A new study has found a majority of medical devices recalled because they harmed patients were part of a class of devices that were not tested in humans.
Researchers reviewed devices the FDA recalled from 2005 through 2009, such as defibrillators and brain shunts. The study was published Monday in the Archives of Internal Medicine.
Study Co-author Dr. Steven Nissen of the Cleveland Clinic said of the 113 highest-risk devices, 80 were approved by what he calls a shortcut process. And Nissen said that in 2006, more than 2,800 people died because of faulty devices.
"At least half of those deaths were due to devices that had been approved without any clinical testing," Nissen said. "That is without any testing in patients."
Nissen said the medical device approval process should be far more rigorous.
But AdvaMed -- a medical device trade group -- said the study is flawed in part because the number of devices it cites is a tiny fraction of those sold.
Ralph Hall, a medical device regulation attorney, said it's not possible to test pacemakers and artificial hips the same way new drugs are tested.
"The vast majority for the reasons for the recalls are issues for which a human clinical study could not have found it or is certainly not the best way to find it -- a manufacturing mistake, the wrong label was put on the product, for example," Hall said.
Hall also said the study is misleading because it focused on a tiny fraction of devices sold.