I generally agree with Susan's point--who the hell knows what's going to happen? "Interventions", while dramatic and probably somewhat satisfying to the participants who feel they are doing something concrete about a frustrating situation, aren't very effective, any more than any other coercive treatment attempts.
In the Virginia Tech case though, there are a couple of really egregious points.
-They had a double homicide on campus and no apprehended suspect, basically no idea what was going on except that an unidentified killer had struck on their grounds and was loose. They should have closed the campus instead of pursuing "business as usual".
-It appears they automatically, as standard procedure, investigated the status of the first female victim's boyfriend and prepared to search his home for weapons he was known to possess. This is a reasonable avenue of investigation but does not excuse the sweeping and tragic conclusion, repeatedly stated by the soon-to-be-unemployed (I suspect) university president, "We thought it was an isolated incident".
-It seems persons with a history of stalking, mental instability and threatening behaviour have no problem buying firearms and ammunition in Virginia. Might want to reconsider that. Oh, sorry, didn't mean to point out the elephant in the living room.
These questions are quite different from hypotheticals of what should have been done before the situation turned tragic.
We consider and discuss these things, whether our points are valid or not, because we want to believe it didn't have to happen, and that we can learn how to make sure it doesn't happen again. If only.