Assessment of cancer screening effectiveness in the era of screening programs

Whilst the benefit of detecting and removing cancer precursors in the cervix and the large bowel is unequivocally documented, the balance between benefit and harm due to screening for early detection of breast and prostate cancer remains uncertain after 30–50 years of randomized control trials (RCTs). Concomitantly, traditional RCTs become increasingly unfeasible due to low compliance, contamination of control groups, difficulties to enroll screening naïve individuals, and prohibitively high costs. Therefore, we have systematically reviewed pros and cons with alternative study designs to quantify screening effectiveness. Whilst traditional RCTs should remain gold standard whenever feasible, observational cohort and case–control studies are unlikely to provide the valid evidence needed for health policy decisions. Instead, we advocate two methodological approaches that are novel in assessment of cancer screening effectiveness. One—the difference-in-difference design, well established in health policy and public health research—has a sophisticated ecologic design which unfortunately is feasible only in few countries. Another implies that sequential RCTs become embedded whenever population-based screening programs are implemented.