Seol mar théacs é seo: Statistical issues in randomized trials of cancer screening