Monday, January 6, 2014

Guidelines, Complexity and Headlines

Guidelines are a way to systematize medical knowledge in order to make good medical care more accessible.  A single doctor knowing the right way to manage a person’s specific problem is wonderful, but scaling that single interaction into the correct care for everyone is a challenge that guidelines are meant to address.  But healthcare providers must also correctly apply those guidelines in a unique manner for the specific person seeking care.  That is actually a very difficult dilemma – one that makes the communication and use of these guidelines prone to errors. For a guideline to be useful, it must support the humanism that good physicians and other health professionals bring to the table, not substitute a rigid single approach that may actually impede care for those in need.

Every time a new guideline is published in a medical journal, the news media feels compelled to share that expert advice.  I admit to some frustration, not by the effort to make this information available to the public (it should be), but by the compact messaging of a headline that results in misleading information.  I worry that this end result implicitly supports a rigid approach and leaves out the complexity and the unique application that should be an integral part of the way any guidelines are applied. 

A December 18, 2013 USA Today headline read: “Advice would put fewer Americans on blood pressure meds.”  The article highlighted the new guideline developed by the 8th Joint National Committee (JNC8) on the management of high blood pressure.  The news article made it appear the experts were recommending decreasing the use of hypertension medication.  It simplified extremely complex work by an expert group and may have unknowingly misled the public about the recommendations’ true nature. 

Guidelines like this rate the recommendations within the guidelines on the basis of how strong the evidence is that they are likely to produce benefit.  In general, while a number of groups produce guidelines, the ratings systems all tend to be similar. In general, their recommendations are letter-graded, ranging from “A” (which is a “strong recommendation,” with “high certainty based on evidence”) to “D” (which is a “recommendation against”) or “E” (“expert opinion without evidence)

The grade a particular recommendation receives is based on a consensus methodology and is more art than science. 

What the news reports don’t show is the strength of each recommendation within the guideline.  So, with the nine recommendations in the high blood pressure guideline, only 1½ are A (“strong”) recommendations.  Many of the other recommendations within the JNC8 hypertension guideline are rated “E,” which means that there is no evidence for the recommendation -- that it is only based on the opinions of the experts on the panel.


The new USPSTF guideline was, in fact, given only a “B” recommendation -- one that is only of moderate benefit based on the evidence.  The guideline recommends that adults age 55 to 80 who have a 30-pack-year smoking history and currently smoke or have quit within the past 15 years receive an annual lung cancer screening with a low-dose computerized tomography (CT) scan.  ”Pack years” are determined by the number of packs per day a person smokes, multiplied by the number of years they have been smoking.  Thus a person who smokes one pack a day for thirty years is a 30 pack year smoker and so is a person who smokes two packs a day for fifteen years.
What the Wall Street Journal also does not mention is that two accompanying editorials in the Annals of Internal Medicine -- where the new guideline was published -- implicitly challenge the “B” grade as being too strong based on the evidence reviewed.  Why? It turns out that 95% of all positive CT results do not lead to a cancer diagnosis. The committee may have minimized the risks of false positives in setting the recommendation as a “B.”  They also assumed only the right people would be screened.  As the authors of one of the editorials stated, “How patient selection actually occurs is worth careful consideration, because ample evidence shows underuse of cancer screening in populations for which it is indicated and overuse in those for which it is not.”  Thus the headlines, which attempt to inform the public about new thoughts on high blood pressure and screening for lung cancer, may actually confuse rather than inform. 
All this both gives me hope and makes me nervous.  I am heartened by the fact that guidelines are getting more complex and more specific to sub-populations of patients, and more often based on strong evidence rather than solely on opinion.  At the same time, I am nervous about their implementation. They are becoming harder and harder to communicate to those who would most benefit – and more difficult for physicians and nurses to use in daily practice.  No doctor can remember all the countless recommendations within those hundreds to thousands of guidelines. 
To properly implement guidelines and their recommendations, we need to use smart electronic medical record systems --and I contend those types of systems do not yet exist.   Healthcare professionals need to come to an agreement as to how they’ll use recommendations that have less than “A” grades.  And ultimately, we have to find a way to use these the guidelines effectively and uniquely in a 10-or 15-minute office visit in which we may need to use multiple guidelines with multiple recommendations. 

My headlines for stories about new guidelines would differ from those we see in USA Today and the Wall Street Journal. In fact, I could sum up most news stories with a single headline: “Evidence-Based Guidelines Getting More Complicated, Harder to Use and Understand.”  This means we all need caring, thoughtful health professionals to intelligently use guidelines in the context of a single person’s life more than ever. We also need the right technology to facilitate their use.  And finally, we need better public education on what guidelines really do mean, and not only the simple headlines that hide complex facts.