CANADIAN ATRIAL-FIBRILLATION ANTICOAGULATION STUDY - WERE THE PATIENTS SUBSEQUENTLY TREATED WITH WARFARIN

Objective: To determine the effect of the results of clinical trials o n the behaviour of patients and physicians, the authors ascertained th e proportion of patients participating in the Canadian Atrial Fibrilla tion Anticoagulation (CAFA) study who started or continued warfarin th erapy at the end of the study and identified factors affecting the dec ision to use or not use warfarin. The CAFA study was a double-blind, r andomized, placebo-controlled, multicentre study to evaluate the effic acy of warfarin in preventing stroke among patients with nonrheumatic atrial fibrillation. Recruitment and follow-up were stopped early beca use two other similar studies had shown a decrease in the rate of stro ke among patients treated with warfarin. Design: Mail survey 21 months after the end of the study. Participants: The personal physicians of 336 patients who had participated in the CAFA study. Outcome measures: Type of antithrombotic therapy the patients had received since the CA FA study ended, for patients who were not receiving warfarin, the reas ons they were not. Results: Questionnaires concerning 254 (76%) of the patients who had participated in the study were returned. Since the e nd of the CAFA study, 153 (60%) of these patients had been treated con tinually with warfarin, 14 (6%) had been treated with warfarin but had subsequently stopped taking it, 59 (23%) had taken acetylsalicylic ac id (ASA) continually, 5 (2%) had been taking ASA but had subsequently stopped taking it, and 23 (9%) had not taken either drug. The respondi ng physicians stated that 58 (67%) of the patients who were not treate d with warfarin did not wish to take the drug. The patients who had re ceived warfarin during the CAFA trial were more likely to be treated w ith warfarin after the trial (75%) than were those who had received a placebo (56%) (p = 0.001). The probability of the patients' being trea ted with warfarin also depended on which study centre they had been tr eated in (p = 0.001). Conclusions: Of the patients in the CAFA study f or whom questionnaires were received, only 167 (66%) had been treated with warfarin after the end of the study. The patients were more likel y to have been treated with warfarin after the study if they had recei ved warfarin during the study. The positive results of clinical trials , on their own, are not enough to fully change the behaviour of patien ts and physicians.