CardioExchange Archive

With the recent release of the 2014 AHA/ACC Guideline for the Management of Patients with Valvular Heart Disease, CardioExchange editors were interested to know how, in the present climate of rapid change and controversy, the writing committee achieved consensus on the future treatment of valvular disease. We asked committee co-chairs Rick Nishimura, Professor of Medicine at the Mayo Clinic College of Medicine, and Catherine Otto, Professor of Medicine at the University of Washington School of Medicine, to discuss how the guideline was updated. 

Our basic approach for updating the guideline was to start with the evidence, basing recommendations on the published body of knowledge for each clinical issue.  First, we made a detailed outline for the scope of the document, including each valve lesion as well as prosthetic valves, endocarditis, noncardiac surgery, and pregnancy in women with valvular heart disease.

Within this outline, a parallel structured format was used with sub-headers for each type of valve disease for diagnosis and testing, medical therapy, and the timing and type of intervention. This outline was refined and expanded early in the writing process, but the parallel structure was rigorously maintained to ensure that recommendations were complete and consistent throughout the document.

Next, committee members were assigned as primary or secondary writers for each section, ensuring the absence of potential conflicts of interest for that specific topic.

The first responsibility of the primary writer was to:

• Identify the clinical issues we needed to answer for each section.

• Perform a systemic review of the published evidence with preparation of detailed evidence tables (which are published as an online supplement to the guidelines).

• Draft initial recommendations based on the evidence.

• List the references justifying a Level of Evidence A or B for each recommendation. Our goal was to provide recommendations with a Level of Evidence A or B; if there was no evidence to support a recommendation, we did not write one. This approach decreased the number of Class C recommendations from over 70% in the 2006 guidelines to less than 50% in the 2014 document — still not optimal, but better than before given the paucity of randomized controlled clinical trials for valve disease.

The committee was split into two groups under the leadership of the co-chairs to facilitate discussion and consensus building for the draft recommendations. Once each group achieved consensus, the entire committee discussed each recommendation again.

Consensus was achieved by writing recommendations that accurately reflected the published evidence, were congruent with the clinical expertise of the committee, and were worded precisely and unambiguously. Although we often had to rely on observational or nonrandomized data, we based definitions of disease severity and criteria for intervention on published study entry criteria and clinical outcome data whenever these data were available.  All of the committee members were willing to modify opinions based on their personal clinical experience when the published data were convincing.

The next steps for each primary writer were to:

• Write concise text for each recommendation, referring to the evidence tables and expanding on details as needed for clinical implementation.

• Add key references to each text section.

• Prepare flow chart diagrams for each major valve lesion to ensure the recommendations were internally consistent.

The text was reviewed by the entire committee with discussion, revision, and re-review as needed to ensure that all agreed with the final version.

After the committee members voted on all the recommendations, the entire document was sent for external review.  Reviewers were selected both by the organizations collaborating on the document and from suggestions by the writing group. We received a total of 1463 comments from 49 reviewers — we responded to each of these comments before sending the document to the American College of Cardiology Board of Trustees (BOT). After the very rigorous review from the reviewers and the BOT, any resulting changes in the Class Recommendations went back to the entire committee for approval. We hope this process of iterative internal review and extensive external review will ensure rapid and widespread acceptance of the guideline recommendations.

The modular format of these new valve disease guidelines will improve quick access to the specific information needed for patient care, particularly when these guidelines are available in an interactive digital format. In addition, this format allowed us to make rapid changes, even at the end of the review process; individual recommendations can be updated in the future, without rewriting the entire document, as justified by new published data.

We look forward to seeing these guidelines become a “living” evidence-based document with continuous updates to reflect the rapidly changing knowledge base in valvular heart disease.