Skip to main content
Fig. 2 | Trials

Fig. 2

From: Bounding the per-protocol effect in randomized trials: an application to colorectal cancer screening

Fig. 2

Age-standardized bounds (for ages 50–64) for the per-protocol 10-year risk difference under various sets of assumptions. Sets of assumptions include: a No assumptions. b The instrumental conditions (relevance, exclusion restriction, and exchangeability). c The instrumental conditions plus an assumed maximum risk under screening in the “never-takers” of 2 %, 1 %, and 40 % for the CRC incidence, CRC mortality, and all-cause mortality, respectively. d The instrumental conditions plus an assumed maximum risk under screening in the “never-takers” of 1.5 %, 0.75 %, and 30 % for the CRC incidence, CRC mortality, and all-cause mortality, respectively. e The instrumental conditions plus an assumed maximum risk under screening in the “never-takers” of 1 %, 0.5 %, and 20 % for the CRC incidence, CRC mortality, and all-cause mortality, respectively. f The instrumental conditions plus additive effect homogeneity. g The instrumental conditions plus multiplicative effect homogeneity. The dotted line indicates the intention-to-treat effect

Back to article page