Age-specific Incidence in the Absence of Screening
Trends in breast cancer have been steadily increasing in the US for decades although the reasons have not been
well understood. The pattern is partly due to screening; however, previous studies have suggested that screening
does not explain all of the trends in incidence1. An age-period-cohort analysis was performed to
separate the portion of the trend in incidence that was due to screening and the portion that reflects changing
risk over time. Change in risk over time is represented in the model as a cohort factor with different birth
cohorts at varying risk for breast cancer assumed due to differential exposures to risk factors. Results from
the age-period-cohort model were used to estimate what incidence would have looked like if mammography screening
had never been introduced into the population. Incidence estimates without mammography screening2 by year of birth and age were provided by Theodore Holford
through personal communication.
For additional information on the age-period-cohort model or how incidence rates without mammography were used in the base case modeling, please contact Statistical Research and Application Branch of the National Cancer Institute.
1 Holford TR, Roush GC, McKay LA. Trends in female breast cancer in Connecticut and the United States.
J Clin Epidemiol 1991;44:29-39. [PubMed]
2 Incidence Rates in the absence of screening
(74 KB - PDF)
|
