Blinded End Point Adjudication in the ‘Danish Multicenter Randomized Study on Fibrinolytic Therapy versus Acute Coronary Angioplasty in Acute Myocardial Infarction’ (DANAMI-2 Trial)

Abstract

Background: Allocation concealment in trials that compare interventional procedures with medical therapy is impossible. Bias introduced by the investigators when end points are adjudicated is a methodological problem in such trials. Objective: To develop and test a novel method of blinded end point adjudication in open trials. Method: In the DANAMI-2 trial, a randomised open trial comparing primary angioplasty to fibrinolytic treatment of acute myocardial infarction, end points were evaluated by an independent end point committee (EPC). The cardiologists who presented the clinical end point data to the EPC were asked not to disclose any information on the randomised treatment, in order for the EPC to pass a truly blinded end point adjudication. The EPC categorised their evaluation of each end point as evaluated blinded or evaluated not blinded. Results: Blinded end point adjudication was possible in 84% of cases. All deaths reported by local investigators were confirmed by the EPC. Local investigators reported 78 cases of reinfarction and 31 (40%) of these events were rejected by the EPC, whereas the EPC discovered 35 unreported cases of reinfarction. Thirty-one strokes were reported by local investigators of which 2 (6%) were rejected by the EPC which, however, included 2 previously unreported strokes. A total of 383 cases of bleeding were reported of which 115 (30%) were rejected, whereas the EPC included 16 another cases of unreported bleedings. Blinded end point adjudication therefore contributed significantly to the final number of end points. Conclusion: Blinded end point adjudication is feasible in open randomised trials and is likely to reduce bias.