I want to study a residual risk of transfusion transmitted viral infections (HIV, Hepatitis B and Hepatitis C) in blood that has been approved for transfusion in Tanzania (using Polymerase Chain Reaction). My hypothesis is that 1% of the blood approved for transfusion is infected with HIV, (and for Hepatitis B and C the prevalence might be less). The overall prevalence of HIV in Tanzania is about 5.7%. How can I calculate the sample size?