[Primary prevention of varicose bleeding: current status]

1. Surgical shunts should be abandoned for prevention of the first bleeding because of an increased mortality rate, despite a marked reduction in frequency of bleeding. 2. Sclerotherapy is not recommended for clinical practice in preventing the first bleeding. The meta-analysis of numerous randomized clinical trials shows a small effect on bleeding risk and mortality rate. But these effects are remarkably heterogenous in various studies and are compensated by the complications of sclerotherapy. 3. Prophylactic treatment with beta-blockers has similar results as sclerotherapy documented by meta-analysis of several studies. In contrast to sclerotherapy both risk and side effects are minimal. More controlled clinical trials are needed to define subgroups of patients with higher benefit of the prophylaxis. In individual high-risk patients (size of varices, liver dysfunction, red-colour-signs) beta-blocker prophylaxis may have possible benefit.