How do we decide how representative our donors are for public health surveillance?