Duration of Maintenance Treatment After a Single Remitted Psychotic Episode
A year symptom-free after one clean psychotic episode, a patient wants to stop her medication. Guidelines themselves don't agree on whether that's a reasonable next step or a real relapse gamble.
R.A., a 24-year-old woman, has spent the fourteen months since her single psychotic episode rebuilding a life that, by her own description, looks almost exactly like it did before: full-time work as a paralegal at a mid-sized firm, a close relationship with her parents who live nearby, and a small group of friends who stayed with her through the worst of it. She had one psychotic episode fourteen months ago — two weeks of paranoid delusions and disorganized speech during a period of severe sleep deprivation and a breakup, treated with risperidone, fully remitted within six weeks, and stable ever since.
She works full-time, has had no further symptoms of any kind, and now asks her psychiatrist directly whether she can stop the medication she has been taking for over a year, saying she feels like herself again and doesn't want to be on it indefinitely. She describes this not as impatience but as something she's thought through carefully over several months, watching herself stay well and wondering, reasonably, what continuing treatment is still protecting her from.
This is not a fringe question with an obvious answer. Some guidelines recommend at least one to two years of maintenance after a first remitted episode before any discontinuation is even considered, given that relapse substantially raises the risk of a harder-to-treat, more chronic course. Others hold that a genuinely low-risk patient — clear precipitant, rapid and complete remission, strong insight, stable psychosocial supports — is a reasonable candidate for a supervised taper well before that window closes, since indefinite treatment carries its own real cost in side-effect burden and stigma for someone who may never have a second episode at all. Nothing in her presentation is unusual enough to settle which reading applies to her specifically.
Continue, taper, or wait
I'd hold at fourteen months — not because I doubt her judgment, but because of what actually happened in the trial this guidance comes from. In Wunderink and colleagues' original randomized trial, published in JAMA Psychiatry, comparing dose reduction/discontinuation to maintenance after a remitted first episode, the dose-reduction group relapsed more often than the maintenance group during the two years the trial itself ran. She's fourteen months out. That's squarely inside the window that finding describes.
You're right about the short-term numbers — that's real, and it's the same study I'd cite. But it's not where the story ends. The same patients were followed out to seven years, and by then the dose-reduction group had reached functional recovery at more than double the rate of the maintenance group — roughly forty percent versus under twenty — with no significant difference in symptomatic remission between the two arms.
You called fourteen months "squarely inside the window" the relapse finding describes — I'd say it's also squarely inside the window where deciding on that early number alone, without weighing what the same cohort looked like five years further out, risks optimizing for the wrong outcome. Staying on the medication isn't itself the goal. Staying functionally well is, and by that measure the longer view tells a different story than the two-year snapshot does.
The detail that actually moves me is one you both mentioned but didn't sit with: symptomatic remission barely differed between the two groups. That means the recovery gap wasn't the maintenance group relapsing and falling behind — it opened up specifically in functioning and social recovery, in people who by symptoms alone looked equally well. So this was never really a question of who stayed sicker; it's a question of who got their life back further.
From where I sit, seeing her for unrelated care, that reframes what a year of her own demonstrated stability is worth — it's not just eligibility on paper for the lower-risk profile in that trial, it's a specific, observed track record in the person actually in front of us. I'd hold now, as the relapse data argues, but I wouldn't leave the next step open-ended: a named date to revisit, not a "someday."
Agreed: continue risperidone at the current dose for now, with a scheduled taper conversation at the eighteen-month mark rather than an open-ended 'someday.'
Not agreed, and left explicitly open: whether eighteen months is really the right threshold for her specific risk profile, or whether it's simply the guideline's default number applied without adjustment.
R.A. was told directly that this isn't a no — it's a timeline, with her own low-risk features already logged as the reason it's being considered earlier than a default indefinite-maintenance approach would allow.