Items in FPM with MESH term: Bias (Epidemiology)
ABSTRACT: Many patients expect to undergo screening tests for cancer. In evaluating screening procedures, physicians must take into account the known effects of lead time, length and screening biases, all of which can result in an overestimation of the benefits of screening. The gold standard by which a screening test is evaluated remains the prospective, randomized controlled trial, demonstrating reduced morbidity and mortality. The magnitude of benefit from screening is best expressed in terms of the number of patients needed to screen. This value ranges from approximately 500 to 1,100 for proven screening interventions. These concepts are illustrated by controversies in current screening recommendations for cancers of the cervix, lung, colon, breast and prostate, which together account for more than 50 percent of cancer deaths in the United States.
Spoofing Spurious Science - Editorial