Assessing the impact of early detection biases on breast cancer survival of Catalan women
Tipus de documentArticle
EditorInstitut d'Estadística de Catalunya
Condicions d'accésAccés obert
Survival estimates for women with screen-detected breast cancer are affected by biases specific to early detection. Lead-time bias occurs due to the advance of diagnosis, and length-sampling bias because tumors detected on screening exams are more likely to have slower growth than tumors symptomatically detected. Methods proposed in the literature and simulation were used to assess the impact of these biases. If lead-time and length-sampling biases were not taken into account, the median survival time of screen-detected breast cancer cases may be overestimated by 5 years and the 5-year cumulative survival probability by between 2.5 to 5 percent units.
CitacióRoso-Llorach, Albert [et al.]. Assessing the impact of early detection biases on breast cancer survival of Catalan women. "SORT", Desembre 2014, vol. 38, p. 139-160.