This Article Full Text (PDF) All Versions of this Article: 0272989X07302166v1 27/4/438    most recent References Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Thomson, R. Articles by Chinn, D. J. Search for Related Content PubMed PubMed Citation Articles by Thomson, R. Articles by Chinn, D. J. Social Bookmarking                   What's this?

A Cautionary Note on Data Sources for Evidence-Based Clinical Decisions: Warfarin and Stroke Prevention

Public Health Research Group, School of Population and Health Sciences, richard.thomson{at}newcastle.ac.uk

Martin Eccles, MD

Centre for Health Services Research University of Newcastle upon Tyne, United Kingdom

Ruth Wood, MSc

Public Health Research Group, School of Population and Health Sciences

David J. Chinn, PhD

Public Health Research Group, School of Population and Health Sciences

Background. Stroke risk in nonvalvular atrial fibrillation can be reduced by warfarin or aspirin; the choice of therapy requires the assessment of risks and benefits. The authors compared methods of risk assessment and their implications for risk communication and treatment. Methods. Stroke risk was compared in 193 patients with atrial fibrillation using the Framingham equation; an atrial fibrillation—specific Framingham equation; the Congestive heart failure, Hypertension, Age, Diabetes and Stroke (CHADS₂) score; the Stroke Prevention and Atrial Fibrillation (SPAF) scheme; and the Scottish Intercollegiate Guidelines Network (SIGN) guidelines. Treatment guidance from SIGN, a simple prediction rule, and a decision analytical approach was compared. In the latter, patients were classified as risk too low to benefit from warfarin if the risk of cerebral bleeding on warfarin approximated to, or exceeded, thromboembolic stroke risk reduction. Results. Framingham equations gave lower stroke risks overall than SIGN or SPAF. CHADS₂ was intermediate. Using SIGN, warfarin would be given to all 103 patients without a history of stroke/transient ischemic attack and for whom warfarin was not contraindicated but only to 73 patients using the simple prediction rule and 48 patients using the decision analysis. Conclusion. Community-based cohorts give lower stroke risk estimates than CHADS₂; both give lower estimates than schemes from control groups from randomized controlled trials. Using community-derived risks would lead to fewer patients being treated with warfarin than guidance derived from randomized controlled trial controls, which may lead to many low-risk patients being treated with high-risk therapy. This raises the debate about appropriate sources of data for risk assessment to support risk communication and effective clinical decisions.

Key Words: Key words: atrial fibrillation • stroke • risk assessment • decision making • patient preferences. (Med Decis Making 2007:27:438—447)

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?