Wonder if anyone will see this, but I have a general question I might as well ask here.
What if, you put an addicted person in a padded cell, and supply progressively decreasing doses of the drug(s) the person is addicted to?
The dosage gradient could be mathematically computed to find the lowest possible harmful withdrawal effects.
The person would stay in the cell environment until completely past the hardest phases of withdrawal (however
long this takes, whether days, weeks, months(?)), but they would be free to move around (within the secure complex) as they please, and have limited access to technology (anything enabling communication would be heavily vetoed, but eg unlimited access to a video library could work well).
It seems that the current models in use depend entirely on the will of a person who is heavily compromised. I honestly have to say, that seems fundamentally and ridiculously broken, and yeah, even suspiciously so: it's so obvious that this approach is not going to work for most, so why does every "treatment center" continue down this path?
The only caveat I can think of with the suggestion above is about the violent phase of withdrawals, and sedating a person enough when in that phase so they can eat/drink/etc. (I'm not sure how long it is - I vaguely recall it's just a couple of days right?)
Also, about the padded room - I'm not talking "psychotically mad-proof" padding, but just enough for this context. Obviously that's going to change per drug and person, and maybe thick padding is what would be required.
I'm just thinking about cost here (cost of sedation and padding), because I know that a) everything health-related in the US seems to attract the most markup and overhead in any field in the world, and b) because (as per my notation that the current model is _suspiciously_ broken) I wouldn't be surprised if anyone who tried this would find it mysteriously hard to get funding.
On the other hand, I must admit that I've never really tried to find out if this idea has been attempted or not. Would/does it not work?