Medical Device Daily National Editor

TORONTO – Evidence-based guidelines – a growing medical emphasis, for physicians to guide patient therapy and disease management, and for payers; the byproduct of randomized clinical trials and a growing mass of data; and, more often these days, a repetitive mantra considered so self-evidently useful and valid that they receive little critical debate.

But the unquestioned precision of evidence-based guidelines wasn't the point of a first-day satellite symposium at this year's scientific sessions of the Heart Failure Society of America (HFSA; Minneapolis), titled "Device Therapy for Heart Failure: Debating the Guidelines – Which Should We Really Follow?"

Introducing the session as moderator, Clyde Yancy, MD, of Dallas, suggested that the debate would focus on physician failure to use evidence-based guidelines consistently. And, echoing an earlier satellite symposium, he noted significant reductions in the use of ICD therapies for blacks, as well as for women, with the suggestion that this could be the result of some discriminatory biases in the practice of using the use the guidelines recommending these therapies.

But he then indicated an alternative interpretation: that there may be significant difficulties in interpreting and applying the guidelines.

This interpretation received the primary emphasis during the symposium presentations by three physicians, and the subsequent case study debates concerning whether to use these therapies on specific patients underlined the point.

And this observer heard a great many questions expressed concerning the evidence on which medical guidelines are based, how they are developed and the unquestioned ability of physicians to understand and interpret them.

Mariell Jessup, MD, of Philadelphia, the first presenter, suggested, for instance, the inconclusive nature of the initial version of American College of Cardiology/American Heart Association guidelines for ICD use, developed in 2005, based on the MADIT trial, and then the updated guidelines based on MADIT II.

Given the increased confidence concerning ICD use, and the resultant detailed guidelines, the initial, two-paragraph guidelines might have looked as providing little help to clinicians, and other presentations noted how often guidelines tend to lag behind new evidence.

JoAnn Lindenfeld, MD, of Denver, further reinforced the tentative and evolving nature of the evidence-based guidelines by describing the evolution in the development of HFSA's comprehensive guidelines for ICD use in 2006.

She said that these initial guidelines were developed by writers who first reviewed the basic literature, "wrote what we thought, submitted it to the executive committee, and then to the society at large" – then adding that this was "not a very systematic process."

Lindenfeld went on to note the subsequent controversies which arose, listing a variety of issues raised concerning the validity and usefulness of some sets of guidelines, and fostering much greater care in guideline development.

Besides the lack of timeliness, these were some other issues:

That they are too often "cumbersome" and thus difficult to use.

The presence of multiple, and differing, sets of guidelines for the same disease.

"Variable" rather than consistent formats.

Sometimes reflecting writer bias (Lindenfeld citing a set of pharmacologic guidelines for the use of statins, written by those with interests in this drug type).

Not making "explicit" the expected outcomes, such as a realistic look at mortality, exercise capacity and quality of life.

Expanding on this final point, she said that while evidence-based guidelines may provide guidance to the cardiologist, they leave a whole variety of questions unanswered for the patient, or for how the cardiologist communicates with the patient.

They don't provide much direction, she said, for helping the patient understand the "absolute risk and benefit" of a therapy. "The benefit may be greater, but how much greater? ... We need to tell our consumers what the timing of the benefit is – right away or take some time."

On the risk side, she noted that while ICDs have been shown to save lives – from seven to eight people out of 1,000 implanted – patients also should be told that "30 patients will die anyway with an ICD." So, guidelines could be improved, she said, by helping "all of us individualize" what they mean for patients – "how to apply them to my individual patient."

She expressed a ray of hope on these issues by saying," We are almost there." But she qualified this statement by quoting fitness guru Jack LeLanne, who at 93 said he felt good and "had sex almost every day" – underlining how relative and unspecific the word "almost" is.

Leslie Saxon, MD, of Los Angeles, however, suggested that evidence-based guidelines aren't likely to be structured as patient-friendly statements, following her careful review of a combined set of guidelines by the ACC/AHA and European Society of Cardiology. She excused her drill-down detailing of the statements by saying, "Every time I put an ICD in, I fill out a three-page form about why I did it and what my thinking was. We're not going to get paid unless this thing is filled out."

That, she said, "is the benefit of these guidelines and the detail is pretty clear to us."

The session was wrapped up with the presentation of case studies, with the discussion casting further doubt on the definitive ease of use in a disease with so many variables – dyssynchrony exactly in which ventricle? – and when dealing with human beings with lives and circumstances lying outside clinical parameters.

Considerable debate was raised by the case of a 56-year-old male hospitalized for "increasing dyspnea," after a year of heart failure symptoms, but responding "well" in the hospital to diuretics and characterized as "treated for a week, had something for a year."

Put in an ICD or not?

While the three presenters opted to treat him pharmacologically and hope for improvement, session moderator Yancy said he would implant, because "for this patient to get better he would have to [show] reverse modeling; he would have to have improvement in EF [ejection fraction] greater than 10."

But Jessup threw in a real-world variable from the experience of one of her own patients similar to the hypothetical case: Her patient worked on a tugboat, at a strenuous job. "With an ICD, he would lose his job – and probably his insurance."

The MDD reporter's conclusions from this debate: Evidence-based guidelines aren't exact. And they had better be complemented by superior physician judgment and excellent communication with the atient.