Deprescribing a Long-Standing Antipsychotic in a Stable Adult with Intellectual Disability
Legacy-medication continuation versus a genuine taper attempt — real clinical uncertainty when the original indication for a decades-old antipsychotic prescription was never clearly documented.
Nobody currently on D.L.'s care team was working with him when the haloperidol was first started, and that fact alone has become the actual clinical problem. D.L. is 52, has moderate intellectual disability, and has taken an antipsychotic continuously since a psychiatric hospital admission in his late twenties — records from that admission, tracked down only after considerable effort, describe "severe behavioral dysregulation" without a documented psychotic diagnosis, and no note in the chart since has ever revisited whether the medication was still treating something real or had simply become a fixture nobody wanted to be the one to touch. For the past decade at minimum, by every available record, D.L. has been behaviorally stable, living in a supported apartment, working three days a week at a community garden program he has attended for eleven years.
A new physician taking over his care flagged the medication during a routine annual review, not because anything acute prompted it, but because twenty-plus years of continuous antipsychotic exposure with an undocumented original indication is exactly the pattern national deprescribing reviews have identified as needing active reconsideration rather than passive continuation. D.L. himself, when asked directly and with appropriate support to understand the question, says he doesn't remember a time before the medication and doesn't have a strong preference either way — he trusts whatever his team decides, which puts real weight on the team getting this decision right on his behalf.
What makes tracking down the original rationale so difficult isn't just the passage of time; it's that three different care teams have managed him across three different agencies since that first hospitalization, each transition carrying forward the medication itself but not the reasoning behind it, the way a photocopy of a photocopy keeps the shape of a document while losing whatever notes were written in its margins.
The clearest documented fact in the entire chart, oddly, is also the simplest one: for as long as anyone currently involved in his care has known him, D.L. has been calm, engaged at the garden program, and has not had a single incident requiring intervention — a stability so complete and so long-running that it has become almost invisible as evidence, the way a smoke detector's silence stops registering as information after enough years of nothing going wrong.
Twenty years of stability, an unclear original reason, and a new finding
The mild orofacial dyskinesia found on today's exam changes the calculus directly — that's a possible early sign of tardive dyskinesia, a real, potentially irreversible risk of long-term antipsychotic exposure that grows with cumulative years on the drug. Combined with a decade of documented stability and an indication nobody can actually confirm was ever appropriate, this isn't a borderline case for attempting a taper. It's close to the textbook scenario the deprescribing literature was written to address.
I agree with the direction but want real caution about the mechanism, not just the goal. De Kuijper and colleagues' controlled discontinuation trial in adults with intellectual disability on long-term antipsychotics found many patients could be successfully tapered without behavioral relapse — but a meaningful minority did relapse and required reinstatement, and the trial's own design used a slow, structured taper with close monitoring specifically because abrupt or rapid discontinuation carried real risk of destabilizing a functioning placement.
That's exactly the taper design I'd propose — slow, structured, with an explicit relapse-response plan agreed in advance, not a rapid stop. The dyskinesia finding argues for starting the process now rather than delaying it, not for skipping the caution built into how it's actually done.
Given his stable community placement and employment, the real risk to manage isn't just behavioral relapse in the clinical sense, it's any disruption significant enough to threaten either one. A gradual taper over several months, with his group home and employment supervisor briefed on early warning signs to watch for and report, extends the monitoring net beyond clinical visits alone — appropriate given what's actually at stake for him if this goes wrong.
Agreed: gradual structured taper initiated, with an explicit written plan for what warning signs trigger pausing or reversing the taper, shared with both his residential and employment supports.
D.L. was included directly in the conversation about what was changing and why, in accessible terms, and confirmed he understood the plan could be paused if problems arose — a step the team treated as necessary regardless of his stated lack of strong preference either way.