paramedics arrive dramatically reduces the chances of a victim's heart developing dangerous disturbances in the next hour.

The drug is generally administered intravenously once heart-attack victims reach hospital emergency rooms, to prevent or correct abnormal rhythms common in the early hours of a heart attack. But lidocaine injections are not routinely given prior to arrival because until now such pre-hospital treatment had not been proved effective.

Of the 650,000 Americans who die of heart attacks each year, more than half die of abnormal heart rhythms before they arrive at hospitals, according to the American Heart Association.

In some cases, paramedics on ambulances equipped for advanced life support give the drug intravenously. But the majority of the nation's ambulances do not carry the drugs or electric-shock devices used in advanced life support.

The number of deaths that might be prevented by lidocaine would depend on how quickly it was given after pain began, but "we might expect a possible 30 percent decrease in pre-hospital death" if the treatment were widely adopted, according to Dr. Rudolph W. Koster, the Amsterdam cardiologist who directed the study.

"The findings . . . represent an important advance, and their recommendations, if implemented, will help save many of those who would otherwise die" from abnormal heart rhythms, said Dr. Bernard Lown, a Harvard Medical School cardiologist, in an editorial published in the same issue of the journal.

Lown's editorial went on to disagree with a suggestion by the Dutch researchers that injectable lidocaine be made available directly to patients with heart disease, so that they or their family members could give it at home at the beginning of a suspected heart attack, even before medical help arrived.

Lown said this would result in the drug being overused by patients, with the likelihood that deaths from severe reactions would overshadow lidocaine's lifesaving potential.

In the 33-month study, more than 6,000 individuals with chest pain were assigned randomly to one of two groups. Patients in the treatment group were given 400-milligram lidocaine injections into a shoulder muscle; patients in the control group were not.

Eight treated patients and 17 controls developed the potentially lethal heart rhythm called ventricular fibrillation. However, after the first 15 minutes following the injection, only two treated patients developed the dangerous rhythm, compared with six in the control group.

Paramedics in the Dutch study administered the drug with an automatic injecting device called the LidoPen, manufactured by Survival Technology Inc. of Bethesda. A spokeswoman for the company said the LidoPen sells for about $15. It is six inches long, about the diameter of a lipstick and injects lidocaine into a muscle automatically when a safety cap is removed and the device pressed against the skin.

The Food and Drug Administration approves the device only for emergency use by doctors or paramedics or by certain heart patients when used with an electronic monitor that transmits heart rhythms to a doctor's office over telephone wires.

In the past, injection of lidocaine into muscle has not been widely adopted for heart-attack victims because of the belief that their circulation might not be good enough to promote absorption. In an interview, Koster said a potential side effect of lidocaine most feared by the researchers -- complete standstill of the heart -- had developed only in one treated patient and in one control patient.