Okay, so pick your psychiatric diagnosis-- only don't pick Adjustment Disorder, or Major Depression, single episode. Pick a psychiatric diagnosis where we Know that it recurs and where long-term treatment is indicated. Let's say schizophrenia, or bipolar disorder, or recurrent major depression with a bunch of episodes. Let's say the episodes are bad and the patient gets lots of symptoms and life gets ugly.
So pick your medicine to treat Illness X. The patient takes the medicine and most of the symptoms get much better, the patient feels better, everyone takes a deep breath, the side effects are minimal or non-existent. Life is good, though the patient still has some problems (ah, don't we all....) and lives a bit on the edge in a way that leaves us wondering-- is there a personality disorder here? A developmental issue? A social issue? Or are there perhaps some residual symptoms? Maybe this is just one of those people who will never fit neatly into a boxed corporate-climbing life, or for whom meds and therapy won't be complete answers.
We're moving along okay, nothing scary is happening, the patient is mostly well, the medicine is tolerated, life is looking up. And then an episode hits---this is not "supposed" to happen. But we all know that the medicines decrease the likelihood of a recurrence of illness, while they are no guarantee.
So we take our ill patient and we do what one might do: raise the dose, assess symptoms, increase the frequency of sessions, get thee to a lab: check levels, look for other things that could account for the sudden symptom exacerbation, think about drug interactions and what's that thyroid doing anyway?
The patient returns. Ah, much better, the symptoms have abated, the patient feels better than ever. The obvious signs of illness are gone. For the sake of clarification, in psychiatry "signs" are thinks we can see-- psychomotor slowing or activation, abnormal movements, changes in the rate of speech, disordered thoughts, conversations with non-existent people...fill in the blanks. The patient is eating and sleeping better, functioning better, less irritable, less chaotic.
One little thing, Doctor: "I stopped the medicine."
Oy. So the patient stopped taking the medicine that treats the illness and gets much better. Maybe the problem wasn't a breakthrough of symptoms, maybe it was that the patient was having unrecognized side effects from the medicine and feels better without it? Nope, the symptoms were classic illness symptoms, not side effects. Why would they get "better" from the psychiatric symptoms when the med stops? I have no idea. And yes, I promise you, the patient had the symptoms before any psychotropic medication was ever started-- this isn't simply an adverse reaction to the medication. The best I can do is that the episode was self-limited and happened to end as the medication stopped, but that feels a little lame even to me.
So now what? The patient had numerous episodes of the illness before getting diagnosed and treated. But, really, you can't say to a patient: This is the gold standard of treatment for your illness, take the medicine even though you feel much better since you stopped it. Oh, I guess you could say it, but no patient will listen.
It's hard to prophylax well patients. We could try another medication on the theory that it may protect against future episodes, but if someone is feeling well, there is very little immediate up-side to prophylaxis: You feel well now and you may get side effects (oh, and you'll have to get labs and EKGs and maybe the new medicine will give you lovely adverse effects). We could do nothing and wait: it's pretty clear that it's just a matter of time and the "well" patient is a time bomb.
It's not supposed to work this way.