Once again, I would suggest that the potential risks of exposure are being seriously downplayed here by people who don't appear to have any significant background in radiation and the related helath effects.
Relatively little is truly
known about the effects of tritium exposure on the body, and (in the event of a gas release) routes of exposure are not entirely limited to inhalation of the dry gas (which is apparently the least dangerous). Broken exit signs are considered hazardous by the EPA, and a review of this
webpage would suggest that it's not an incident to be treated lightly (note some of the vagueness of the language used here, like "gas...
most likely will quickly disperse " - this would suggest a lack of full understanding).
I'm not trying to scare anyone, I own plany of tritium myself and consider it to be quite safe. I just personally feel it should be handled (especially when released) with an abundance of caution just because it's not worth the risk. I also think its a disservice to the CPF community to not make people aware of the potential risk, regardless of whether or not it is significant - this is why I have repeatedly posted in these threads with links to relevant information. To state things as fact which have little or no scientific study to back them up is just careless, IMO.
So, in short, if you break a marker or glowring - don't freak out! But it doesn't hurt to wash things which could have been in close contact with escaping gas (especially for large tubes like in glowrings), and dispose of the broken bits carefully. Most glowrings don't contain huge amounts of tritium, but some can contain levels where the risk starts to become a concern (in excess of 1 curie) and it's entirely possible that a significant internal dose from such a source could pose some health risk (although getting all that tritium into your body from one single incident isn't very likely). A recent study on the internal effects of tritium exposure was completed in the UK and can be found
here. This study suggests that the risk is greater than previously thought, enough so to recommend an elevation in hazard classifcation.