Yeah, you could do competing risks. I see yclist.com does distinguish between 'dead' and 'exit', although your CSV doesn't. You can use 'comprsk', I remember it as being no harder to use than 'survival' when I tried it out briefly for https://www.gwern.net/DNM%20survival . Any competing-risks model going to be very imprecise, though, since if you can't estimate the overall survival curve very precisely, you definitely can't estimate it split into multiple outcomes.