The evidence for those recommendations was graded A to C. A recommendation got an A if its evidence came from multiple controlled trials, a B if evidence came from one randomized trial or non-randomized trials, and a C if the recommendation was based on clinical judgment, with little or no scientific evidence.
Just 12 percent of current recommendations -- 314 of the 7,196 -- got a grade of A, and 45 percent earned a C.
Those grades have been dropping in recent years, the study found. Though the number of recommendations increased by 48 percent over the 24 years covered by the study, there has been a steady shift toward lower levels of supporting evidence.
The lack of evidence is greatest in treating disorders of the heart valves, Tricoci said. "This is the area with the highest number of level C recommendations," he said.
The findings do not mean that cardiologists should ignore the guidelines, Tricoci said. "What the paper says is that cardiologists in daily practice face several situations where there is not scientific evidence and not a consensus about the best approach," he said. "This quantifies the gap in knowledge that I'm sure every cardiologist is familiar with every day."
One example of the uncertainty they face is the recommendation for prescribing the clot-preventing drug Plavix for someone who has had a stent implanted after an artery-opening procedure such as angioplasty, Tricoci said.
"That is uncertain even in the guideline," he said. "It says, 'for at least a year,' but there is no trial that shows how long Plavix should be given."
"There are lots of issues raised by this discussion," Gardner said. "To us, it is a validation of the usefulness of guidelines, plus pointing out that we don't have as many multi-center randomized trials as we would like to have."
But existing guidelines, Gardner said, are basically the best that can be set at this time
All rights reserved