Clinical Practice Guidelines and Scientific Evidence

To the Editor: Dr Tricoci and colleagues¹ published an analysis indicating that American College of Cardiology/American Heart Association (ACC/AHA) guidelines are largely developed from lower levels of evidence or expert opinion. They concluded that there is a “need to improve the process of writing guidelines and to expand the evidence base from which clinical practice guidelines are derived.” Their first conclusion could have been better substantiated by a critical appraisal of the ACC/AHA guideline methods.²

First, the ACC/AHA does not appear to follow Institute of Medicine recommendations to separate the systematic review process from guideline formulation.³ Rather, the same writers appear to perform both processes.

Second, the ACC/AHA uses an overly simplistic, outdated hierarchy of study design (randomized controlled trials [RCTs] at the top) to assess level of evidence. Different types of clinical questions are best answered with different study designs. Moreover, in several recent ACC/AHA guidelines I could find no mention of quality assessment of the trials included in the body of evidence.

Third, the ACC/AHA does not document the systematic review underlying each recommendation using standard reporting criteria.⁴ It appears that citations supporting each recommendation may be selected by guideline authors.

Fourth, class I recommendations are those where “ . . . there is evidence and/or general agreement that [an intervention] is useful and effective.” If the expert panel agrees, then supporting evidence is not required to recommend an intervention, inconsistent with an evidence-based process.

Fifth, a class IIa recommendation is based on conflicting evidence or on divergent expert opinions, yet guideline writers can use phrases such as “is reasonable” or “is probably recommended.”

Sixth, I find no explicit process for translating evidence into recommendations, other than level of evidence (ie, study design) and the opinion of experts. It is important to understand what other factors are considered and how the usefulness of an intervention is determined.