Well, one thing to think about is that there are certain things you teach to SINGLE ENGINE students that help to set them up for more complex airplanes with more engines. You NEVER EVER let them get slow, you make them keep their hand glued to the throttle during any critical phase of flight (takeoff, landing, low alts). You look for disconnects such as students that don't pitch down automatically when they increase flaps, you look for students that don't slow down and manage airspeed before descents, you have them do a short field landing on the numbers and see if they are prematurely "pulling up" because they are over the dirt and freaked out by it, you look for students that are fixated on the runway during base that pitch up while they are turning to final, you look for the signs that they don't really understand coordination (yaw) during stalls and other maneuvers (like chasing the inclinometer, which the purpose of is to tell you that you've already screwed up
). None of these are multi-engine specific, yet they are all the reasons that someone will kill themselves in a more complex or multi-engine aircraft. I'm not saying that you have to go back to SE aircraft with these kinds of students (well, maybe, haha), it's just that you are looking for a lot of the same things, and you obviously address these before you worry about attempting a Vmc demo or single engine approach.
Nothing in here is radically different if you take the same approach with SE airplanes. Brief your plan of action, then execute your plan of action if something goes wrong. It used to bug me to no end on checkrides when and experienced student would brief "and if something goes wrong, I'll give you the controls". Always be prepared to do what you brief. Chances are you'll do it really crappy, but at least survive