In Reply. —
We have a number of technical concerns about the article by Steinberg et al1 that lead us to believe that their confidence interval for breast cancer risk after 15 years of estrogen replacement therapy (ERT) is too narrow. Steinberg et al1 fit a simple exponential model with a single slope parameter to complex data from multiple studies. They obtain a separate slope parameter estimate from each study and then take a weighted average of these estimates to determine the duration-risk effect from all studies. Unfortunately, most of the studies that have looked at the duration effect have followed up women for less than 10 years, and there have been only three studies2-4 that have observed substantial numbers of patients who have taken ERT for more than 15 years. Thus, by taking a weighted slope estimate from all studies, regardless of their duration, they are extrapolating
Dupont WD, Page DL. Estrogen Replacement Therapy and Risk of Breast Cancer: Results of Two Meta-analyses-Reply. Arch Intern Med. 1992;152(5):1090–1093. doi:10.1001/archinte.1992.00400170156031
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: