(AP) -- One in five heart defibrillators may be implanted for questionable reasons without solid evidence that the devices will help, according to a first-of-its-kind analysis.
Implanted defibrillators shock the heart back into a normal rhythm when it starts beating irregularly. They can prevent sudden death in people with advanced heart failure, but researchers haven't found a benefit for other patients.
Patients who've had a recent heart attack or recent bypass surgery aren't good candidates for defibrillators, for example. Guidelines don't recommend them for people newly diagnosed with heart failure either and those so sick that they have very limited life expectancies won't be helped.
But in the new study, which examined nearly four years of national data, 22 percent of the implant surgeries were in patients who fit one of those categories.
Some may have been appropriate, said lead author Dr. Sana Al-Khatib of Duke University School of Medicine in Durham, N.C., but it's likely that many were done despite the research evidence.
"It's lack of knowledge. It's ignorance. It's not keeping track of the guidelines," she said. "And we may have some physicians who don't agree with the guidelines or don't think the guidelines apply to their patients."
The study, appearing in Wednesday's Journal of the American Medical Association, examined national registry data from nearly 112,000 patients in 2006-2009.
The researchers found that the patients who got implants according to guidelines were less likely to die in the hospital and suffer complications than the patients whose surgeries clearly fell outside the guidelines.
The surgeries cost thousands of dollars, raising questions about wasted resources, Al-Khatib said. "It's all about improving the quality of care," she said.
The 22 percent rate is disturbingly high, said Dr. Douglas Zipes of the Indiana University School of Medicine in Indianapolis. He wasn't involved in the study but has helped write research-based guidelines for the devices.
"If the patient is not benefiting, we shouldn't be doing it," Zipes said. He noted that doctors with special training in heart rhythms, had slightly better rates of evidence-based implants than did other types of doctors in the study.
Doctors should follow guidelines, he said, and patients should choose a specialist with training in heart rhythm problems.