March 25th, 2019
Clinical practice guidelines are an essential tool for clinicians. Health professionals rely on guidelines as “cheat sheets” to get treatment recommendations from the latest research. Unfortunately, as we have pointed out, guidelines are often affected by flawed research and conflicts of interest, and can even be hard to find.
In two pieces published recently in the Journal of the American Medical Association, researchers have identified important issues in cardiology guidelines, including the lack of clarity in summary statements and a worsening evidence base.
A team of cardiologists and researchers from Duke University, including former FDA commissioner Robert Califf, conducted a systematic review of more than 50 current cardiology guidelines, an update to a previous guideline review in 2009. The team analyzed each guideline recommendation and categorized them based on the level of evidence that supported the recommendation.
Disappointingly, the level of evidence in cardiology guidelines has not gotten better over the past ten years; in fact, it has declined. While 11.5% of recommendations in 2008 were supported by evidence from multiple randomized controlled trials, only 8.5% of recommendations in 2018 were classified as having this level of supportive evidence. About half of recommendations in 2018 were supported by a single RCT or observational evidence, and 41.5% were supported by expert opinion only.
This is important because we’ve seen many procedures and drugs accepted as beneficial and adopted into standard of practice without strong evidence, and later find that they weren’t beneficial at all. For example, the 2017 ORBITA trial showed that stenting for stable angina, a long-held practice to relieve symptoms, did not actually reduce patients’ pain better than a sham procedure.
The lack of evidence in cardiology guidelines could be a timing issue; randomized controlled trials generally take many years, perhaps we just don’t have the data yet. Another contributor could be that there are few incentives to conduct randomized trials of procedures and drugs once they have been adopted as the norm. Why would the companies that manufacture these products fund a trial that could show they aren’t effective?
As Dr. Bernard Lown said about why procedures like stents persist without evidence: “It’s hard to influence a social enterprise like health when powerful economic interests are pushing in the opposite direction.” The Duke study should make us reconsider not only the evidence behind our current standard of care, but also our future research priorities.
In a JAMA editorial, family medicine physician Dr. Neil Skolnik takes issue with the 2018 American College of Cardiology (ACC)/American Heart Association (AHA) guideline for managing cholesterol. As Skolnik points out, there is a key discrepancy in the language of the guideline that could lead to overtreatment. In the body of the guideline, the authors write that there is not strong evidence showing that patients older than 75 would benefit from taking a statin if they haven’t previously had a heart attack. One would think the authors, on the basis of this lack of evidence, would not recommend older adults taking a statin, given the known risk of side effects.
However, in the summary section of the guideline (the part that doctors are most likely to read) the guideline reads, “In adults 75 years of age or older with an LDL-C level of 70 to 189 mg/dL, initiating a moderate-intensity statin may be reasonable.” “It may be reasonable to” and “it may be reasonable not to” mean essentially the same thing, writes Skolnik, which leaves the real meaning of the guideline summary ambiguous. Almost all adults over 75 have this level of LDL cholesterol, Skolnik points out. Should doctors be taking this guideline as a recommendation to give statins to all healthy older adults? The unclear wording of this summary could lead to high levels of overtreatment, if doctors aren’t careful to look deeper into the actual evidence presented in the guideline.