I can say there is good evidence suggesting that people on average engage in, say, a methadone program five to seven times before they're able to manage their substance use long term.
There is evidence on the recovery side. For example, there was a randomized clinical trial of buprenorphine and some counselling, versus just counselling and no pharmacotherapies. It found that the mean abstinence rate for the counselling group was 5%, compared with 43% in the buprenorphine group. Again, there is really good evidence to suggest that pharmacotherapies can provide more effective abstinence compared to recovery-oriented non-pharmacotherapy treatments.
Again, I would reiterate that the catch-22 that happens where people can't find housing because they're using substances but can't access treatment because they don't have housing is a major issue that is going to prolong the life cycle of this overdose epidemic.