Objectives: The beneficial effect of oral anticoagulants after infrainguinal venous bypass surgery is compromised by bleeding complications. We developed a model to identify patients, treated with anticoagulation, at risk of major haemorrhage and estimated whether this complication could have been prevented if patients had received aspirin.
Design: Randomised clinical trial.
Methods: Data of patients who participated in the Dutch Bypass Oral Anticoagulation or Aspirin Study were reanalysed using Cox regression. After infrainguinal bypass surgery these patients were randomised to oral anticoagulants (n = 1326) or aspirin (n = 1324).
Results: Predictors of major haemorrhage for patients on oral anticoagulants were increased systolic blood pressure (> or = 140 mmHg, hazard ratio [HR] 1.62), age > or = 75 years (HR 2.77) and diabetes mellitus (HR 1.60). If the 345 patients in the highest risk quartile had received aspirin, major haemorrhages would have been reduced from 46 to 22, with no major changes in ischemic events and graft occlusions. In the subgroup with venous bypasses major haemorrhages would have been reduced from 27 to 13, at the cost of seven more ischemic events (mostly fatal) and 17 more graft occlusions.
Conclusions: Treating patients at highest risk of major haemorrhage with aspirin instead of oral anticoagulants would have resulted in a reduction of non-fatal haemorrhages, but for venous bypasses this reduction was outweighed by an increase in ischemic events and graft occlusions. We still recommend treatment with oral anticoagulants after peripheral venous bypass surgery.