Is there direct evidence that screening reduces morbidity or mortality?
Direct evidence for the efficacy of screening is ideally provided by large randomized, controlled trials (RCTs) with longitudinal follow-up to compare morbidity and mortality outcomes. Randomization is the only way to eliminate length of time bias (i.e., the greater likelihood of detecting cases with long rather than short preclinical phases) and volunteer bias (i.e., the tendency of volunteers to be healthier and have more positive health behaviors) among those subjects who are screened compared with those who are not screened. RCTs are still vulnerable to lead time bias, but the impact of this bias can be limited by using morbidity or mortality rates rather than case-fatality or survival rates. The feasibility of RCTs to generate direct evidence depends upon the length of follow-up needed to identify differences in morbidity or mortality. If the natural disease course between the condition targeted by screening and the health outcome to prevent is short, then obtaining direct evidenc
