These large databases are very hard to pull out reliable data from

"There is a very important story to be told about unnecessary carotid surgery and stenting in the U.S.," Dr. David Spence, a stroke prevention researcher at the University of Western Ontario in London, Canada, told Reuters Health last week.

According to his research, 90 percent of symptom-free patients who get surgery would be better off without it, he said in an email.

Dr. Stanley L. Barnwell, the medical director of intervention neuroradiology at Oregon Health and Science University in Portland, said he's less convinced by the results of the new study than by earlier research showing stenting is just as safe as surgery.

In an earlier experiment called the CREST trial, patients were randomly assigned to receive either of the procedures to clear up the plaque in their arteries.

This type of project set out in advance to control for differences between the patients and to carefully monitor their outcomes, whereas the current study is a look back on medical records from patients who were not necessarily part of a study.

"These large databases are very hard to pull out reliable data from," Barnwell, who was not involved in either study, told Reuters Health.

Kallmes said the studies complement each other -- his study had more numbers and reflected real world practice, while the earlier trial was carefully controlled but limited in the number of patients involved.

"Maybe they give us bookends of the real rate of hemorrhage [after these procedures], and the truth lies somewhere in between," Kallmes said.

In his study, published in the journal Stroke, Kallmes and his colleagues call for more research to explain why they saw such a big difference between the two procedures.