TransLink in Vancouver does not report suicides, instead choosing to vaguely announce a “medical emergency” at one of their stations. The term does not always refer to suicide, but when stations are closed as the result of a medical emergency, this is very often the case.

Over the past year, there have been an average of two “medical emergencies” per month resulting in closed SkyTrain stations. Many of these are confirmed to be suicides by eyewitnesses on social media describing what they saw.

Statistics on suicide by train are recorded in the UK, where in 2013 there were 308 incidents, up from 279 the previous year. This means that it is more likely than not, on any given day, that somebody is going to jump in front of a train in Jolly Ol’ England, in spite of one advantage the London Underground and other systems have over Vancouver’s SkyTrain: Some stations have “suicide screens” installed, which are protective barriers blocking access to the tracks until the train has stopped in the station.

Almost all stations in Tokyo have glass barriers between the tracks and platform, but it remains the world’s capital for this method of suicide. And all methods of suicide, really, on account of it being Japan. There are notices posted in stations with mental health information and suicide hotlines, and passengers are even encouraged to practice the procedure for responding to a jumper. It sounds ridiculous, but some stations play videos of kittens to help dissuade the depressed from ending their lives.

Unfortunately, if just one station is not outfitted for suicide prevention, people can just ride the very transit system they intend to use to end their lives, right to the best station for the job. Platform barriers are certainly a good idea to stop accidental deaths and spontaneous impulses, but most suicides are thought out well in advance.

The question at this point is how far we should go to stop people from killing themselves, because a determined individual is going to be successful one way or another. There are almost certainly cases where we have no business trying to stand in someone’s way. However, suicide by train is a problem that impacts witnesses much more than it does victims (since they’re dead and shit). For each suicide, there are potentially dozens of people who will need to seek counselling to feel safe on public transit again, and for some people, the result of witnessing something so shocking and gruesome can be much worse.

One way to stop people from killing themselves in terrifying, public displays, would be to make access to superior methods more readily available. In Switzerland, Dignitas has legally assisted in the suicides of over 1,000 individuals. These are not just patients with terminal illnesses or the elderly; in 21% of cases these are people suffering from “weariness of life.”

Suicide rates in Switzerland have steadily declined over the past 30 years. 70% of those approved for suicide change their minds during the process. It seems that, while having doctors prescribe suicide to patients might make a lot of people uncomfortable, it comes with large benefits. We’re definitely not doing a good job in North America.

Still, there are those who want to kill themselves publically and draw attention to their suicide, and train platforms remain near the top of the list of options. Compared to another popular choice, like jumping off of a bridge, the number of witnesses very near the event is usually going to be greater. It’s also typically a much more horrific way to die than anyone expects, taking an average of seventeen minutes to kick in (I Googled it) thanks to your typical train’s uncanny ability to ziplock and cauterize the human body during bisection. Gross. Also neat. But gross.

If suicide screens cannot be installed in all stations of a particular system, or if determined people can simply get around them, another option might be to just slow trains down before they arrive at the platform. The extra few seconds required to do most of the braking outside of the station, and then roll in casually, is not going to dramatically impact system efficiency.

I don’t have a strong list of answers to these problems, but it’s on my mind today after reading a bunch of tweets about it. I would be failing to take my own best advice if I didn’t make an effort to bring up the uncomfortable.