Individual effects for well-informed shared decision making for atrial fibrillation thromboembolic prophylaxis: WISDM for A FIB




Poster session 3 Friday: Evidence Tools / Evidence synthesis - creation, publication and updating in the digital age


Friday 15 September 2017 - 12:30 to 14:00


All authors in correct order:

Alper B1, Oettgen P1, Durand M2, Blaine A3, Erkmann J1, Malone-Moses M1, Elwyn G2
1 EBSCO Health DynaMed Plus, USA
2 Dartmouth College, USA
3 EBSCO Health Shared Decision Making, USA
Presenting author and contact person

Presenting author:

Brian Alper

Contact person:

Abstract text
Background: Options for embolic stroke prevention in atrial fibrillation include anticoagulants (apixaban, dabigatran, edoxaban, rivaroxaban, warfarin), antiplatelet agents (aspirin with or without clopidogrel), and left atrial appendage closure. Comparisons of options have limited accuracy for individual decisions due to selective use of evidence for relative-risk estimates, not applying relative-risk estimates to individual risk, or misalignment of outcomes used for relative risk and individual risk estimates. Accurate individualised estimates need to be communicated clearly for clinicians and patients to make informed decisions.

Objectives: We systematically determined the evidence needed for informed shared decision making and produced a tool to make it easy to use.

Methods: We used DynaMed systematic literature surveillance to identify meta-analyses and randomised trials for eight options. For aspirin and warfarin, we extracted relative-risk estimates compared to no treatment for ischemic stroke and for major bleeding. For clopidogrel plus aspirin, we extracted relative risk estimates compared to aspirin. For all other options, we extracted relative risk estimates compared to warfarin. We selected CHA2DS2-Vasc and HAS-BLED scores as the most accurate predictors for individual risks for embolic stroke and major bleeding, and developed an interactive form to view an individual’s estimated annual risk of embolic stroke and major bleeding with selected treatment options. We used Option Grid methods to present the results for patient use for shared decision-making support.

Results: shows an online tool providing clinician-facing and patient-facing information. The tool includes relative risks of ischemic stroke and major bleeding with each option, absolute risks and number needed to treat or harm for clinicians, and numbers per 1000 people for patients.

Conclusions: Use of WISDM for A FIB can provide accurate, individualised estimation of benefits (in terms of embolic stroke prevention), harms (in terms of major bleeding and other complications), and burdens (descriptions of use of the treatment) to facilitate SDM.