[Problems in using survival data of cancer patients as a method of gauging the effectiveness of cancer control efforts].

Survival rates of cancer patients have been widely used as a method of gauging the effectiveness of cancer control efforts. Sometimes, however, they cannot be considered reliable indicators due to many biases, such as a lead time bias, a length bias, a self-selection bias and an overdiagnosis bias. Instead, cancer mortality rates should be used for evaluating screening programs and for assessing the recent progress against cancer, although some problems are encountered in the accuracy of death certificates.