I'm one of the people who's done some disaster work (I volunteered with the Red Cross after the Northridge quake, in 1994) and also crisis work before that, so I'll comment on informed consent in these environments.
We worked in a Red Cross service center (literally a 40'x48' tent, set up in a park). People came in looking for help with rent, food, clothes, etc. -- not for psychological help -- but in relating their story to the family services worker, often trauma feelings would emerge. That's where I came in: keeping an eye out for such incidents, I'd walk over, introduce myself ("Hi, I'm David") while pointing vaguely toward a table in the back with a 'mental health' sign over it, and offer to talk with them if they wanted, either then or when they'd finished the family services interview. This function was invariably seen as helpful both by the victims and the ARC workers (who didn't have to deal with people crying, so much).
Sometimes just that interruption was enough; people might or might not come by my table afterwards. Most times they did and we talked, but usually briefly. It was always voluntary, in the sense that this was offered as something they could do and they were free to come by or not. People could come back later if they wanted, but didn't much: there were no appointments and also no continuing relationship in the sense of multiple psychotherapy sessions. [Continuing or dual relationship issues potentially arise more often with other volunteer workers than with disaster victims, since we worked together in 12-hour days, usually seven days a week. My role was to work with both groups, and included some aspects of personnel management.]
Disaster work like this is typically quite brief. No time for an hour session when there's a mash tent atmosphere, no privacy, and I might need to deal with a far worse situation at any minute. Maybe we'd have 5 minutes; maybe less. Lots of giving out of information, to help normalize the "odd & crazy" feelings typical after something so traumatic -- thats the most common "intervention" I did, I suppose, besides handing out teddy bears or crayons to the kids.
[I probably spent almost as much time trying to weasel more teddy bears out of the regional headquarters than I did doing "therapy" (a resident puppy might have been as good, but we didn't have one). But the work isn't so well defined as that, even: once it rained, wetting the wires so our phones went out. I fixed them by cleaning the connections with alcohol swabs (from the nurses) and a pencil eraser, then sticking plastic forks through holes in the tent walls to keep the lines off the wet ground. Shades of my days as an Eagle scout.]
Sometimes, I was able to sit and talk a bit (listening, mostly) with someone -- if the tent seemed quiet and the other psychologist was around so she could deal with anything that might arise. It helped when we could discover connections between the quake trauma and earlier upsetting things that had happened to them. These connections helped make some sense (or meaning) out of their feelings, and I could see the relief when this happened. Complex cases were referred to local providers of course, but I have no idea how often these were followed up. Very much an informal sort of consent, I suppose; generally you did what seemed like the right thing to do at the time.
Disaster mental health work is mostly seat-of-your-pants kind of stuff. For me, it was a wonderful and very informative "field trip" to observe trauma responses when (unlike in my practice) most everyone I met had been affected by the same objective traumatic event. Roles here are never well-defined, and it requires a high tolerance for ambiguity (thats me) to thrive when everything is always up in the air. The professionals who came in all dressed up, expecting an office and appointments and the like (!), really didn't do very well in these situations. The issues are more like in crisis work, or Ram Dass' excellent How Can I Help? book, than traditional therapy.
Consequently, though still needed, informed consent is much more informal here, and the risks and issues are different enough that you can't just apply the forms and approach as would work (and are necessary) in private practice.