Hi Nick
I wonder why cannot you acquire images with the other linac when the original linac goes down? That is what we always do, no matter how well the patients positioning has gone in the original linac .
Here, the breast treatments are pretty much the only group, which are allowed to go without daily IGRT. If we want to treat breast patient after three/four fractions without IGRT, weekly verification tangential image is however mandatory and systematic errors in isocenter and in rotations should be near zero in those first three/four first images in the first place to make a decision to proceed with SGRT only.
As well, we can continue with CT SIM surface in SGRT only workflow, but there are cases where new setup reference surface needs to be taken, like you suggest. In those cases, this new setup reference surface should be verified in the next fraction(s) with IGRT and again online match shifts should be near zero to make a decision to go on with SGRT only. It is always departments decision what kind of possible systematic error is allowed (on the background) to proceed with SGRT only. However, it is not optimal if the daily random errors exist around the value of for example 3 mm or 2 degree (systematic errors) if we can easily fine-tune our setup reference surfaces in SGRT only workflow such that the random errors exist around near zero systematic error.
Marko