> If they can do it then a university can definitely do it.
I'm not sure how comparable the two situations are, although there are certainly measures in the article you linked to that could be applied anywhere. The two populations in question (Detroit mental health services patients vs. MIT students) are very different, as are the environments.
> better access controls to rooftops
This was done while I was at MIT, but I think the controls were relaxed again some years after I left. One of the issues with any institution over time is that different administrations have different priorities, and often change things for no good reason. MIT, unfortunately, is not immune to this problem.
http://www.henryford.com/body.cfm?id=46335&action=detail&ref...
If they can do it then a university can definitely do it.
I do understand the sentiment behind "zero is the aim, but pragmatically we can't stop all suicides", but that attitude allows people to die.
Even if we do accept that we can't stop every death by suicide we can say that the rate is currently far too high.
As one example: People have mentioned deaths from tall buildings. Research shows that you can reduce suicide rates by restricting access to means and methods, so better access[1] controls to rooftops, with anti-suicide fencing, and suicide-prevention telephone number signs, will reduce the numbers of people dying by suicide.