When clot-busting drugs fail to stop a heart attack, surgery to reopen a clogged artery is the most effective way to treat a patient, cutting the risk of death in half, a new study shows.

The study, published in this week’s New England Journal of Medicine, could revolutionize the way smaller U.S. hospitals treat heart attack patients, forcing them to set up arrangements with larger hospitals capable of performing the surgery, known as angioplasty.

The data, funded by the British Heart Foundation and conducted at 35 British medical centers, also found that angioplasty, where a tube is threaded into the heart so a balloon can reopen the clogged area, reduces by half the risk of a stroke, heart failure, or later heart attack.

Angioplasty is helpful even when factoring in the extra time it may take to move a patient to a hospital where the procedure can be performed, the study shows.

“It’s certainly changed our practice, and rescue angioplasty is the standard of care now in the UK,” chief author Anthony Gershlick told Reuters.

Gershlick, who works at the University of Leicester, said many U.S. hospitals do nothing if an electrocardiogram shows that clot-busting drugs have not worked. He noted that in 30 to 40 percent the drugs are ineffective.

The new study of 427 volunteers found that giving more clot-busting drugs offered virtually no benefit over conventional treatment with the blood thinner heparin. “It’s almost like doing nothing, Gershlick said.

About 30 percent in both groups died, had another heart attack, suffered a stroke or developed severe heart failure.

But among those who had the option for angioplasty when the clot-dissolving drugs failed, the rate was 15 percent, even though many patients had to be transferred to another hospital for the surgery. That transfer usually delayed the surgery by an average of 84 minutes, the study shows.

Six months after the initial heart attack, the Gershlick team found that 6 percent of those who had received the surgery had died while nearly 13 percent of the patients who had been treated only with drugs had died.

The likelihood of a second heart attack was also far lower for the angioplasty group - 2.1 percent compared to 8.5 percent among those who received conventional care and 10.6 percent with the second dose of clot-busters.

One side effect of the drugs is bleeding. Five patients died of bleeding problems in the group that received a second dose of clot-busting drugs; three died with conventional treatment; and there were no bleeding-related deaths in the angioplasty group.

Gershlick said it was understandable that community hospitals were not pushing for angioplasty if clot-busters failed. Until now, there had been little evidence to support a backup treatment, except for an old study from the 1990s suggesting that a second dose of the drug might help. But that research used an early, less-effective clot-busting drug.

When clot-dissolving drugs fail, “you need to call the ambulance, get the cath lab team in, and do an angioplasty,” Gershlick said.