A Guide to CV Guidelines
Abstract & commentary
By Allan J. Wilke, MD, MA, Professor and Chair, Department of Integrative Medicine, Ross University (Bahamas) Limited, Freeport, Grand Bahama, The Bahamas. Dr. Wilke reports no financial relationship to this field of study.
Synopsis: Guidelines to assess cardiovascular risk are often inconsistent and poorly developed.
Source: Ferket BS, et al. Systematic review of guidelines on cardiovascular risk assessment: Which recommendations should clinicians follow for a cardiovascular health check? Arch Intern Med 2010;170:27-40.
Among our myriad functions, family and internal medicine physicians see adults for health maintenance exams. As part of these exams, we screen patients for risk factors to predict and, perhaps, prevent disease. The particulars of these screens are almost always dictated by guidelines issued by specialty societies or governmental agencies. By their nature, guidelines can vary in quality and, at times, can contradict each other. This can occur based on what evidence the authors used to construct the guidelines or by inherent bias. So, whose guidelines should you use and trust?
These investigators from the Netherlands and Boston focused on cardiovascular (CV) risk assessment guidelines. They looked at CV guidelines that could be applied to adults without established CV disease (CVD) or already receiving treatment for diabetes, hypertension, or hypercholesterolemia and were designed to prevent a first CVD event. They performed a systematic review of the English literature from 2003 to 2006, searching MEDLINE, CINAHL, the National Guideline Clearinghouse (USA), the National Library for Health on Guidelines Finder (United Kingdom), the Canadian Medical Association Infobase, and the G-I-N International Guideline Library. An astounding 1984 citations were identified. After appropriate exclusion, 114 were fully reviewed and another 87 were excluded (again, appropriately: not the most recent version, developed before 2003, not focused on screening, not an asymptomatic population, etc.), leaving 27 for formal review.
The authors applied the Rigor of Development domain of the Appraisal of Guidelines Research and Evaluation (AGREE) tool to each guideline to ascertain the quality of the guideline's development. AGREE considers conflicts of interest and the reporting of the evidence search methods, evidence selection criteria, recommendations formation methods, benefits and risks, supporting evidence, external expert review, and process for updating. A single guideline might have multiple recommendations. Each guideline received a percent rigor score with 50% as the cutoff for a well developed guideline. Of the 27 guidelines, 17 received scores 3 50%.
The groups issuing at least one high-scoring guideline included: the National Institute for Health and Clinical Experience (NICE, UK), the Scottish Intercollegiate Guidelines Network (SIGN), the American Heart Association (AHA), the American Stroke Association (ASA), the New Zealand Guidelines Group (NZGG), the World Health Organization (WHO), the European Society of Cardiology (ESC), the Canadian Cardiovascular Society (CCS), the U.S. Preventive Services Task Force (USPSTF), the National Heart, Lung, and Blood Institute (NHLBI, USA), National Health and Medical Research Council (NHMRC, Australia), European Association for the Study of Diabetes (EASD), the Canadian Diabetes Association (CDA), the Canadian Task Force on Preventive Health Care (CTF), and the American Association of Clinical Endocrinologists (AACE). All of the USPSTF guidelines scored 3 95%. Not every group fared as well. For instance, the AHA's guideline on total cardiovascular risk scored 76%, but its dyslipidemia and dysglycemia guidelines (issued jointly with the American Cancer Association and the American Diabetes Association) scored only 14%.
Among the 27 guidelines, there were areas of disagreement. For instance, the USPSTF found the evidence insufficient to recommend screening for dysglycemia. The NHMRC recommended screening individuals 3 45 years with a body mass index (BMI) 3 30 kg/m2, but found insufficient evidence to screen individuals 3 55 years without another risk factor. Furthermore, there were inconsistencies across guidelines regarding the use of aspirin, statins, and hypertensive agents in patients with diabetes.
Commentary
In primary care, we want to practice evidence-based medicine, but we do not have the time to review the literature and develop our own strategies for application. We depend on expert colleagues to do this. CVD is the No. 1 killer in the Western world. To learn that CV guidelines (even well-constructed ones) are not consistent is discouraging. I have relied on the USPSTF to "guide" me, and this has worked well for me since generally they are conservative and usually noncontroversial (notwithstanding the recent update of mammography guidelines1). Some variation in recommendations is inevitable since, even among relatively well-off countries, there are differences in populations and what is expected of their health care systems. Certainly, depending on what populations are studied and what studies are used to develop a particular guideline, we could end up with incongruent recommendations, but we don't expect major differences.
In an invited commentary, Smith writes, "Yet concerns remain about the process by which guidelines might be developed, and a lack of congruence among their recommendations is one reason for low implementation by health care providers."2 I want to emphasize that the authors of this review of CV guidelines did not look at the evidence that was used to develop the guidelines. They looked at how they were developed. That said, I think it's important that guidelines be developed in accordance with the AGREE tool, especially the disclosure of conflicts of interest. Smith concludes, "Further progress will depend on our ability to develop guidelines that are free from external bias, an agreement on a definition for global cardiovascular risk that is broadly applicable, and the identification of target populations in terms that take into account important regional variations in risk and health care delivery." To me, it's a matter of trust.
References
1. "New Breast Screening Limits Face Reversal." Wall Street Journal Jan. 12, 2010. Available at: http://online.wsj.com/article/SB126325763413725559.html.
2. Smith SC Jr. Comment on "Systematic review of guidelines on cardiovascular risk assessment: Which recommendations should clinicians follow for a cardiovascular health check?" Arch Intern Med 2010;170:40-42.