In this particular example, the patient is exposed to being crushed, so the hazard is a crushing hazard, which is related to the possible harm. The fact that the patient moved or was moved is not related to the harm.
I think that this is a little backwards. Saying that there is a "crushing hazard" is like saying (in another context) that there is a "death hazard", which is wrong. Crushing, or patient crush, or getting crushed, is a harm. "A crushing hazard" is just a non-insightful / non-specific way of relating to several different hazards that could lead to the crushing harm. For example, a patient may be crushed because the device moved them against something, or because a heavy device installed above the patient (and not intended to move the patient) dislodged and fell on them. These are two different hazards that lead to a similar harm through different hazardous situations. The risks associated may be different and so may be the mitigation means.
In my example, the fact that the patient was moved
is related to the harm, because that harm wouldn't have come about (through this particular scenario) if the patient wasn't being moved. There would still be a
potential for harm (a hazard and a risk), but there would not be actual harm. This is where probabilities come into play.