Contact

Duke Health News

919-660-1306

Despite billions of dollars and decades of research, only a minority of recommendations for cardiology care in the United States rest upon scientific evidence from large, randomized clinical trials, the gold standard of research, according to investigators in the Duke Clinical Research Institute (DCRI) and the University of North Carolina at Chapel Hill.

The finding stems from a systematic review of the quantity and quality of scientific evidence underlying clinical practice guidelines endorsed by the American College of Cardiology and the American Heart Association and is published in the Feb. 25 issue of the Journal of the American Medical Association.

The finding may be especially surprising because cardiology has a reputation of being one of the disciplines most firmly grounded in evidence.  Researchers say the study reflects a critical need to expand research funding and to create a national research agenda that could fill in major gaps in knowledge about some of the most fundamental aspects of heart care.

"Most of the clinical trials in this country are sponsored by business interests trying to bring a new drug to the marketplace," says Pierluigi Tricoci, MD, a cardiologist at DCRI and the lead author of the study. "The fact of the matter is that there are many questions in cardiology care that still need to be answered, and the only way to answer them properly is through large, randomized trials. Those trials cost a lot of money, however, and there aren't many funding sources interested in supporting initiatives where a new product is not involved."

Tricoci and a team of researchers reviewed the evolution and growth of ACC/AHA guidelines generated between 1984 and 2008.  Practice guidelines provide a framework for cardiology care and are designed to help physicians in their day-to-day decisions regarding treatment options.

Guidelines are rated according to the level of the scientific evidence supporting them and the class of the recommendation. The level of evidence (A, B, or C) refers to the type and strength of the studies supporting the recommendation: Level A reflects recommendations arising from multiple randomized trials; Level B refers to evidence from non-randomized trials or just one randomized trial, and level C are recommendations with little or no scientific evidence.  The class of the recommendation (I, II, or III) refers to the strength of the endorsement the evidence elicits, and is generally a more subjective evaluation based upon such factors as relative risks and benefits and degree of conflicting data.

In examining the entire body of guidelines over the 24-year period, researchers reviewed 53 guidelines on 22 topics with 7196 recommendations. They found that only 11 percent of current recommendations (314 of 2711) carry level A evidence, while 48 percent carry level C evidence.

Focusing solely on 12 current guidelines where there had been at least one revision, they found that the number of recommendations had increased over time by 48 percent (from 1330 to 1973) with a progressive shift toward more class II recommendations, where there are lower levels of evidence and expert opinion, often due to conflicting data.

"Most current guideline recommendations are emerging from weaker evidence. There is also a growing proportion of recommendations subject to uncertainties, says Tricoci.  "In short, we are generating a lot more information but very little definitive knowledge."

Sidney Smith, MD, Professor of Medicine at UNC, former chair of the ACC/AHA Guidelines Task Force and senior author of the study, says "the ACC/AHA guidelines have an established record of excellence in improving outcomes for patients with cardiovascular disease."  Still, he says, there is a growing number of clinical circumstances where more information from randomized clinical trials is needed.

"It is possible that many of these evidence gaps could be covered by good use of money dedicated to comparative effectiveness studies in the newly-approved American Recovery and Reinvestment Act," says Rob Califf, MD, director of the Duke Translational Medicine Institute and a co-author of the paper.

Smith agrees, but adds that "if not, it will be important to identify alternative funding sources because cardiovascular diseases remain the leading cause of death and disability in our country."

The study was funded by the Agency for Healthcare Research and Quality and a grant from the Azus Fund, Sharp HealthCare Foundation. 

Additional co-authors include Judith Kramer, MD, Duke Center for Education and Research on Therapeutics; Rob Califf, MD, director of the Duke Translational Medicine Institute; and Joseph Allen, from the ACC's Cardiology and Science Division.