With a stepped wedge design, in which, over time, the intervention is turned “on” in all participating sites (Hughes et al. 2015), one might expect that implementation at participating sites would be fairly seamless; if the intervention is working, sites can simply leave it turned on. However, there is a question of timing—does one turn off the intervention while waiting for results or leave it on? And, based on the experiences of the Collaboratory, there is ample variation among sites and individuals regarding the fidelity to an intervention.

In the chapter on Dissemination and Implementation, we describe a case example from the Lumbar Imaging with Reporting of Epidemiology (LIRE) trial.