—It appears that Dr Swift and colleagues misunderstood our use of matching on time since certification in our study of breast cancer among 105 385 US radiologic technologists.1 While the possibility of overmatching on a correlate of exposure may be a concern, matching on year of certification was not equivalent to matching on years of work. In fact, the distribution of years worked was quite varied within intervals of year of certification, reflecting independent decisions by women to cease active work following certification for various reasons such as raising a family. Even before matching, this pattern, shown here in the Table for all female questionnaire respondents certified between 1940 and 1959,2 was fairly uniform across 5-year intervals of year of certification. That is, women were equally likely to work fewer than 5 years (10%-14%) as to work 30 to 34 years (9%-15%) within categories of certification year. Thus,
Boice JD, Doody MM, Mandel JS. Breast Cancer Among Radiologic Technologists-Reply. JAMA. 1996;276(5):369–370. doi:10.1001/jama.1996.03540050029010
Customize your JAMA Network experience by selecting one or more topics from the list below.