Physician and patient preferences for stroke prophylaxis in atrial fibrillation

Purpose. To derive and compare relative preferences of physicians and patients for selecting oral antithrombotics in atrial fibrillation (AF). Methods. Elicitation task: Time trade-off (TTO) and best worst scaling (BWS) choice experiments were constructed from literature review and expert opinion, reflecting four attributes relevant to oral antithrombotic selection in the setting of stroke prevention in AF – frequency of laboratory monitoring, annual risk of stroke, annual risk of major bleed, availability of reversal agent. Analysis. Utilities based on the patient TTO data were derived and analyzed for subgroup differences. Relative utilities based on the BWS choice data were derived using the conditional logit model and latent class analysis. The Wilcoxon signed-rank test was performed to assess the difference in preference for each attribute level between the Best-Worst score for the two groups. Results. 33 physicians and 58 patients completed the choice experiment. Both groups favoured “annual stroke risk of 0%” as the most valued attribute-level, and “annual stroke risk of 10%” was the least favourable attribute-level. Significant preference differences between the two perspectives for several of the attribute levels were also found. The results points out that while both groups had strong positive preferences for “annual stroke 0%”, the physicians valued this attribute level to be much more desirable than the patients. Similar observation applies to “annual stroke 10%”, where physicians had much stronger negative preferences for the attribute level compared to the patients as well as the range of annual bleeding rates, suggesting that physicians in general, were more stroke and bleed averse than the patients. Conclusions. There is a general congruence in physician and patient preferences for stroke prophylaxis, however, the strength of preferences differ for several attributes differ between the two groups. Using a BWS choice experiment, a preference elicitation method based on the random utility theory, this is the first study that quantitatively evaluated the preferences of physicians and patients for stroke prophylaxis in atrial fibrillation and provides important insights into clinical decision-making in a patient-centered care model.