I think Dinah raised an interesting topic in her last post. Can psychiatrists ever refuse to treat someone, or to treat them only under certain conditions?
I'll leave Dinah and Roy to talk about the free society way of handling that; I can address what happens inside the walls.
The patients I treat get locked up because of persistent misbehavior or persistent high-risk behavior. (I don't call it 'self-destructive' behavior---even though it is---because inside the walls that term has a very different connotation.) I don't really ever have the option of not treating someone. My clinic is never too full to accept new patients and our jail/emergency room never goes on 'fly-by' status due to lack of bed space. So, I get all comers.
That being said, I do have certain limits. By the time I get my patients they are generally ready to accept the idea that their behaviors are getting them into trouble. The biggest issue is how to handle the interventions from that point on.
The biggest trick with treating substance abusers is that there's a big difference between what they say they want and what they actually want. They say they want to be drug or alcohol-free, to give up 'fighting the system' and to get themselves together. But at the same time they want to be in control of treatment, in control of their environment and to have it all done on their own timeline.
This is a setup for frustration.
Fortunately, it's all manageable. The key is to be upfront about expectations and limits and to be true to your word. Inmates can accept a 'no' but it's a very very bad idea to imply you can or will help them with something you have no control over. I don't make housing or cellmate changes, I don't order lower bunks or special diets, I don't make phone calls or transmit messages for inmates. That's pretty much the easy part.
Occasionally inmates aren't happy about that. Very rare inmates may escalate their requests along a continuum that I can pretty much predict: a hint for a favor turns into a direct request which turns into a demand which turns into a threat. Once you reach the demand/threat point treatment stops.
So getting back to the idea of conditional treatment, my only real condition is the requirement for safety. Once a patient gets to making a threat of violence to me, to himself or anyone else, treatment stops and safety interventions begin. That's really the only way to get things back on track.