*3.1. CNBSS Findings*

Figure 3 illustrates the excess in cancers found in the Study arm compared to the Control arm in each of the CNBSS trials. For invasive cancers, at the end of 5 years of the screening intervention (defined here as time 0), there was an excess of 59 and 79 cancers in the younger and older groups, respectively. When all cancers (invasive and in situ) were considered, the corresponding excesses were 92 and 115 cancers, respectively. It can then be inferred that there were 33 excess in situ cancers found in CNBSS1 and 36 in CNBSS2 over the period of screening. However, there is a dramatic difference between the trials in how the excesses varied as time elapsed after the screening intervention in the two trials.

**Figure 3.** Excess invasive cancers in the study group compared to the controls in the two CNBSS trials.

Using the definition employed by Baines et al. for observation periods between t=0 and 20 years after succession of screening (t = 20 y), for invasive cancers, the estimated overdetection fraction ranged from 28% to 48% at t = 0 and at t = 20 y, respectively, in the women in their 40s and between 29% and 5% for women in their 50s. Estimates were higher (36–53% in the younger group and 35% to 15% at t = 0 and 20 y, respectively, in the older group) when both all cancers (in situ and invasive) were considered.

In CNBSS2, for invasive cancers, the behavior over time is as would be expected with the excess gradually diminishing over the first 10 years after the period of screening as cancers gradually surface (without the benefit of screening lead time) in the control group. When the in situ cancers are included, the absolute excess that they represent at t = 0 appears to persist over time.

In CNBSS1, perplexingly, the excess of invasive cancers actually increases over the 20-year post-screening follow-up period. As in CNBSS2, the absolute contribution of in situ cancers to the excess appears to be essentially constant.
