Ok I'll interject my two cents as I've lived and worked in Los Angeles for several years doing both 911 and IFT. Isolating a single activity as increasing the risk or morbidity and mortality is near impossible, despite what statistics will say, I can manipulate the data or exclude data to prove my point every time. Yes I've taken statistics and also taken courses where we worked the numbers to research an issue, additional have read many papers that the author shows proves their point, but when you look at the data, it's quite obvious that the data doesn't match the results the author talks about. A majority of marathons are run early in the morning and by noon all the streets closed are open again, additionally as every marathon participant is accounted for they opened up the roads behind the runners, so you could cross those streets much earlier then when the marathon was complete. There are also several parades, mass gatherings, concerts, and even extensive road construction that can shut down roads for even longer periods of time. Are you proposing that nothing can be done in the roadway at any time due to the possibility of a person dying because of some event? If there is such a huge rush, then more planning needs to be taking place to figure out maybe a different station needs to cover that territory while the event is going on and maybe a certain hospital can't be reached, so another one automatically will become the closer one due to the traffic diversions to get there.