Medical science is evolutionary. We learn as we go and adjust our remedies as we better understand the problems we confront. Guidelines for medical practicioners and for consumers can help us make intelligent choices, but their credibilty can be jeopardized if they stray from recounting the scientific data into the minefield of promoting the policy preferences of the expert scientists who draft the guidelines.

We can and should preserve the credibilty of guidelines by injecting them with a healthy dose of humility.

Scientists have developed a system to grade the guidelines based on the quality of the evidence supporting the advice. The process is as valuable for clinical advice given in the doctor's office as it is for general population advisories such as the Dietary Guidelines for Americans.

The February 25 issue of the Journal of the American Medical Association contains an important article by Pierluigi Tricoci et al which grades the evidence underlying the joint cardiovascular practice guidelines of the American College of Cardiology and the American Heart Association. The results were important in their own right. Authors from Duke University and the University of North Carolina agreed that most of the guidelines are based on "lower levels of evidence or expert opinion" and disturbingly documented that "(t)he proportion of recommendations for which there is no conclusive evidence is also growing." Nearly half (48%) of the guidelines were based solely on expert opinion, not data. And, an icreasing percentage of guidelines are not based on "Class I" evidence but more on "Class II" which are supported by "conflicting evidence and/or a divergence of opinion about the usefulness/efficacy of a procedure or treatment." Fully 41% of the ACC/AHA guidelines are based on contentious Class II evidence.

The authors called for a more evidence-based approach.

We agree. The background discussion of the paper should be required reading for all expert groups preparing guidelines. After all, at least the ACC and AHA report that their recommendations are based on "Level of evidence C" -- expert opinion, the lowest level of evidence after "Level of evidence A: recommendations based on evidence from multiple randomized trials or meta-analyses" or "Level of evidence B: recommendations based on evidence from a single randomizd trial or nonrandomized studies."

In contrast, the process of preparing the Dietary Guidelines for Americans is purely "Level of evidence C: recommendation based on expert opinion, case studies, or standards of care." Sure, the experts select evidence of quality scientific studies, but the process is inverted: their cited literature follows the experts' conclusions (level C) rather than drive the process (levels A and B).

In an accompanying editorial , Terrence M. Shaneyfelt and Robert M. Centor opine:

Current use of the term guideline has strayed far from the original intent of the Institute of Medicine. Most current aticles called "guidelines" are actualy expert consensus reports. ...

The overreliance on expert opinin in guidelines is problematic. All guideline committees begin with implicit biases and values. However, bias may occur subconsciously and, therefore, go unrecognized. Converting data into recommendations requires subjective judgments; the value structure of the panel molds these judgments.

While many focus on direct financial conflicts of interest as the motivating bias of experts, Tricoci et al dig a bit deeper:

Recommendations based only on expert opinion may be prone to conflicts of interest becaue, just as clinical trialists have conflicts of interests, expert clinicians are also those who are likely to receive honoraria, speakers bureau, consulting fees, or research support....

The list could be much longer and include career and other professional advancement considerations.

The overall message is one of humility and restraint. Don't over-interpret the evidence and fit the evidence to desired policy options. That's what robs guidelines of credibilty and leaves practicioners and the general public whip-sawed with conflicting advice.

For guidelines on dietary salt, the lesson is that we need to adhere more faithfully to an evidence-based approach embodied in the approaches of the international Cochrane Collaboration and the U.S. Preventive Services Task Force, both of which have concluded there is insufficient evidence to justify a population guideline on salt intake levels. That's a "Class I, Level of evidence A" conclusion we should incorporate.

The article and editorial are timely for two reasons:

1. One of the authors, Dr. Robert Califf of Duke University, is reportedly a top contender to become the new Commissioner of Food and Drugs in the Obama Administration and
2. The Dietary Guidelines Advisory Committee has just begun a periodic reassessment of the Dietary Guidelines.

Our vote is for Dr. Califf to head FDA and for a new measure of data-driven humility in rendering dietary advice on Americans' eating habits, converting the Guidelines from expert opinion to a true evidence-based product.