Tammy Bungard and colleagues1 correctly identify patient-, physician-, and health care system–related barriers to warfarin use in patients with atrial fibrillation. However, they do not identify patients' preferences as an important element in decision making for treatment of this condition.
Protheroe et al2 have recently completed a study examining patients' preferences for warfarin in an elderly population (aged 70 to 85 years) with atrial fibrillation. Patients with atrial fibrillation (N = 260) were selected from 8 general practices in Bristol, England; 97 of these patients (50% of those eligible) had no contraindications to warfarin, no other concomitant medical problems, and were able to express their preferences as measured by the time trade-off method. Just under half (n = 48; 49%) were already taking warfarin. When their individual probability of suffering a stroke was combined with their utility scores by means of decision analysis, 59 (61%) would accept warfarin therapy for their atrial fibrillation. This contrasts with the proportion deemed to benefit from treatment if treatment was offered solely on the basis of absolute risk (n = 70, 72%) and consensus conference guidelines (n = 89; 95%).3,4 Furthermore, the chance-corrected measure of agreement (κ statistic) for treatment preferences based on individualized decision analysis compared with absolute risk and consensus recommendations was 0.25 and 0.09, respectively, indicating "poor" agreement. Of those patients already taking warfarin, 17 (35%) would not take warfarin if their preferences were taken into account by means of decision analysis, while 28 patients (57%) who were not taking warfarin would accept treatment according to their decision analysis.
Fahey T, Montgomery AA, Peters TJ. Warfarin Therapy for Atrial Fibrillation: The Patient's Preference Is Important. Arch Intern Med. 2000;160(15):2393. doi: