Dr. Kasner said he would not recommend closure “for patients with multiple typical stroke risk factors (older age, hypertension, diabetes, smoking, etc.), patients with lacunes, patients with diffuse atherosclerosis, or other likely contributors to future stroke risk.”

But he said, “I am recommending closure for patients who meet the criteria used in the REDUCE study: generally under age 60 with an embolic-appearing stroke (typically a cortical or cerebellar infarct, not small deep infarcts), PFO, and no other cause for the PFO.”

Dr. Spence, a coauthor of REDUCE, was one of the doctors urging caution in applying the new findings. “I think there is substantial risk that in the United States, where inappropriate carotid stenting is rampant, there will be a flood of percutaneous closures of PFO in people who are more likely to be harmed than helped,” he said. “A lot of cardiologists are going to be throwing PFO devices into people who can’t benefit from them.”

“Among people with a stroke and a PFO, the PFO didn’t cause it in 80 percent of the cases,” said Dr. Spence. Among the best candidates for closure: people with larger holes, people whose strokes are associated with prolonged sitting, shortness of breath at the onset of a stroke, waking up with a stroke, or a history of deep-vein thrombosis. “I think most doctors aren’t aware of these factors,” he said.

The Gore REDUCE study was funded by W.L. Gore and Associates, the RESPECT study by St. Jude Medical, and the CLOSE study by the French Ministry of Health.