To the point about prognosis, my research has shown.... We have actually, my co-authors and I, extensively looked at the question of prognosis prediction in psychiatry, looking at treatment-resistant depression as a paradigm case, looking both at clinicians' predictions and precision medicine. The conclusion is that we cannot predict prognosis. Contrary to what Dr. Trew was saying, the state-of-the-art science says that even when we use precision medicine, the best prognosis prediction in the long term is at the level of chance. That is what the science says. That is what has been published on this topic.
To the point about suicidality and autonomy that you're raising, I want to say this: Even if we agree, and we can, that some cases of persons with mental illness who have a wish to die warrant our compassion and assistance, we need to reckon with the fact that other cases of persons with a mental illness who want to die will warrant suicide prevention. No one believes that MAID should replace suicide prevention. The problem is that we don't have parameters to decide when to accept and when to reject patient autonomy on this.
It's helpful to clarify that when we talk about autonomy and if we want to be serious about autonomy, we talk about informed consent. The trouble is that many cases of patients who today receive suicide prevention meet the requirements for informed consent, so if we want to be serious about patient autonomy and if we want to legalize MAID for mental disorders, we first need to have a major overhaul of the way we do suicide prevention.